decoration decoration
Stories

GROKLAW
When you want to know more...
decoration
For layout only
Home
Archives
Site Map
Search
About Groklaw
Awards
Legal Research
Timelines
ApplevSamsung
ApplevSamsung p.2
ArchiveExplorer
Autozone
Bilski
Cases
Cast: Lawyers
Comes v. MS
Contracts/Documents
Courts
DRM
Gordon v MS
GPL
Grokdoc
HTML How To
IPI v RH
IV v. Google
Legal Docs
Lodsys
MS Litigations
MSvB&N
News Picks
Novell v. MS
Novell-MS Deal
ODF/OOXML
OOXML Appeals
OraclevGoogle
Patents
ProjectMonterey
Psystar
Quote Database
Red Hat v SCO
Salus Book
SCEA v Hotz
SCO Appeals
SCO Bankruptcy
SCO Financials
SCO Overview
SCO v IBM
SCO v Novell
SCO:Soup2Nuts
SCOsource
Sean Daly
Software Patents
Switch to Linux
Transcripts
Unix Books

Gear

Groklaw Gear

Click here to send an email to the editor of this weblog.


You won't find me on Facebook


Donate

Donate Paypal


No Legal Advice

The information on Groklaw is not intended to constitute legal advice. While Mark is a lawyer and he has asked other lawyers and law students to contribute articles, all of these articles are offered to help educate, not to provide specific legal advice. They are not your lawyers.

Here's Groklaw's comments policy.


What's New

STORIES
No new stories

COMMENTS last 48 hrs
No new comments


Sponsors

Hosting:
hosted by ibiblio

On servers donated to ibiblio by AMD.

Webmaster

Gordon v. Microsoft Exhibits

Gordon vs. Microsoft Exhibits

These are the exhibits used in the antitrust trial of Gordon v. Microsoft, which took place from March 18-April 15, 2004. The antitrust litigation settled during the trial, but we at Groklaw saved the exhibits first, before they disappeared. We did some group work on them, then we got distracted, and now, finally, here they are, the entire collection for us and for historians.

The numbers take you to PDFs we made of the exhibits; the links marked TIFF take you to the exhibit as it originally appeared on the Plaintiff's website. Where we have finished a text version, you will find a link marked Text.

Below the list of the exhibits, you will find text versions of the ones we have completed. Also, you will find links to the text within the list. Feel free to help us complete that task.


Gordon v. Microsoft Trial Exhibits
Jump to:Plaintiff [0100], [0200], [0300], [0400], [0600], [0800], [0900], [1000], [1200], [1500], [2100], [2300], [2400], [3200], [4400], [4500].
Respondent [0000], [0500], [0600], [0700], [1000], [2300], [3500].
Plaintiff's Exhibits
0017 Microsoft/HP license agreement, May 23, 1988 [107 pages] [Tiff]
0027 Microsoft/GO Corporation Nondisclosure Agreement, signed by Bill Gates, dates 7/8/88 [Tiff] [Text]
0035 Microsoft internal emails, one from billg: "You never sent me a response on the question of what things an app would do that would make it run with MSDOS and not run with DR-DOS." [2 pages] [Tiff] [Text]
0035A [Tiff] [Text]
0058 GO Corporation letter to Jeff Raikes, dated Feb. 19, 1989, documenting a joint project between the two companies [Tiff] [Text]
0062 [Tiff] [Text]
0068 [Tiff]
0094 IBM internal memo, dated May 3, 1989, re Microsoft "back doors" [Tiff]
0101 J.Kaplan memo to "rc", May 10, 1989, suggesting tactics to contain Microsoft access to documents [Tiff] [Text]
0108 Bill Gates 6-page memo to Steve Ballmer et al, May 18, 1989, Subject: Operating System Strategy - "The DOS gold mine is shrinking and our costs are soaring - primarily due to low prices, IBM share and DR-DOS.... Going full bore on Windows leaves the following major problems: 1. OS/2 consistent; 2. OS/2 more powerful...Government standards are a problem - I think we need to invest more people in this. I have been very successful explaining how a desktop binary standard is better than a source standard when I get to talk to the right people. The EEC meeting has me feeling better about this." [Tiff]
0109 Microsoft 10-page memo from Mark Chestnut to Russ Werner, Status Report for April 1989 [Tiff] [Text]
0115 3 pages of emails dated May 4 - June 8, 1989 reporting on Microsoft meeting with GO Corporation, on what they were doing. "The one key piece we need in order to put Windows on flat computers made by other manufacturers is the handwriting recognition software...Eventually, we will have to decide if we want to acquire this technology or build it ourselves." [Tiff] [Text]
0116 Robert Carr, Agenda, 6/9/89 [Tiff]
0117 Confidential GO Corporation memo from Robert Carr to file, with cc to Jerry K, dated June 16, 1989, Subject: "Draft Response to Microsoft Pitch for Cooperation" [2 pages] [Tiff] [Text]
0135 Bill Gates 2-page memo, August 6, 1989, Subject: OS strategy. "IBM never knew our plan and if they did they shouldn't like it..." [Tiff]
0136 Emails regarding an 1989 memo from a Far East Product Engineering product manager: "Bill Gates ordered to all application business units to include checking routines of operating environments and if it is Microsoft DOS, nothing will happen. But if it is non MS-DOS (such as DR-DOS), application will display messages saying that 'This application has been developed and tested for MICROSOFT MS DOS. Since you use different environment, this application may not work correctly..." [2 pages] [Tiff] [Text]
0141 Email asking, "The question is, do you think it is feasible to document 'DefineHandleTable' for ISV programmers to use?... I am not sure about this since Microsoft's public position has been that Apps Division programmers do not have special hooks into Windows, when, in fact, they do." [Tiff]
0161 IBM confidential internal memo from J. A. Cannavino, attached to Microsoft letter from Bill Gates, dated October 16, 1989, to Cannavino, suggesting IBM adopt a Windows strategy: "I am confused by the IBM strategy." [4 pages] [Tiff]
0162 Confidential IBM memo, dated October 18, 1989, re "Windows vs. OS/2 Comparison" with comparative charts attached. "There is no hardware platform where Windows is clearly better than OS/2." [8 pages] [Tiff]
0178 Joint Microsoft/IBM press release re expanding partnership and setting future DOS and OS/2 directions. [5 pages] [Tiff]
0181 Bob Frankston email, November 16, 1989, Subject: Update on Microsoft's view of the world. "Before leaving I dropped by Microsoft's reception for user groups. Gates summary of the IBM announcement is that 75% of the 286/386 machines shipping have 2MB or less and are thus Windows machine and that all they did was to cede a part of the remaining 25%. It wasn't an appropriate place for me to challenge him or accuse of changing the story." [Tiff]
0186 Microsoft memo from Steve Ballmer, dated December 1, 1989, Subject: Windows/OS/2 Positioning and COMDEX Events. "It is now over two weeks since Bill and Jim Cannavino reaffirmed the Microsoft/IBM relationship at Comdex... Let me clarify our positions on a few points. We will work to correct the Press and IBM's understanding of these issues." [21 pages] [Tiff]
0212 Email from cameronm to peterm, Jan. 20, 1990. "SCO is a DOMINANT Unix supplier and they make NO MONEY AT ALL. ZERO PROFIT, a very tiny cash reserve, and I will add, probably a lot lower development coests than we have. AND they have an apps business to boot!... Yes, we have to beat UNIX, but we have to beat it ONLY BECAUSE it's a long term strategic threat. It isn't a market that is profitable ...Can we back off OS/2 as the successor to DOS? How do we beat a successful retreat path? How much of the industry do we alienate? I will just assume that OEMs and IBM are irrelevant, that they are so jaded it doesn't matter what we do. What about large corporations who have bought our story, what about ISVs ho have invested in OS/2? What will the press say? I don't buy that we can hide behind this insipid claim that we've always pushed Windows and that it's their own fault. It just isn't true, we've told ISVs to develop for OS/2 and clearly put a lot more behind getting OS/2 apps than Windows apps.... Ultimately we have to think about whether we make more money... by keeping on our present course ... or whether we will make more money in repositioning Win 3 strongly on the desktop and OS/2 as the server/Unix alternative platform. I hope we realize, or that I have shown, there isn't a lot of money we can wring from the Unix market in the next couple of years.... We can't outwardly reposition OS/2 immediately (until we are ready)." [6 pages] [Tiff]
0215 MS/IBM Systems Software Plan: 1990-92 [51 pages] - "Flirtation with UNIX: some corps attracted by 'open'/standards message... govt giving leadership to UNIX movement. Currently limited issue, but could become large scale movement if viable, alternative vision not supplied.... else risk ceding share in large way to UNIX over time.... OS/2 is having mixed/poor acceptance... not differentiated sufficiently from Windows... both can't win desktop in 1991." [P. 36: AT&T System V.4 listed as having 5% marketshare of servers; <1% on desktops. P. 37: SCO System V.3.2 listed as having 1% market share on desktops, 3% on servers. P. 46 Novell NetWare listed as having 60-70% market share.] [Tiff]
0243 Internal confidential memo from J. A. Cannavino, dated 3/14/90, with attached letter to him from Bill Gates. "I am disappointed that we did not meet before IBM made a font decision.... I was disappointed to see that the AIX networking strategy conflicts with the PC networking strategy.... I was surprised to read that a group in IBM is implementing Novell protocols on the mainframe.... We have discussed the need for IBM to disclose its video strategy to us so we can plan to support your hardware aggressively....The only way I think that OS/2 will take off quickly is to get it bundled with 386 hardware." [4 pages] [Tiff]
0259 Confidential IBM memo, subject: Microsoft, dated April 9, 1990. "Attached please find several items which came out of Microsoft work sessions... Let me put one caveat on all of this material. It all assumes we continue to dialog with Microsoft on the complete 'partnership' we have been discussing. Should different direction come out of the April 11 meeting, a new strategy, agenda and 'IBM interests and concerns' position for each product will have to be developed....Microsoft is unwilling to have a partnership on DOS and Windows at all. IBM is simply to be another OEM. Microsoft does not want IBM to participate in managing these products. Throughout the history of our relationship, IBM has always played an important role in planning DOS. We cannot agree to abdicate that role now for such an important product.... Microsoft now also wants to manage OS/2 in a way that is inconsistent with the existing JDA." [12 pages] [Tiff]
0275 Internal Microsoft memo from Scott Old, dated April 30, 1990, cc to Bill Gates and Steve Baller, et al, Re: FY91 USSMD Strategic Planning: "We should create a means to ensure that our product demos and attract loops are a permanently installed on the resellers demo machines. We should capture the loyalty of the reseller sales rep via training programs and a reward system for selling Microsoft products. We will launch the Microsoft Programs Reseller in order to do the aforementioned on an on-going basis." ("Produced using Microsoft Word for the Mac and the Apple LaserWriter Plus printer.") [6 pages] [Tiff]
0285 IBM memo with attached OS/2 marketing slides. [9 pages] [Tiff]
0286 Memo to Bill Gates regarding DRI, DOS 5. [Tiff] [Text]
0293 Memo to Bill Gates et al Re "GO competition", dated May 28, 1990. "We did not convince them (Intel) not to invest in GO.... One appropriate followup is probably for us to write a letter to Grove with a follow-up phone call by Bill." [9 pages] [Tiff] [Text]
0298 Steve Ballmer letter to IBM's Lee Reiswig, dated June 1, 1990, 2 pages. "If we had distinctly different markets this divergence might not be disastrous. But there are many areas where our products will conflict." [Tiff]
0305 Microsoft Board of Directors' Report/Finance & Administration. [180 pages] [Tiff]
0310 [Tiff]
0319 [Tiff]
0349 [Abstract: [Pages 1 2 3 ] Microsoft email From gregw Jul 31, 1990 To: billg darrylr jeffr mikemap Cc: bradsi jabeb lloydfr tonyw Subject: Re: GO threat. [Page 3] of the email has a reply-included email from Bill Gates dated Jul 31, 1990.] [Tiff] [Text]
0355 [Abstract: Page 1: Microsoft Memo From: lloydfr To: billg darrylr, jeffr, mikemap, bradsi [handwritten underline] jabeb, tonyw, gregw, [s?]andyt, pradeeps, marline Subject: GO Corp Info. Date: 8/1/1990 Pages 2-14: Various screenshots of the Go PenPoint computer user] [Tiff] [Text]
0359 [Abstract: Microsoft email from lloydfr to mikemap Cc: cathyw Aug 9, 1990 Subject: GO Info. lloydfr tells mikemap to return confidential information that lloydfr had given mikemap the week before, so that lloydfr can destroy it.] [Tiff] [Text]
0370 [Abstract: Microsoft email. jeffr asks debem to print a 3-page email "for the GO file" from claraj to jeffr rice, dated Aug, 29 1990, subject: GO] [Tiff] [Text]
0373 [Tiff] [Text]
0443 [Abstract: Pages [1 2]: Microsoft email. jeffr asks debem to print a two-page email from lloydfr to billg gregs jeffr mikemap pradeeps, dated Oct 31, 1990, subject: Approaching GO] [Tiff] [Text]
0461 [Tiff] [Text]
0462 [Tiff] [Text]
0465 [Tiff]
0466 [Tiff]
0478 [Abstract: [Page 1 2 3 4 ] Jerry Kaplan notes for meeting with Jim Cannavino December 2, 1990 GO Corporation Confidential Mar 5, 1994] [Tiff] [Text]
0502 [Tiff]
0619 [Tiff]
0638 [Tiff] [Text]
0645 [Tiff]
0687 [Abstract: Page [ 1 2 ] Page [ 1 ] Fax Cover Sheet Page [ 2 ] Apr 18, 1991 Fax of a letter from Jerry Kaplan to Bill Gates. (Letter is dated Apr 17, 1991.)] [Tiff] [Text]
0690 [Tiff]
0709 [Tiff] [Text]
0718 [Abstract: Pages [ 1 2 ]: May 3, 1991 Letter from Bill Gates, Chairman, Microsoft Corporation, to S. Jerrold Kaplan, Chairman of Go Corporation] [Tiff] [Text]
0733 [Abstract: Pages 1, 2: May 13, 1991 Letter from S. Jerrold Kaplan, Chairman of Go Corporation, to Bill Gates, Chairman of the Board & CEO MICROSOFT CORPORATION] [Tiff] [Text]
0746 [Abstract: Pages 1 2 3 4 5: Mar 16, 1992 Microsoft email From nathanm To: pradeeps Cc: billg bradsi cameronm gregs jeffr jonl karenh lloydfr marline makemap raleighr robg steveb Subject: RE: PCs that aren't PCs] [Tiff] [Text]
0786 [Tiff]
0805 [Tiff] [Text]
0809 [Abstract: Page [ 1 2 ]: Jul 3, 1991 Bill Gates letter to S. Jerrold Kaplan, CEO of Go Corporation, (2 pages) Page [ 3 ]: Jul 14, 1988 Bill Gates email To: gregm jeffr joachimk jons mikemap steveb Subject: GO corporation Cc: charless jeffh tandyt (1 page) Page [ 4 ]] [Tiff] [Text]
0828 [Tiff]
0841 [Tiff] [Text]
0843 [Tiff]
0859 [Abstract: [Page 1 2 3 4 ] Microsoft Memo by Greg Slyngstad Jul 24, 1991 "Re: Challenges to Pen Win" "Threats to Windows For Pens"] [Tiff] [Text]
0868 [Tiff]
0869 [Tiff] [Text]
0874 [Tiff] [Text]
0959A [Tiff] [Text]
0978 [Tiff]
0979 [Tiff] [Text]
0981 [Tiff] [Text]
0984 [Tiff] [Text]
0985 [Tiff] [Text]
0986 [Tiff] [Text]
0990 [Tiff] [Text]
1002 [Abstract: [Page 1 ] Microsoft Email Oct 1991 Fujitsu Poquet "Go Corporation" toshim richardf joachimk jeffr "No mercy is an option for us."] [Tiff] [Text]
1048 [Tiff] [Text]
1077 [Tiff] [Text]
1079 [Tiff]
1081 [Tiff]
1130 [Tiff] [Text]
1133 [Tiff] [Text]
1142 [Tiff]
1146 [Tiff] [Text]
1183 [Tiff]
1192 [Tiff]
1219 [Tiff]
1221 [Tiff]
1222 [Tiff]
1232 [Tiff]
1243 [Tiff] [Text]
1265 [Tiff]
1273 [Tiff]
1324 [Tiff]
1394 [Tiff] [Text]
1397 [Tiff] [Text]
1413 [Tiff]
1416 [Tiff]
1513 [Tiff] [Text]
1519 [Tiff]
1520 [Abstract: Jan 13, 1993 letter from Joachim Kempin, Vice President OEM Sales, Microsoft, to Eckhard Pfeiffer, President and Chief Executive Officer, Compaq Computer Corporation, on Microsoft's desire for a "Strategic Partnership" Agreement and partnership with Comp] [Tiff] [Text]
1521 [Tiff]
1522 [Abstract: Compaq PDA Operating System Selection - Portable and Software Marketing - PC Division A January 13, 1993 powerpoint presentation arguing the risks v. return of angering Microsoft by choosing to partner with GO Corporation.] [Tiff] [Text]
1660 [Tiff]
1744 [Tiff]
1745 [Tiff]
1797 [Tiff]
1999 [Tiff]
2132 [Tiff]
2151 [Tiff]
2158 [Tiff]
2214 [Tiff]
2235 [Tiff]
2246A [Tiff]
2251 [Tiff]
2253 [Tiff]
2258A [Tiff]
2264 [Tiff]
2274 [Tiff]
2299 [Tiff]
2311 [Tiff]
2322 [Tiff]
2348 [Tiff] [Text]
2350A [Tiff]
2373 [Tiff]
2374 [Tiff] [Text]
2399 [Tiff]
2417 [Tiff]
2418 [Tiff] [Text]
2426 [Tiff]
2428 [Tiff]
2446A [Tiff]
2529 [Tiff]
2562 [Tiff]
2605 [Tiff]
2662 [Tiff]
2706 [Tiff]
2904 [Tiff]
3159 [Tiff]
3174 [Abstract: Two similar screenshots: Page [ 1 ] PenPoint computer table of contents, Page [ 2 ] Microsoft NotebookApplet] [Tiff] [Text]
3190 [Tiff]
3247 [Tiff] [Text]
3383D [Tiff]
3475 [Tiff] [Text]
3505 [Tiff]
3534 [Tiff]
3667 [Tiff] [Text]
3668 [Tiff]
3686 [Tiff]
3793 [Tiff]
3994 [Tiff]
4178 [Tiff] [Text]
4396 [Tiff]
4400 [Tiff]
4401 [Tiff]
4402 [Tiff]
4407 [Tiff] [Text]
4410 [Tiff] [Text]
4411 [Tiff] [Text]
4421 [Tiff]
4422 [Tiff]
4423 [Tiff] [Text]
4424 [Tiff]
4425 [Tiff]
4426 [Tiff]
4427 [Tiff]
4428 [Tiff]
4429 [Tiff]
4430 [Tiff]
4435 [Tiff]
4437 [Tiff]
4438 [Tiff]
4439 [Tiff]
4440 [Tiff]
4448 [Tiff]
4453 [Tiff]
4455 [Tiff]
4470 [Tiff]
4471 [Tiff]
4482 [Tiff]
4497 [Tiff]
4516 [Tiff]
4517 [Tiff]
4521 [Tiff]
4546A [Tiff] [Text]
4559 [Tiff] [Text]
4560 [Tiff] [Text]
4561 [Tiff]
4562 [Tiff] [Text]
4563 [Tiff] [Text]
4564 [Tiff] [Text]
4574 [Tiff] [Text]
4576 [Abstract: Internal IBM memo from Administrative Assistant, Office of the AGM, Market Development and Operations: Briefing Sheet for Lou Gerstner Call to Bill Gates May 26[, 1993]] [Tiff] [Text]
4588 Bill Gates 2-page memo to Intel's Andy Grove, with internal cover memo, dated May 25, 1995, regarding NSP and software standards, presenting offer to cooperate with competitive benefit for each company [Tiff] [Text]
4602 [Tiff]
 
Respondent's Exhibits
0017 [Tiff]
0071 [Tiff]
0080 [Tiff]
0082 [Tiff]
0116 [Note: same as Plaintiff 0116.] [Tiff]
0164 [Tiff]
0178 [Tiff]
0333 [Tiff]
0417 [Tiff] [Text]
0478 [Tiff]
0515 [Tiff]
0517 [Tiff]
0534 [Tiff]
0541 [Tiff]
0543 [Tiff]
0547 [Tiff]
0548 [Tiff]
0570 Nov. 1, 1989 letter from IBM VP James Cannavino to Bill Gates re OS/2 partnership concerns [Tiff]
0574 [Tiff]
0585 [Tiff]
0595 [Tiff]
0604 [Tiff]
0608 [Tiff]
0609 May 5, 1995 internal memo from IBM's Dan Lautenbach re "IBM OS.2 Warp Brand Plan" [Tiff]
0624 [Tiff]
0625 [Tiff]
0627 [Tiff]
0629 [Tiff]
0631 [Tiff]
0633 7/27/94 internal IBM memo from Neil Isford to J.M. Thompson, Subject: OS/2 (why poor sales results: "Lead by Gates and his media machine, Microsoft is spreading fear, uncertainty, and doubt to the media about OS/2 to the press, our channels, and most recently directly to our large accounts.... Influencers like Gartner are now carrying the Microsoft message...." [Tiff]
0642 John W. Thompson August 7, 1995 ISV (Solution Developer) Strategy report [Tiff]
0653 [Tiff]
0664 [Tiff]
0665 [Tiff]
0666 [Tiff]
0667 [Tiff]
0669 [Tiff]
0671 [Tiff]
0672 [Tiff]
0675 [Tiff]
0684 [Tiff]
0689 [Tiff]
0710 [Tiff]
0721 [Tiff]
0723 [Tiff]
0738 [Tiff]
0759 [Tiff]
0768 [Tiff]
0771 [Tiff]
0793 [Tiff]
0794 [Tiff]
0805 [Tiff]
0810 [Tiff]
0812 [Tiff]
1026 [Tiff]
1057 [Tiff]
1061 [Tiff]
1205 [Tiff]
1271 [Tiff]
1436 [Tiff]
1448 [Tiff]
1659 [Tiff]
1660 [See Plaintiff 1660] [Tiff]
2318b [Tiff]
2491 [Tiff]
2494 [Tiff]
2496 [Tiff]
2497 [Tiff]
2499 [Tiff]
2502 [Tiff]
2507 [Tiff]
2508 [Tiff]
2529 See Plaintiff 2529 [Tiff]
2535 [Tiff]
2639 [Tiff]
2646 [Tiff]
2673 [Tiff]
2690 [Tiff]
2706 See Plaintiff 2706 [Tiff]
3511 [Tiff]
3715 [Tiff]
3725 [Tiff]
3726 [Tiff]
3727 [Tiff]
3734 [Tiff]
3735 [Tiff]
3744 [Tiff]
3745 [Tiff]
3747 [Tiff]
4431 [Tiff]
4468 [Tiff]
 
Text transcripts

PLAINTIFF'S EXHIBIT 35 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 35 Gordon v. Microsoft p. 1 of 2

will get upset. There is a strong likelihood however that we could get rid of debug exe3bin and link. That is 60 pages and 15% of the user's ref/guide

One thing we can get rid of though is the user's guide which is geared at the user which has never used MS DOS before That's 62 pages and about 15% more of the manual right there.

Finally as far as GW BAsic [sic] is concerned we need to incorporate a quick reference or whether we like it or not we will get zillions of calls for the manual.

From: mikeswl [indistinct]
To: mikescale [indistinct]
Cc: srkamr [indistinct]; tomle
Subject: DOS 4.0 Retail Upgrade Tree
Date: Thursday, September 22, 1988 12:17PM

Srkam [indistinct] and I created a project for the DOS 4.0 retail upgrade. We put it on \\TROJAN\DOS2. The project is called 40RETAIL.

-Mike

Susan Johnson

From: billg
To: pascalm; russw; tomle
Cc: philba
Subject: Dr dos
Thursday, September 22, 1988 12:41 PM

You never sent me a response on the question of what things an app would do that would make it run with MSDOS and not run with DR-DOS. Is there any version check or api that they fail to have? Is ther[e a] feature they have that might get in our way? I am not looking for something they can't get around. I am looking for something that their current binary fails on.

This is a fairly urgent question for me and I have received nothing.

Susan Johnson

From: tomle
To: pascalm; rossw
Cc: philba; tomle
Subject: Dr dos
Date: Thursday, September 22, 1988 1:28 PM

Page 43

[STAMPED] Exhibit 7
[STAMPED] X 565988 CONFIDENTIAL
[STAMPED] MS98 0185971 CONFIDENTIAL
[STAMPED] EXH 1 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER

Plaintiff's Exhibit 35 Gordon v. Microsoft p. 2 of 2

I am assuming your [sic] handeling [sic] this one based on the info you got from AaronR pascal. Please correct me if I am wrong?

Tom

P.S. I am referring to Bill's quiestion [sic]


Susan Johnson

robertre
anthonys; davidma; vich
msSteam; rsype
NP faults in device drivers
Thursday, September 22, 1988 1:35PM

After discussing this with Ross Cook, I decided to do some research. On page 8-91 of the IBM OS/2 Technical Reference, Vol. 1 (VerifyAccess) I see the following paragraph:

Once the process has been verified as having the needed access to a specific address location, the device driver doesn't need to request access verification each time it yields the CPU during task-time processing of this process's request. If the process makes a new request, then the device driver must request access verification.

If I were reading this, I would not assume that I had to do any kind of locking of the segment. The next paragraph says

Note also that, prior to requesting the Lock on user process-supplied addresses, the device driver must verify the user process's access to the memory with the VerifyAccess DevHlp call. The device driver must not yield the CPU between the VerifyAccess and the Lock, *otherwise the user process could shrink the segment before it hsa been locked* [then, in square brackets: ] [emphasis mine]. Once the user access has been verified, the device driver may convert the virtual address to a physical address and lock the memory. The access verification is valid for the duration of the lock.

So the documentation implies that the only reason you'd want to lock a segment would be to prevent a shrink (and presumably a free). There is no mention anywhere in the VerifyAccess man page that says anything about having to protect the segment from swapping or discarding.

I suggest someone check our documentation on this.

It may be wrong, but it's documented, albeit badly. A question: can we detect if a device driver was built with 1.0 tools, or something similar? We may be able to do something to let 1.0 DOs [sic] run unhindered. I asked why this was never seen before, and Ross told me that the EE guys tend to run machines with lots of memory, so they've just never had this happen before.

Based on the above, I'm leaning towards either taking out the message or putting it under vmstrict control. Comments?

-rpr

Page 44

[STAMPED] X 565989 CONFIDENTIAL
[STAMPED] MS98 0185972 CONFIDENTIAL

PLAINTIFF'S EXHIBIT 35A Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 35A Gordon v. Microsoft p. 1 of 1

[STAMPED] Depo. Ex. 8
[STAMPED] Depo. Ex. 132

>From pasacala Thu Sep 22 14:55:51 1988
To: billg
Cc: aaronr philba russw timle
Subject: Re: DR DOS
Date: Thu Sep 22 14:47:50 1988

Here follow the three "differences" (between DR and MS DOS) that Aaron has been able to find so far. Except for these differences, the two OSs behave similarly, including documented calls.

The bottom line is that, given Aaron's current findings, an application can identify DR DOS. However, most apps usually have no business making the calls that will let them decide which DOS (MS or DR) they are running on.

Do you think differently?

[STAMPED] MATERIAL REDACTED
                         :
This is the list of differences aaron was able to find.
The DR DOS BOOT RECORD is different. It contains the OEM ID string "DIGITAL" in it.

Undocumented DOS system call 52H returns a pointer to an internal DOS structure known as the "sysinit variables". The DR DOS structure does not match well with MS-DOS:

SysInitVers STRUC
SYSI_DPS DD ? ; DPS chain
SYSI_SFT DD ? ; SFT chain
SYSI_CLOCK DD ? ; CLOCK device
SYSI_CON DD ? ; CON device
SYSI_MAXSEC DW ? ; maximum sector size
SYSI_BUF DD ? ; buffer chain
SYSI_CDS DD ? ; CDS list
SYSI_FCB DD ? ; FCB chain
SYSI_Keep DW ? ; keep count
SYSI_MMIO DS[?] ? ; Number of block devices
SYSI_MCDS DS[?] ? ; number of CDS's
SYSI_DEV DD ? ; device list
SysInitVers ENDS

SYSI_DFS == 0:0 on DR DOS, never see this on MS-DOS
SYSI_BUF == 0:0 on DR DOS, never see this on MS-DOS
SYSI_CDS == 0:0 on DR DOS, never see this on MS-DOS
Undocumented DOS system calls 12H and 1FH return a pointer to an internal DOS structure known as the "Drive Parameter Block". On MS-DOS all of the DFSs are linked together into a dword linked list. The DR DOS DFSs all have FFFF:FFFF is [sic] the link field and do not form a linked list (this is consistent with the fact that SYSI_DFS == 0.0).

[STAMPED] EXH 2 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS-PCA 1175720 CONFIDENTIAL
[STAMPED] X0196084 [or X01960S4, intelligible]
[STAMPED] PLAINTIFF'S EXHIBIT 26 C.A. No. 2:96CV645B

PLAINTIFF'S EXHIBIT 109 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 1 of 10


[STAMPED] MSC00474985
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] EXH 3 DATE 5/17[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS-PCA 1175831 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT 51 D.A. No. 2:96CV645B

Microsoft Corporation [address/tel/telex/fax]

[MICROSOFT branding mark]

Microsoft Memo

TO: Russ Werner

FROM: Mark Chestnut

SUBJECT: Status Report for April, 1989

DATE: 5/22/89

The following summarizes major product marketing activities this past month. A summary of the current status of DOS group projects is also attached.

ROM DOS

This was a great month for ROM DOS, as two very high volume potential OEMs who were on the verge of signing with DRI both committed to MS ROM DOS. Vendex/Hcadstart, which is launching a very low cost 8088 machine for distribution through the mass merchant channel, committed to 250K units of ROM DOS and $1 million for the first year. Emerson, which is introducing a full line of low cost PCs (8088, 286 and 386SX), also for distribution through mass merchants, also committed to 250K units and $1.75 million for the first year.

Both Vendcx and Emerson were planning to sign with DRI because it did not appear that we could deliver ROM DOS in the very short term, which is when they needed it Closure of both of these deals was made possible by our decision to do a ROM copy version of DOS 3.31 with large disk partition support, for shipment by the end of this month. Thanks to Tom Lennon and the entire DOS team for moving very quickly and making it possible for us to get this business.

Good progress was also made in getting Intel to move closer to licensing ROM DOS for their Wildcard product I told Intel about our short term plans for ROM copy DOS, and they have agreed to reference sell this with the Wildcard for the time being. Once we have our ROM executable version done, indications are that they will license this from us and distribute it with the Wildcard.

Philba, Tomlc and I also had a meeting with Bruce McCormick, Intel's marketing guy for Flash EPROM. Intel is really pushing their FEPROM technology (they claim to be 2 years ahead of the Japanese), and are actively trying to get OEM design-ins. A big plus for them towards this end is support in DOS for FEPROM which we happen to be working on and will have completed in a couple of months. Intel is very interested in doing some type of joint PR announcing the availability of the FEPROM file system for DOS from MS and Intel's next generation of FEPROM products (target late summer timeframe). I proposed that, at the same time, we announce our ROM executable DOS and that Intel is licensing it and distributing with the Wildcard. Intel is very receptive to this. I will be taking Rich Freedman (our summer intern who will focus on the embedded market) to meet with Intel in early June to discuss this further. Getting Intel to license ROM DOS for the Wildcard would be a big win, and the joint PR could give us a great head start in attacking the embedded market

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 2 of 10

RUP

The RUP plan was revised this month so that some additional utilities could be added. Final release is now targeted Tor Sept 30. We .have looked closely at several utilities and feel at this point that we can prqbably include the following in time for a 9/30 release:

  1. QuickHelp
  2. command line editor
  3. undelete command
  4. EMM386 with "loadhigh" option

We also looked very closely at doing "Extended Where", a Magellan-type disk organizer that would be integrated with the shelL It would also have an extensible architecture that would allow for ISVs to easily develop "viewers" (ability to view the files created in a specific application in native form) for their applications. This would be a major advantage over Magellan, which is a closed system (Lotus has to develop the viewers for each application that Magellan supports, resulting in limited app support). We ultimately concluded that we can't do Extended Where for a 9/30 release, but we'll continue to work on it for RUP version 2.

There are some issues with the RUP which need to be resolved. They include:

  1. 3rd party peripheral support. There are concerns about our ability to support weird mass storage devices, WORM drives, etc
  2. OEM support We are currently lacking cooperation from NEC and Toshiba, which represent several hundred thousand machines in the US.
  3. Cost to support the product through PSS.

We will be putting together a plan to address the above and will present this along with our recommendations to Steveb on 5/25.

DOS/Win

The DOS group's action items from the last IBM DOS/Win meeting have been completed, and IBM is working towards presenting a plan for DOS/Win to Hanrahan on 6/7. Tomle is coordinating our end of this and is working on a plan designed to gain MS development control over the DOS portion of DOS/Win.

DOS 4.01

DOS 4.01 continues to gain momentum, as several more OEMs began shipping DOS 4.01 in April The list of OEMs currently shipping now includes:

  • Compaq
  • Unisys
  • Olivetti
  • Siemens
  • Amstrad
  • Intel
  • Dell
  • Compuadd
  • Phoenix

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 3 of 10

Another big boost to DOS 4.01 momentum will come in the first part of June, when we expect to issue a "DOS 4 works great with networks" press release. The release will announce 3 COM's shipment of the 3+ redircctor that works with DOS 4, and will also include favorable statements from Novell and Banyan that their products work fine with DOS 4, etc. All three companies have agreed to participate, and I will be working with Marianne Allison to get this thing finalized over the next couple of weeks.

PRI Competitive Response

The first MS product with the nonnested DOS warning code. Quick Pascal, was released. Tom Reeve and Cindy Kasin have committed to implementing it in all new MS application and language releases from this point forward, including international.

I am also planning to hire an independent DOS guru to do an in-depth comparative analysis of MS vs. DR DOS, with the idea of somehow making those results available to the press. This could be useful ammunition to have against DRI, and will be of value even if we choose not to make it public. I approached Ray Duncan about doing this, but he finally said "no thanks", so I am now talking to Rick Wilton, another DOS guru who writes for MS Press.

DOS Royalty and Packaged Product Business

DOS OEM Royalty, Domestic and International

Royalty business continues to be very strong, with both domestic and international OEM well above budgeted unit shipments for the year (see attached spreadsheet and charts). Domestic revenue is only 92% of budget, however, primarily due to the effect of several OEMs having prepaid balances (meaning that they previously paid royalties on units shipped this year, which explains why units shipped exceeds revenue recognized for the year). .

International royalty revenue is 174% of budget, so overall DOS royalty revenue is $19 million ahead of budget year to date.

DOS Packaged Product, Domestic and International

Domestic packaged product shipments continue their upward trend for April was the best month this year with over 22K units shipped. International packaged product shipments were down in ApriL Both domestic and international packaged product shipments for April show an increase in DOS 3-3 shipments relative to DOS 4.01 shipments. This is because DOS 4.01 shipments were temporarily suspended in April until the Amstrad bug was fixed and all inventory re-worked. I therefore expect DOS 4.01 shipments to continue to increase relative to 3.3, as was the trend in the months prior to April

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 4 of 10

Performance Against April Objectives

1. Get commitment from Vendex to license ROM.DOS

Done.

2. Get commitment from Emerson to license ROM DOS

Done.

3. Finalize plans for utilities, online doc, etc. to be included in RUP

Progress.

4. Set up European trip with MS subs, get all info needed for RUP support of European OEMs.

Turned over to Tomle.

5. Reach agreement with IBM on functional spec and plan for 6/90 seamless product

Progress.

6. Finalize ROM DOS business plan

Done.

May/June Objectives

1. Finalize and issue "DOS 4 works great with networks" press release

2. Finalize plans for utilities, online doc, etc. to be included in RUP

3. Ship ROM DOS 1.0

4. Get closure on Intel licensing ROM DOS for distribution with Wildcard

5. Bring Rich Freedman on board, get him going on embedded market analysis

6. Get closure with DOS guru on MS/DR DOS comparative study


Plaintiff's Exhibit 109 Gordon v. Microsoft p. 5 of 10

MS-DOS year-to Budget vs Actual

Domestic OEM, Royalty

Q1 Budget Q1 Actual % Budget Q2 Budget Q2 Actual % Budget Q3 Budget Q3 Actual % Budget YTD Budget YTD Actual % Budget
Units 484,200 769.800 159% 726.700 817,900 113% 760.900 1,054.300 139% 1.971,800 2,642,000 134%
Revenue $12.658,000 $12,958,000 101% $14,184,000 $14,337,000 101% $20,721,000 $16,638,000 80% $47,763,000 $43,933,000 92%
 
Domestic OEM, Packaged Product

Q1 Budget Q1 Actual % Budget Q2 Budget Q2 Actual % Budget Q3 Budget Q3 Actual % Budget YTD Budget YTD Actual % Budget
Units 35,200 8,704 25% 46.100 23,200 50% 40,300 52.024 129% 121.600 83,828 69%
Revenue $2,146,000 $543,000 25% $2,812,000 $1,456,000 52% $2,457,000 $2,643,000 108% $7,415,000 $4,642,000 63%
 
International OEM, Royalty

Q1 Budget Q1 Actual % Budget Q2 Budget Q2 Actual % Budget Q3 Budget Q3 Actual % Budget YTD Budget YTD Actual % Budget
Units 595,600 1.191.100 200% 636,500 $1,217,000 191% 778.500 2,422,400 183% 2,010,600 3,838,500 191%
Revenue $12,845,000 $15,613,000 122% $14,506,000 $19,613,000 135% $18,264,000 $20,933,000 115% $32,326,000 $56,159,000 174%
 
Int'l OEM. Packaged Product

Q1 Budget Q1 Actual % Budget Q2 Budget Q2 Actual % Budget Q3 Budget Q3 Actual % Budget YTD Budget YTD Actual % Budget
Units 17,285 14.503 84% 23,645 27.112 115% 24.249 30.369 125% 68,646 71.984 105%
Revenue $1,420,436 $953,224 67% $1,949,640 $1,937,190 99% $2,013,299 $2,126,057 106% $5,383,375 $5,016,471 93%

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 6 of 10

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] MSC 00474990
[STAMPED] MS-PCA 1175836 CONFIDENTIAL
[The exhibit page is a bar graph. Approximate numerical values have been eyeball-estimated from the bargraph for purposes of this transcript.]

Worldwide DOS OEM Royalty Revenue, FY [19]89 Actual vs Budget

Q1: Budget $25,000,000 Actual $30,000,000
Q2: Budget $30,000,000 Actual $35,000,000
Q3: Budget $41,000,000 Actual $40,000,000
YTD: Budget $97,000,000 Actual $102,000,000

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 7 of 10

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] MS-PCA 1175837 CONFIDENTIAL
[STAMPED] MSC 00474991
[The exhibit page is a bar graph. Approximate numerical values have been eyeball-estimated from the bargraph for purposes of this transcript.]

Worldwide DOS OEM Packaged Product Revenue, FY [19]89 Actual vs. Budget

Q1: Budget $3,800,000 Actual $1,700,000
Q2: Budget $4,900,000 Actual $3,800,000
Q3: Budget $4,200,000 Actual $4,900,000
YTD: Budget $12,900,000 Actual $9,000,000

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 8 of 10

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] MSC 00474992
[STAMPED] MS-PCA 1175838 CONFIDENTIAL
[The exhibit page is a bar graph. Approximate numerical values have been eyeball-estimated from the bargraph for purposes of this transcript.]

International DOS Packaged Product Units Shipped
[Presumably, 3.3 Refers to DOS Version 3.3. 4.01 Refers to DOS Version 4.01]
Jan: Version 3.3: 2,000 Units. 4.01: 7,400 Units.
Feb: Version 3.3: 1,900 Units. 4.01: 7,900 Units.
Mar: Version 3.3: 2,000 Units. 4.01: 10,000 Units.
Apr: Version 3.3: 3,700 Units. 4.01: 5,000 Units.

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 9 of 10

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] MSC 00474993
[STAMPED] MS-PCA 1175839 CONFIDENTIAL
[The exhibit page is a bar graph. Approximate numerical values have been eyeball-estimated from the bargraph for purposes of this transcript.]

Domestic DOS Packaged Product Units Shipped
[Presumably, 3.3 Refers to DOS Version 3.3. 4.01 Refers to DOS Version 4.01]
Jan: Version 3.3: 4,300 Units. 4.01: 8,100 Units.
Feb: Version 3.3: 8,000 Units. 4.01: 12,000 Units.
Mar: Version 3.3: 8,100 Units. 4.01: 11,900 Units.
Apr: Version 3.3: 11,900 Units. 4.01: 11,800 Units.

Plaintiff's Exhibit 109 Gordon v. Microsoft p. 10 of 10

DOS Group Projects

5/22/89

Project Status Major Issues
 
1. ROM DOS - Development on schedule for July release
- FEPROM development on schedule
- Vendex, Emerson deals closed
- Business plan complete
- Need to do 3.31 level ROM executable
 
2. RUP - Stevcb meeting 5/25 to finalize utilities, etc.
- Development finishing sizing of utilities
- European status: some OEMs have agreed to cooperate, others needed
- Corpcomm communications plan complete
- Marketing Plan complete
- Initial BOM complete
- Need to plan European OEM trip for June - PSS support issues - Wang, Zenith issues
 
3. DOS/Win Merge - DOS group action items complete
- IBM presenting to Hanrahan 6/7
- Externals group work items not complete
 
4. Non MS/PC DOS warning code - Code complete and fully tested
- Code shipping with QuickPascal
- Apps, languages, int'l program mgrs. have committed to incorporating into all MS products from this point on
 
6. Packaged Product - packaging to be changed in June to reduce COGS by ~$3 (GW Basic manual to be removed)
 
7. DOS 4 Maint - no resources to address other reported bugs currently assigned - Several potentially severe bugs need attention
 
8. LIM Issues - MS involvment in Quarterdeck response complete
- Awaiting Intel proposal for LIM 4.1



PLAINTIFF'S EXHIBIT 136 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 136 Gordon v. Microsoft p. 1 of 2

year old for a day when my regular sitter couldn't one day last month. she lives on 156th just north of Redmond campus. 5 minutes away. she called last week and asked me to let people know she was available periodically. she lives in an apt., has 2 kids (4 and 7), great outdoor play areas. my daughter had a great time there. you might give her a call ([telephone number]) let me know if it works out.
theresa

>From bjbank We Aug 9 15:31:08 1989
To: tomle
Cc: bclee bjbank yasukim yongchi
Subject: dos clone check on windows
Date: Fri Jul 03 13:52:46 PDT 1992

Hi Tom,

I need your help again.

MSH of Korea is localizing windows 2.10 and they want to make it check whether the system is real MSDOS or not. If it is it not, it should generate a warning message. I talked to philba and davidw, and davidw said you could help us.

If you read the following emails, I think you will understand our current situation.

Thanks.
bjbank
-------

>From davidw Wed Aug 9 14:04:57 1989
To: bjbank
Cc: jodys philba tomle
Subject: dos clone check on windows
Date: Wed Aug 09 14:01:11 1989

Windows has no such code in it. Since we do not have any code that does this we put up no warnings. However our testers have informed us that we crash while booting Windows if running under some dos not our own.

Tomle can give you checking for a warning message.

>From bjbank Wed Aug 9 13:10:07 1989.
>From philba Wed Aug 2 15:53:18 1989
Sender: jodys Wed Aug 2 15:47:16 1989
Sender: bjbank Wed Aug 2 15:41:33 1989

Hi, Jody.

I am a project manager at Far East Product Engineering group. My main responsibility is to support far east subsidiaries (MSKK and MSCH) technically. You have been great help to us and we really appreciate it. Recently, I have an inquiry from Korea about the windows product.

Plaintiff's Exhibit 136 Gordon v. Microsoft p. 2 of 2


The question from MSCH is "How to check the DOS is MS-DOS or clone". MSCH wants to include such routine in Hangeul Windows so that Hangeul windows can run only on Hangeul MS-DOS.

Could you tell me to whom I can ask to resolve this problem?

Thanks

bjbahk/Far East Product Engineering Group

-------------
Sender: yongchi Thu Jul 20 18:08:50 1989
To: bjbahk
Subject: DOS clone check on Windows
Cc: bclee dkkim ischoi ivys sangc ucmoon yhjeon
Date: Thu Jul 20 18:08:49 1989

Hello BJ,

We will finish our Windows project before October. Now I must decide our product retail Windows sepc. Would you please let me know how Windows(or other Apps too) check current dos is MS-DOS or MS-DOS clone ?  And if it is MS-DOS clone how original WIndows [sic] handle them - stop windows, warning messages ?

Thanks,   yongchi

-------------

>From tomle Wed Aug 9 16:06:08 1989
To: ibmboca?b391747
Cc: philba tomle
Subject: Busmaster interface review
Date: Fri Jul 03 13:52:50 PDT 1992

I understand the busmaster interface proposed by Ralph Lipe is going in front of the architectural review board tomorrow (8/10). I would like to talk to you for a moment to find out if you are on that board as well as getting your impression of the appropriate process to follow. We have not had any previous dealings with the architecture board and so are uncomfortable with the lack of any crisp process to gain acceptance of the interface.

If you don't object, I would like to call you first thing in the morning to talk about this.

Thanks,
Tom

>From billg Wed Aug 9 16:32:07 1989
To: markcl
Cc: darrylr paulma tomle
Subject: Extended attributes
Date: Fri Jul 03 13:52:50 PDT 199

[document footer:]
WinMail  1.21  brucen  Fri Jul 03 13:42:36 199Page: 150

[STAMPED] X 571035 CONFIDENTIAL
[STAMPED] MS-PCA 1148521 CONFIDENTIAL


PLAINTIFF'S EXHIBIT 286 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 286 Page 1 of 1]

[STAMPED] Depo. Ex. 822

>From russw Wed May 16 07:36:52 1990
To: billg steveb
Subject: I have asked philba to spend a lot of time making dos 5 happen fast
Cc: ericst jeremybu joachimk markche richardf
Date: Wed May 16 07:34:40 1990

given the aggressive stance taken by dri this week, I have asked philba to spend more of his time on dos 5 than on win 3.1 planning -- like 60-70% to make sure this product happens as fast as it can and that it checks out and to get the troops pumped up.

we are also planning some more extensive pr around our beta ship ( in about 3 weeks) to get acros the message that the product is a now thing vs. a later thing -- dri is spending time saying that our dos 5 won't be around for a long time.

also, I believe that their ems and load-hi modes are incompatible with Win 3.0. we are trying to verify this by the announce timeframe.

[STAMPED] PLAINTIFF'S EXHIBIT 118 C.A. No. 2:96CV645B
[STAMPED] XO196018

Plaintiff's Exhibit 841 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 841 Page 1 of 4]

File ; c:\bradsi\novell.fld
Messages ; 1 - 19

######################################################## 1
>From davidcol Wed Jul 17 08:46:49 1991
To: bradc bradsi gregle philbe richt russs tomle
Subject: novell
Date: Wdd Jul 17 08:46:14 1991

I think we should use Windows to get a Microsoft OS back onto the Netware clients which will bundle or require DR. DOS. We should alter our plans a bit and move all the DOS 6.0 improvements directly into Windows. When the user starts Windows, they get the Microsoft OS (including networking) and all the other cool features that go with that. When they quit, they get Netware and DR DOS and no Windows apps. The key is getting a piece of MS system software on that client so we can deliver our stategy and vision. We can leverage Windows and Windows apps to to this.

We should not consider things that stop Windows from working on Netware. (Netware here = netware + DR DOS.) If it was just DR DOS alone, then we should prevent Windows from working there. Netware has too much market share and too many customers are loyal to it for us to exclude windows from that market.

I think this dictates that we maintain good relationships with Novell so we can stay abreast wjth what they are doing at the detailed technical level. However, I do think that we should do our own winnet drivers and other Novell provider components. If we are going to take over the desktop when Windows starts, it MUST be all Microsoft written software since Novell won't help us do that.


######################################################## 2
>From jimell Wed Jul 17 08:54:44 1991
To: billg bradsi mikemur paulme steveb
Cc: jimell
Subject; Fw: Novell/Digital Research reach definitive agreement...
Date: Wed Jul 17 08:53:48 PDT 1991

I thought about it all night. Since I came here I said there were two things that concerned me related to Novell: one Novell partnering with IBM and two Novell coming at us at the desktop. Both fears have now come true.

I had planned to call Novell this morning and ask them (I know lots of people there) how they plan to position, etc. But the release below makes it pretty clear.

Given this, I suggest (at least for Systems) that we

a) do not change our public posture

b) do not appear alarmed .. "that's interesting or curious or something" would be a good quote.

c) we get serious about supporting our networking more than others. This is the strategy that Novell will adopt as soon as they have a product to do so. They will of course continue to support "foreign" OSs like Windows and DOS, but their focus will be on building a new generation OS for the desktop.

I understand the difficulties of working with ISVs to get them to change their support from Windows, but we shouldn't kid ourselves. I have been fighthing Novell for years -- we shouldn't underestimate their technical or marketing abilities. Their intentions are clear.

Our posture should be the same. We support out networking and also "foreign" servers such as Novell, etc.

d) I suggest that we only include our software on the diskettes for Windows 3.1 and future DOS versions. We can be open without doing this. (Note that doing Netware and LM dual redirectors and the service provider environment are still critical components of the system.) (We ship compatibility software to other formats -- we should do the same for networking. We don't ship parts of 123 or other products in Excel. I never have agreed with this strategy and still don't. Actually for the current set of products and given where LM has been, that was probably the right strategy, but we must change this for the future.)

e) I suggest that we include winball into 3.1 as a standard feature as soon as its ready with minimal price change.

f) We should provide some part of LM as a standard packaged part of NT. Support pricing must be worked out.o

g) We consider changing our apps to not run unless the OS is our OS.

h) We must leverage our networking and OS strengths together to win this battle. Not doing so already has hurt the networking business and if the release is any indication, it will hurt the OS business in the future.

jlm
| >From rooternef Wed Jull 17 08:12:40 1991
| To: execnews
| Cc: sharonb
| Subject; Novell/Digital Research reach definitive agreement...
|
| Date: Wed Jul 17 08:02:56 1991
|
|
| NOVELL DGTL RESEARCH: Novell and Digital Research sign definitive merger
| agreement
|
| July 17, 1991
|
|
|     PROVO, Utah--(BUSINESS WIRE)--Novell Inc. (NASDAQ:NOVL) and
| Digital Research Inc., a developer of advenced operating system
| software. Including the first DOS 5 operating system, Wednesday
| signed a definitive agreement to merge the two companies, making
| Digital Research, headquartered in Monterey, Calif., a wholly-owned

[Plaintiff's Exhibit 841 Page 2 of 4]

| subsidary of Novell.
|     Under the terms of the definitive agreement signed yesterday,
| existing shares of Digital Research common stock, convertible
| securities and options will be exchanged for $1.5 million newly
| issued shares od Novell common stock.
|     Digital Research is a private company incorpotated in the state
| of California, It is the originator of CP/M, the precursor of
| today's disc operating systems and DR DOS.
|     "Our strategy is to provide computer users and industry partners
| with easier to use, more powerful, software products to support tight
| integration between desktop computers, computer networks and host
| computer systems," said Ray Noorda, Novell's president and chief
| executive officer.
|     "Digital research is an important part of this strategy, Novell
| is welcoming a talented organization with technology leadership not
| only in DOS operating system products, but also forward looking
| expertise in multi-tasking and graphical user interface technology."
|     Novell said it is responding to customer demand for tightly
| coupling network operating system software with desktop and host
| computer operating systems. Novell has already become the largest
| outside investor in UNIX System Laboratories, the developer of UNIX
| System V release 4.
|     With the merger with Digital Research, Novell is adding DOS,
| multi-tasking and real-time operating system technology. The
| combined resources are seen as providing a dynamic technology
| platform for better integrating DOS, UNIX and Netware operating
| system environments.
|     In addition, the company will continue to develop innovative
| software products to support customer utilization of OS/2, Windows,
| Apple Macintosh and other operating system environments.
|     Dick Williams, president and CEO of Digital Research, said the
| merger gives Digital research significant new market reach through
| Novell's relations with leading computer vendors, its presence in the
| systems integration market, and far reachind distribution, marketing,
| education and customer support resources.
|     "We have long understood the value to customers of significantly
| extending the capabilities of DOS. Digital Research and DR DOS have
| already set the standard for DOS capabilities in the 1990s. Tightly
| integrated products from Novell and Digital Research will simplify
| network use and better support our mutual customers and industry
| partners."
|     "With novell, we see ourselves as supplies of total, operating
| system solutions for the enterprise computing environment, from
| banking to industrial automation real-time requirements, to advanced
| graphical user interface technology," Williams added.
|     Digital Research brings two new software engineering centers to
| Novell. In Monterey, Digital develops graphical user interface
| technology and FlexOS, a real-time, multitasking, multiuser operating
| system for the Intel family of microprocessors.
|     FlexOS combines general purpose operating system ease-of-use with
| real-time, transaction oriented capability necessary for
| point-of-sale, and industrial and process control systems. Original
| equipment manufacturers who deliver FlexOS with their systems include
| FANUC, IBM, ICL, TEC and Siemens.
|     In Hungerford, Berks, the United Kingdom, Digital develops its
| general operating system family of products, including DR DOS, DR
| Multiuser DOS and Concurrent DOS. DR DOS represents between 10 and
| 15 percent of the overall DOS market, is translated into major
| languages, and is sold to end-users worldwide through both software
| distribution channels and more than 200 OEM vendors.
|     The merger agreement has been approved by the boards of directors
| of each company, but remains subject to the approval of Digital
| Research stockholders, regulatory approvals and other normal conditions
| to closing. Certain digital Research stockholders, including each of
| its directors, have signed agreements to vote their shares in favor of
| the merger, which is expected to be completed in October 1991.
|     As part of the agreement, Digital research has agreed to pay a fee
| in the event the merger fails to be completed due to a vote of their
| shareholders, or a change in recommendation by the Digital Research
| board. The merger agreement is expected to be accounted for as a
| pooling-of-interests.
|     Digital research, with 273 employees, had revenue of $40.9 million
| in its fiscal year ended Sept. 30, 1990, up from $36.2 million in
| fiscal 1989.
|     Novell, Inc. is the leading providor of network server operating
| system software that integrates desktop computers, servers, and
| minicomputer and mainframe hosts for information sharing.
|     Novell's NetWare network computing products manage and control
| the sharing of services, data and applications among PC workgroups,
| departmental networks and across business-wide information systems.
|
|
| CONTACT: Novell Inc.
| Peter Troop, XXX/XXX-XXXX
| or
| Digital Research
| Joe Taglis, XXX/XXX-XXXX


######################################################## 3
>From philbe Wed Jul 17 08:S9:13 1991
To: bradsi
Subject: ReL novell
Date Tue Jul 16 08:47:52 1991

I think we need to carefully measure our ewsponse. Totally freezing them out will force them to compete on a wider basis and could cause the disaster scenario to occur. Not to mention the thin FTC ice that we would be on. On the other hand, we have an advantage with windows and should press it in the system arena. One approach would be to deny enhanced mode to all but real-mode dos customers. This can be to move kernel (or major components of it) into a VxD and thus significantly up the cost of entry to running 'full' windows. This has the 'side effect' of disallowing enhanced mode under a dpml server and not running the more advance components except in enhanced mode. It is, however, a bunch of work -- my guess is 1+ MM.

As for denying them access to windows info, I think this is something we should not do. Having shitty windows support in netware will only hurt us.

| > From bradsi Tue Jul 16 23:33:12 1991
| To: bradc davidcol greglo philbe richt russs tomle

[Plaintiff's Exhibit 841 Page 3 of 4]

| Subject: novell
| Date: Tue Jul 16 23:32:44 1991
|
| this novell thing bugs me. we need to think through what kind
| of relationship we can/should have with novell now
| for dos, windows, dos6 and win32. i'm
| sure jimall and stevea are thinking the same for winx/win4.
| send me your thoughts on novell, I'll pass along
| some of mine. I want to put together recommendations
| and present to other execs.
|


######################################################## 4
>From tomle Wed Jul 17 09:29:18 1991
To: bradsi
Cc: tomle
Subject: Re: novell
Date: wed, 17 Jul 91 09:28:03 PDT

I spent time last night since I couldn't sleep very well. Her is what I came up with:

Why would Novell do this?

This seems like an odd move if they really understand the position we were placing them in, why would they make the move gauranteed to get us the most angry?

1) Their greedy. They see a marvelous opportunity to take Microsoft's position in the systems market. With IBM and MS going at it, it would be a great time for Novell to leverage their Net dominance, their current relationship with IBM and their advantage in Corporate loyalty to muscle our cash cow away from us.

2) They are planning to use the DR-DOS threat to force MS to do what Novell wants in both Dos and Windows. My understanding from Kaveen and Terry is that Novell has a strong Windows strategy so unless they are moving away from that strategy, this move somehow has to make sense in a Windows strategy.

3) Both of the above.

4) They think that offering a complete Novell solution gives them an edge in the corporate environment. You must question, an edge over what? Can they achieve this knowing that Windows will be everywhere?

The really scary part for me is that we were in the infancy stages of developing a relationship with Novell so we know very little. The first question I want to try to answer is, given Novell control over Dos, what would they do to it? My strategy would be to release Dos products very early on that hit key features on the list of things we think Novell would do.

How do we deal with Novell now that this has happened.

1) Windows gets another reason it must succeed. Because of that I strongly recommend the Windows group NOT close off to Novell. Windows must execute on the strategy of being the best network workstation environment and Novell is the key player there. I have to trust the windows folks to be in bed with my enemy while still being on my side. This is exactly what i asked of the lanman guys in Dos 5.00 and it worked.

2) Dos can't be that close yet. I am sure it was a little disconcerting for the Novell guys to start talking to the Dos group but we gave them assurances and we were building trust. We need assurances from Novell that these groups are maintained separatly and that their NDAs are meanigful before the Dos group can continue developing it's Novell relationship.

Strategy for dealing with Novell DR-Dos

I am assuming they will take over the Dos market if they can and if they can't they will force some control over Windows directions.

DR-Dos 6.00 is real and is in beta. If I am right about what they are doing, they will release DR-Dos 6.00 before adjusting the DR-Dos priorities on a Network focus. We can't let Novell take the lead in the Dos technology market. This will bw a real challange for us but one I think we have to meet.

We must convice Novell there is real value in Windows and Dos having a close intimate relationship.

Tom

| > From bradsi Tue Jul 16 23:33:12 1991
| To: bradc davidcol greglo philbe richt russs tomle
| Subject: novell
| Date: Tue Jul 16 23:32:44 1991
|
| this novell thing bugs me. we need to think through what kind
| of relationship we can/should have with novell now
| for dos, windows, dos6 and win32. i'm
| sure jimall and stevea are thinking the same for winx/win4.
| send me your thoughts on novell, I'll pass along
| some of mine. I want to put together recommendations
| and present to other execs.
|


######################################################## 5
>From russa Wed Jul 17 10:16:16 1991
To: bradsi
Subject: Re: novell
Data: Wed Jul 17 10:15:07 PDT 1991

My initial thoughts on this are that:
1) this is a strategic mistake for Novell
2) If they change their product strategy mush (ie. build net products that are proprietary to DR DOS) it will be to their detriment
3) they have acquired some programers that will make them more agressive (and successful?) at GUI work
4) the biggest danger for us that they try to be very agressive with DR DOS pricing
5) that competition is not all bad: it may have some shirt/longterm effect on our profits, but it will get us all that much more focused
6) relationship-wise we do have to tread lightly with them

[Plaintiff's Exhibit 841 Page 4 of 4]

In a nutshell, I don't see why it will be an easier for them to enter our market (DOS) as it was for us to enter theirs (Netware). I suspect if they try a frontal assault on the DOS/desktop os biz it will cost then big $'s and they will not succeed. On the other hand if they use it as a platform for "proprietary" net products (ie. that don't run on DOS 5.0/6.0) they will be shooting themselves in the foot. And if they offer 2 versions, a "better one" on DR DOS and a "normal one" on DOS 5/6, I think the market will shrug. If they try to beat us to mky with a 32 bit DOS, again, I think the market will shrug, and this might complicate their relationship with IBM. Product wise in desktop or market, if we execute, then I don't see what they gain.

The flip side of this is, I don't see why we would change our product strategy to somehow "exclude" them, which would be suicide from a net mkt perspective. From a winball perspective, I don't think it changes our Netware plans. We still have to co-exist with installed NetWare LANs, and if we can get them to sign up for licensing us the redir/xport, we'd still like to make it easier for the NetWare customer to install winbell/windows on a NetWare LAN. This is good for them, but good for us.

As far as our relationship with them, I think there are some open issues that we need to think through carefully. We obviously need them to continue to support Windows (including winbell), so if they come to us and ask to help them write VxD's etc, I think we have to help. On the other hand, what if they come and want help getting their new peer to peer product to co-exist with winbell? It's windows, but it's a directly competitive product. This is the sticky territory where I think we are bound to have to work with them, but it will be difficult for both companies. One way to look at it is, for the first time we are looking out at a company that is in 2 businesses and that causes conflict for us because we compete with them and need to be friends with them at the same time. Of course this is exactly what companies like Novell have been faced with in dealing with MS for so long. We'll just have to continue to tread the fine lines. Both companies need each other too much.

The 2 biggest residual effects of the deal are that they can cut the price of DR DOS If they choose to, which might cause us some short term price pressure, and they have apparantly acquired some good GUI programmers. This will probably make them more aggressive about graphical interfaces throughout their product line, which will be good for them.

######################################################## 6
>From russe wed Jul 17 11:43:41 1991 To: bradsi
Subject: RE; Novell/Digital Research
Date: Wed Jul 11:44:38 PDT 1991

Between you and me, I don't see what has changed that would make us reconsider this. I hope no one at MS, including myself, gets defocused reconsidering this transaction just because Adrian is at Artisoft now. We are embarked on a dev course that makes sense for and we should be single minded in pursuing it. The decision to do all 32 bit VxD work is looking better everyday. I bet that's why Adrian wants to re-ignite this merger (because he knows that Artisoft probably can't do it on their own), and it will also give us an advantage against whatever peer product Novell comme up with (I assume they are not ding VxD work since they have not come to us for help.

| >From bradsi Wed Jul 17 11:07:00 1991
| To: russs
| Subject: Novell/Digital Research
|
| Date: Wed, 17 Jul 91 11:04:49 PDT
|
| fyi but not for forwarding or discussion with others.
|
| | >From mikemur Wed Jul 17 09:20:55 1991
| | To: billg steveb
| | Cc: bradsi jimll paulme
| | Subject: Novell/Digital Research
| | Date: Wed Jul 17 08:20:48 PDT 1991
| |
| | Adrian King called this morning from his new office at Artisoft.
| | He said that Jack Schoof is "very" interested in re-opening
| | discussions with us regarding MS acquiring Artisoft. Jack
| | believes that Novell is gunning for us. While he understands
| | our decision to do Kate (Winbell) ourselves, he thinks we're
| | at risk on the low-end DOS peer networking market. He feels
| | that Artisoft can plug that hole instantly for us and can then
| | share with us their sound and windows knowledge.
| |
| | They (Jack and Adrian) would like us to take this as a real
| | "offer" to re-open discussions.
| |
|
|


######################################################## 7
>From russs Wed Jul 17 12:08:14 1991
To: bradsi
Subject; REl Novell/Digital research
Date Wed Jul 17 12:07:04 PDT 1991

What I meant by the VxD comment is the following. Today Artisoft has only a DOS real mode product. I am reading into the new merger proposition from Artisoft that the first thing Adrian talked about with Jack was "how do we get a better Windows product, ie. better integrated UI with Win, prot mode support" etc. And the ease answer of course is, have MS buy us. I think a big piece of this decision, if I was Adrian, would be the 32 bit prot mode work, which I know MS is doingm and I know it is market critical, and I know it will be hard to get it developed at Artisoft.

I am also reading into the Novell picture. Given that there plans to do a peer product are serious, they are most likely giong to follow the LANtastic model (the CNN article even said something like they were out to get Artisoft), ie. a DOS real mode peer product that runs under Windows with the UI being an app.

In both these cases we have 2 significant advantages 1) better memory solution since runs in Win prot mode 2) better integration with Win, since we can hack on file and prt mgr

So where we have an advantage is windows - we will have a significantly better windows product. Meanwhile, Novell, Artisoft and the rest will fight it out for the DOS real mode running under WIN market - good decision by us to stay out of that dogfight and focus on where we add value and gain competitive advantage - Windows.


PLAINTIFF'S EXHIBIT 986 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 986 Page 1 of 1]

####################################################### 346
>From philta(?) Mon Sep 30 09:03:27 1991
To: chuckst(?) mikedr scottq
Cc: bradsi davidcol(?) mackm tomle
Subject: Re: Bambi on DR-DOS 6.0
Date: Sat 28 Sep 91 08:46:59 PDT

The approach we will take is to detect dr 6 and refuse to load. The error message should be something like `Invalid device driver interface'.

mike, tom? make - do you have a reliable dr6 detection mechanism?

>From chuckst(?) Sun Sep 29 17:16:46 1991
To: mikedr philba scottq
Subject: Bambi on DR-DOS 6.0
Date: Sun Sep 29 17:16:39 1991

I tracked down a serious incompatibility with DR-DOS 6 -- They don't use the 'normal' device driver interface for >32M partitions. Instead of setting(?) the regular START SECTOR field to 0ffffh and then using a brand new 32-bit field the way MS-DOS has always done, they simply extended the start sector field by 16 bits.

This seems like a foolish oversight on their part and will likely result in extensive incompatibilities when they try to run with 3rd part device drivers.

I've patched a version of Bambi to work with DRD6 ad it seems to run Win 3.1 without difficulty. This same problem may have caused other problems with Win 3.1 and the swapfile under DRD6.

It is possible to make Bambi work, assuming we can come up with a reasonably safe method for detecting DRD6. The runtime hit would be minimal in time and space, although we wuld have a couple of instructions in the main code path for checking the 'special' DRD6 flag.

What do we think? Should we test further with the patched Bambi to see if there are any more incompatibilities????

####################################################### 347
>From bradc Mon Sep 30 09:08:07 1991
To: lorisi(?) sharonh
Cc: bradsi
Subject: steveb meeting
Date: Mon Sep 30 09:09:00 PDT 1991

i have a meeting with steveb tommorrow[sic] at 1pm to go over the dr dos stuff. i'd like brad to attend with me. if he can't see if we can rechedule for  another time tommorrow[sic] when he can.

[STAMPED] EXH 12  DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS 5054017 CONFIDENTIAL
[STAMPED] MS-PCA 1112238 CONFIDENTIAL
[STAMPED] DEFENDANT'S EXHIBIT 5557

PLAINTIFF'S EXHIBIT 4407 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4407 p. 1 of 3

Win 3.1 Beta Plan
Draft
17 December 1990

[STAMPED] 1103 DEPOSITION EXHIBIT
[STAMPED] X 548717 CONFIDENTIAL
[STAMPED] EXH 6 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS-PCA 1148389 CONFIDENTIAL

Plaintiff's Exhibit 4407 p. 2 of 3

Types of Beta Programs:

There are three concurrent Beta Programs planned for Windows 3.1.

  • Technical Beta
  • Pre-Release
  • Preview

Technical Beta is defined as those actively testing the Retail Windows product as part of a Beta Test. Pre-Release is to give customers, primary ISVs, early access to Retail, the SDK and the DDK as appropriate. The Preview program is for those who will not be actively doining testing, but who wish to preview the Retail product (ie. Corporate Accounts).

There are four other groups which will be receiving the Windows 3.1 software early. They are Windows Development Partners, ESP (Early ship program) OEMs (Original equipment manufacturers), general OEMs and some Press/Authors. There may be some general OEMs who are also Technical Beta testers, but it is important to note that they will be receiving additional software besides the retail product

Objectives:

The purpose of the Beta programs is to:

  • Get feedback from external sites on the stability of the product
  • Test the product on a variety of machine configurations provided by the beta test sites
  • Allow early accoss to the software and tools so ISVs gain a *head-start" on updating their applications
  • Provide early exposure for the product

Number of Sites:

There will be approximately 360 sites involved with Iha Technical Beta. This groop will consist of Corporate accounts. ISVs, and End users. The Pre-Release program will have have upwards ot 2500 ISV participants. Approximately 500 retail copies will be given out as part of the Preview program. There are approximately 12 Development partners, 15 ESP OEMs and 300 general OEMs who will be receiving Windows 3.1 software.

Software and Frequency of Shipments:

Program Software Frequency
 
Technical Beta Retail 3x (1/21,3/15,5/10)
Pre-Release Retail/SDK/DDK 3x (2/15,3/29,5/10,6/10)
Pre-Release (Strategic) Retail/SDK/DDK every 4 wks plus 2/15,3/29,5/10,6/10
Preview Retail 2x (1/21,3/15)
ESP OEMs OAK every 4 wks plus 2/15,3/29,5/10
Development Partners Retail/SDK/DDK every 2 wks plus 2/15,3/29,5/10
General OEMs OAK 3x (2/15,3/29,5/10)
Press/Authors/etc. Retail 1x (3/15)

Windows 3.1 Beta Plan

[STAMPED]X 548718 CONFIDENTIAL
[STAMPED]MS-PCA 1148390 CONFIDENTIAL

Plaintiff's Exhibit 4407 Page 3 of 3

Length of Beta:

The Technical Beta is scheduled to start on 1/21. The Pre-Release program oficially starts on 2/15. Primarily he difference between 1/21 abd 2/15 is to allow time to create the SDK, DDK and OAK kits and also for some additional testing time. The next major updata will be on 3/15. Some hardcopy docs (updated File Manger/Control panel and release notes) will be included in the 1/21 and 2/15 shipments with the possibility d of an updated version being sent with the 3/15 version. 5/i0 is theapproximate Retail ship date and 6/10 is the appoximate SDK ship date. See table below far dates and approximate numbers shipped.

Key ISVs ISVs Others Total Number
 
2/15 2/15 2/15 ~1500
3/29 3/29 3/29 ~2000
5/10 5/10 5/10 ~2500 (Retail)
6/10 6/10 ? ~3000
 
~200 500-2500 ~500 ~9300

Beta Coordinator;

The DOS/Windows Beta Coordinator will assist Program Management by providing information on how to start and run a beta program with specific duties in the following areas:
1) With initial input from Program manager, define fields and write database tables using Superbase as startup database.
2) Make recommendations on staffing levels of temporary staff including start date and duration of assignment.
3) Hire and manage temporary staff.
4) Disseminate flow of incoming information, to the admin and tech staff.
5) Coordinate mailings and other projects
6) Maintain inventory of Beta equipment, including printer to do labels (most of this equipment is currentiy in use by the DOS 5.0 beta program).
7) Special projects as requested and time permits.

Support:

Up to two people from PSS (DSBU) will be providing support for the Technical Beta testers. One person from PSS (SSBU) will be answering SDK-type questions from the ISVs (Ie. Pre-Release participants). These dedicated PSS members will be resident in bldg 3 during the Bets Programs. There, will be a dedicated FAX. and answering machine to report bugs. Use of Online 2 will be required by the Pre-Release members. Those customers who are previewing the retail product can Fax or phone in their bugs, but will not be given support unless they buy Online. A bug report sent to Online will not be counted as an SR.

Note: The use of CompuServe has been suggested. I am waiting to see how well CompuServe works for DOS 5 beiore deciding whether it will be used in the Windows 3.1 Beta programs.
Windows 3.1 Beta Plan

[STAMPED]X 548719 CONFIDENTIAL
[STAMPED]MS-PCA 1148391 CONFIDENTIAL



PLAINTIFF'S EXHIBIT 4546A Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4546A p. 1 of  3

Bruce Neiminen

From: bradc
To: galld; mmerker; rond
Cc: bradc; dosmktg
Subject: Key Competitive Info
Date: Thursday, May 09, 1991 11:47AM

Date: Thu May 09 11:42:49 PDT 1991

As an FYI/reminder, there are two key competitive Info docs that need to go to the field/onto SmartPages. Both of these docs are on \pyrexpublic. They are in dosmktgsales.inf.

Document #1 - OEM.DOC
------------
Explains how to sell the MS-DOS 5 Upgrade when you encounter an OEM also selling a DOS Upgrade. It focuses alot on the advantages of our product over the IBM product

Document #2 - Drdos.doc
--------------
Summarizes the key advantages of MS-DOS 5 and the MS-DOS 5 Upgrade versus Dr.DOS. We are doing further research on this and will have even more data soon.

Please work with richf to get these to the people out there that need them...

thanks for all your help

Bruce Neiminen

From: bradc
To: bradsl (bradsi?); steveb
Cc: bradc; joachimk; sergiop
Subject: RE: DRI
Date: Sunday, May 12, 1991 10:38AM

Date: Sat May 11 22:45:38 PDT 1991

consider it done.

>From steveb Tue May 14 19:02:58 1991
To: bradc bradsl (bradsi?)
Cc: joadhimk
Subject: DRI
Page 43

[STAMPED] X 567195 CONFIDENTIAL
[STAMPED] EXH 8 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER

Plaintiff's Exhibit 4546A Page 2 of  3


a customer can buy a network operating system from Novell without buying anything from Microsoft.
Ray Noorda, Novells chief executive, downplayed his company's rivalry with Microsoft, saying he was attracted to Digital Research because its version of DOS is superior to Microsoft's in some ways.
"Now we can add capabilities (to DOS) in advance of Microsoft doing it" Noorda said in a telephone conference.
The deal, which the companies expect will be completed by October, calls for Novell to issue 1.5 million in new shares in return for Digital Research's private stock. Digital Research said it had revenues of $40.9 million for its fiscal year ended last September.
-O- 5 18 pm edt 07-1 $-91 \023

Copyright (c) 1991 Dow Jones and Company, Inc.
Received by NewsEDGE/LAN: O7/16/S1|14:2O

Bruce Neimlnen

From: bradc
To: bradsi; davidcol; greglo; philba; richt; russs; tomle
Cc; bradc; rtcht; sergjop
Subject: RE: novell
Date: Tuesday, July 30,1991 1:41 PM

Date: Thu Jul 18 14:15:35 PDT 1991

One of my thoughts is that we have to thlink about how to short circuit Novell DOS before it gets off the ground. If we can put a daggar in Dr. DOS (or perhaps we should call it Novelal DOS) now then it will put them on the defensive and have customer worried

Now that we have the NSTL data back and some reasonable data on where Dr. DOS is problematic, I'd like to start a 'slow leak' program - every other week or month we try to get the word out on some major dr. dos compatibility problems. At the same time we should be prepared to invest in more third party testing to look for other holes - for example with netware and dr. dos. I'd have to work with PR to develop the specifics of the plan, but if we can get the world to understand that dr. dos has a kit of incompatibities with apps, windows and networks then it will put Novell on the defensive and make it hard for customers or oems (ibm???) to consider dr. dos seriously.

Page 57

[STAMPED]X 567209 CONFIDENTIAL

Plaintiff's Exhibit 4546A Page 3 of 3


> From davidcol Wed Jul 17 08:47:01 1991
To: bradc bradsl greglo phlba richt russs tomle
Subject: novell

Date: Wed Jul 17 06:46:14 1991

I think we shoud use Windows to get a Microsoft OS back onto the Netware clients which will bundle or require DR. DOS. We should alter our plans a bit and move all the DOS 6.0 improvements directly into Windows. When the user start Windows, they get the Microsoft OS (Including networking) and all the other cool features that go with that. When they quit, they get Netware and DR DOS and no Windows apps. The key is getting a piece of MS system software on that client so we can deliver our strategy and vision, We can leverage Windows and Windows apps to to this.

We should not consider things that stop Windows from working on Netware. (Netware here = netware + DR DOS.) if it was just DR DOS alone, then we should prevent Windows from working there. Netware has too much market share and too many customers are loyal to it for us to exclude windows from that market.

I think this dictates that we maintain good relationships with Novell so we can stay abreast wjth what they are doing at the detailed technical level. However, I do think that we should do our own winnet drivers and other Novell provider components. If we are going to take over the desktop when Windows starts, it MUST be all Microsoft written software since Novell won't help us do that.

Bruce Neimlnen

From: russs
To: bradc
Subject FW: more on DR DOS. Artisoft, and Novell from Jonathon Schmidt
Date: Thursday, August 08,1991 1:53PM


Date: Thu Jul 18 17:24:07 PDT1991

Here Is the mail I sent that BradSI and I werfe telling you about If you want to talk to Jonathon Schimdt his # at Performance Technology is XXX-XXX-XXXX. Tell him I sent you.

>From russs Thu Jul 18 1423:27 1991
To: steveb bradsl mikemur jilmall
Cc: davidl johnlu
Subject: more on DR DOS, Artisoft, and Noj/ell from Jonatbon Schmidt


Page 58

[STAMPED] X567210 CONFIDENTIAL



Exhibits Published in the Gordon v. Microsoft Trial on March 19, 2004
Transcription Complete. 29 Exhibits 127 pp.
Last revised 4/28/2004


PLAINTIFF'S EXHIBIT 373 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 373 Page 1 of 12]

[STAMPED] SIGNED ORIGINAL

203.25

AMENDMENT NUMBER 1

Dated September 1, 1990

TO THE LICENSE AGREEMENT

Dated September 1, 1990
Numbered G150-0271

for

Microsoft [reg] MS-DOS [reg], Version 4.01
Microsoft [reg] Works, Version 2.0

Between

MICROSOFT CORPORATION,
a Delaware Corporation

and

VOBIS DATA COMPUTER GMBH,
A German Corporation

[HANDWRITTEN] .G150-0271

14.2.91/rc

AMENDMENT

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222270 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT NO 608 FOR IDENTIFICATION DATE: 5-4-[19]98 RPTR [initials]
[STAMPED] MS98 0186787 CONFIDENTIAL

[Plaintiff's Exhibit 373 Page 2 of 12]


Effective September 1, 1990, the below signed parties agree that the indicated portions of the above referenced License Agreement (hereinafter the "Agreement") are hereby amended by this instrument (hereinafter the "Amendment"), as follows:

1. Exhibit B is hereby amended and as amended shall be read as set forth in the attached Exhibit B.

2. In Exhibit C3 the following paragraph is added in the section ROYALTY PAYMENTS AND REPORTING REQUIREMENTS:

As between the Products set forth in Exhibit C3 and C5, COMPANY agrees that for each applicable CUSTOMER SYSTEM shipped or placed in use by or for COMPANY during the term of this Agreement, COMPANY shall license, at most either the Exhibit C3 or the Exhibit C5 Product, but no combination thereof, i.e. no more than one such Product, for use on that CUSTOMER SYSTEM. COMPANY shall still make royalty reports on a Product by Product basis.

3. In Exhibit C3, the table in subsection (a) of the ROYALTY PAYMENTS AND REPORTING REQUIREMENTS section is hereby amended and as amended shall read as follows:

Customer System / Royalty Rate ($US)
Exhibit M2 / $13.05

4. New Exhibits C5, C6, C7, and C8 are hereby added and as added shall read as set forth in the attached Exhibits C5, C6, C7 and C8.

IN WITNESS WHEREOF, the parties have executed this Amendment as of September 1, 1990. All signed copies of this Amendment shall be deemed to be originals.

MICROSOFT CORPORATION
By [Signature of Michael R. Hallman]
Name(Print) Michael R. Hallman
Title President
Date 4/22/[19]91

VOBIS DATA COMPUTER GMBH
By [Signature of Theo Lieven]
Name(Print) Theo Lieven
Title General Manager
Date March 28, 1991

[Page Number] 2
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X222271 CONFIDENTIAL
[STAMPED] MS98 0186788 CONFIDENTIAL
[HANDWRITTEN] 608-2

[Plaintiff's Exhibit 373 Page 3 of 12]


EXHIBIT B
PAYMENT SCHEDULES

MINIMUM COMMITMENT

First Period of This Agreement

COMPANY agrees to pay a minimum of Three Million Dollars (US$ 3,000,000.00) for Product(s) licensed under this Agreement within the first period of this Agreement as described below. The Minimum Commitment Schedule listed below sets forth the minimum cumulative amounts of payments which COMPANY shall make to MS during the first period of this Agreement. To the extent that actual earned royalties exceed the cumulative minimum commitment payments, COMPANY shall pay MS for actual earned royalties. To the extent that cumulative minimum commitment payments exceed actual earned royalties, such excess shall be known as prepaid royalties and shall be recoupable against future earned royalties during the term of this Agreement. The minumum commitment amount payable upon signing of this Agreement as set forth below is recoupable and refundable puruant to Section 4(b) of this Agreement. All other minimum commitment payments are recoupable but not refundable.

Commencing with the first complete calendar quarter subsequent to COMPANY's shipment to a customer for revenue of Product MS-DOS(R) Version 5.0, German language version, in Exhibit C7, but no later than four (4) months after MS delivers this Product to COMPANY, minimum commitment amounts for royalty periods during the remaining term of this Agreement shall increase an additional Four Hundred Thousand Fifty Dollars ($450,000.00) over the amounts shown in this Exhibit B. COMPANY shall give MS written notice of the first date of shipment of this Product by COMPANY to a customer for revenue within ten (10) days of shipment.

MINIMUM COMMITMENT SCHEDULE (FIRST PERIOD)
Date
Payment Amount ($US)
Cumulative Amount of Payments for Period ($US)
Signing of this Agreement (payment due upon signing) $ 300,000.00 $300,000.00
End of the calendar quarter ("FIRST PAYMENT DATE") during which the first of the following occurs: the date of first COMPANY shipment of any Product to a customer for revenue, or six (6) months after the EFFECTIVE DATE of this Agreement
[HANDWRITTEN "525,000"]
$600,000.00
$900.000.00
3 months after the FIRST PAYMENT DATE
$700,000.00 $1,600,000.00
6 months after the FIRST PAYMENT DATE $700,000.00 $2,300,00.00
9 months after the FIRST PAYMENT DATE $700,000.00 $3,000,000.00
Total First Period Minimum Commitment
$3,000,000.00
$3,000,000.00
[Page Number] 3
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222272 CONFIDENTIAL
[STAMPED] MS98 0186789 CONFIDENTIAL
[HANDWRITTEN] 608-3

[Plaintiff's Exhibit 373 Page 4 of 12]

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222273 CONFIDENTIAL
[STAMPED] MS98 0186790 CONFIDENTIAL
[HANDWRITTEN] 608-4
Except for the amount due on signing, the date of payment for the above calendar quarter amounts shall be as provided in Section 3.

MINIMUM COMMITMENT

Second Period of This Agreement


COMPANY agrees to pay a minimum of Three Million Dollars (US$ 3,000,000.00) for Product(s) licensed under this Agreement within the second period of this Agreement as described below. The Minimum Commitment Schedule listed below sets forth the minimum cumulative amounts of payments which COMPANY shall make to MS during the second period of this Agreement. Payments made during the first period of this Agreement shall not be credited towards the minimum commitment requirement in the second period. To the extent that actual earned royalties exceed the cumulative minimum commitment payments, COMPANY shall pay MS for actual earned royalties. To the extent that cumulative minimum commitment payments exceed actual earned royalties, such excess shall be known as prepaid royalties and shall be recoupable against future earned royalties during the term of this Agreement. The minumum commitment amount payable upon signing of this Agreement as set forth below is recoupable and refundable puruant to Section 4(b) of this Agreement. All other minimum commitment payments are recoupable but not refundable.

Commencing with the first complete calendar quarter subsequent to COMPANY's shipment to a customer for revenue of Product MS-DOS(R) Version 5.0, German language version, in Exhibit C7, but no later than four (4) months after MS delivers this Product to COMPANY, minimum commitment amounts for royalty periods during the remaining term of this Agreement shall increase an additional Four Hundred Thousand Fifty Dollars ($450,000.00) over the amounts shown in this Exhibit B. COMPANY shall give MS written notice of the first date of shipment of this Product by COMPANY to a customer for revenue within ten (10) days of shipment.

MINIMUM COMMITMENT SCHEDULE (SECOND PERIOD)
Date

End of the:
Payment Amount ($US)
Cumulative Amount of Payments for Period ($US)
12 months after the FIRST PAYMENT DATE $ 750,000.00 $ 750,000.00
15 months after the FIRST PAYMENT DATE $ 750,000.00 $ 1,500,000.00
18 months after the FIRST PAYMENT DATE
$ 750,000.00 $2,250,000.00
21 months after the FIRST PAYMENT DATE $ 750,000.00 $3,000,000.00
Total Second Period Minimum Commitment
$3,000,000.00 $3,000,000.00
The date of payment for the above calendar quarter amounts shall be as provided in Section 3.

Exhibit to the License Agreement dated September 1, 1990, between MICROSOFT CORPORATION and VOBIS DATA COMPUTER GMBH.

[Plaintiff's Exhibit 373 Page 5 of 12]

EXHIBIT C5 (SYSTEM COMMITMENT)
PRODUCT: Microsoft [REG] Windows [TM]

VERSION NO: 3.0
LANGUAGE VERSION: (Do not fill in if Domestic USA Version)

PRODUCT DELIVERABLES:
(a) Product in Object Code form.
(b) Standard Documentation in Series Number W770-3Z that MS delivers with the Product.

PRODUCT SPECIFICATIONS:
The Product will have features as specified in the above-referenced Product documentation.

ROYALTY PAYMENTS AND REPORTING REQUIREMENTS:
COMPANY agrees to pay MS, according to, and comply with Exhibit C3.

ROYALTIES FOR NEW VERSION RELEASES:
MS may elect to increase the per system royalty applicable to new Version Releases, which royalty shall apply to succeeding releases until changed hereunder or by agreement of the parties. The maximum per system royalty for a new Version Release shall be determined as follows:

Maximum royalty = R + (R*N*1.5%)

where R is the initial per system royalty described above and N is the number of months (rounded to the nearest whole number) that have elapsed from the Effective Date until MS delivers the new Version Release.

ROYALTIES FOR UPGRADES:

COMPANY may elect to pay MS a royalty of fifty percent (50%) of the highest applicable royalty stated in this Exhibit C (i.e., the initial royalty stated above plus increases, if any, for new Version Releases) for each full or partial copy of the Product, including Update Releases and Version Releases, licensed or otherwise disposed of by COMPANY as an Upgrade during the term of this Agreement. Upgrade Product shall conform to the definition contained in Section 1 of the Agreement.

ADDITIONAL PROVISIONS:

The Product Microsoft Windows described in this Exhibit contains Version 7.0 of the Microsoft mouse driver ("MS Driver
") (files designated as MOUSE.COM and MOUSE.SYS), which is MS proprietary software designed for use with the Microsoft Mouse. In the event COMPANY ships a non-MS mouse or pointing device with the MS Driver to end user purchasers of COMPANY's Computer Systems, COMPANY agrees to pay MS an additional royalty of Five Dollars (US $5.00) for each copy of the Driver so shipped by COMPANY. Reporting and payment of such Driver royalties shall be in accordance with this Exhibit and Section 3 ("Price and Payment") of this Agreement.

[PAGE NUMBER] 5
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X222274 CONFIDENTIAL
[STAMPED] MS98 0186791 CONFIDENTIAL
[HANDWRITTEN] 608-5

[Plaintiff's Exhibit 373 Page 6 of 12]


COPYIGHT NOTICE:
COPYRIGHT (C) MICROSOFT CORPORATION, 1985-1989. All rights reserved.

PRODUCT NAME AND ASSOCIATED TRADEMARK: Microsoft [REG] Windows [TM]

OTHER ASSOCIATED TRADEMARKS AND DESCRIPTOR:

Exhibit to the License Agreement dated September 1, 1990, between MICROSOFT CORPORATION and VOBIS DATA COMPUTER GMBH
02/07/[19]90 0876L

[PAGE NUMBER] 6
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222275 CONFIDENTIAL
[STAMPED] MS98 0186792 CONFIDENTIAL
[HANDWRITTEN] 608-6

[Plaintiff's Exhibit 373 Page 7 of 12]


EXHIBIT CE

ADDITIONAL LANGUAGE VERSIONS

The Product described in Exhibit C5 ("Reference Exhibit") shall include the additional language versions identified below. COMPANY shall license at most one language version of the Product for use on each applicable CUSTOMER SYSTEM. COMPANY'S royalty obligations shall be as set forth in the Reference Exhibit regardless of the language version licensed with each applicable CUSTOMER SYSTEM. Nothwithstanding the preceding sentence, COMPANY shall pay MS an additional language version royalty equal to fifteen percent (15%) of the royalty otherwise payable to MS for each full or partial copy of each language version of the Product identified below which is licensed or otherwise disposed of by COMPANY during the term of this Agreement. COMPANY shall report, on a language version by language version basis, and pay such language version royalties pursuant to and as part of its royalty payment and reporting obligations under the Royalty Payments and Reporting Requirements section of the Reference Exhibit. The Product Name and Associated Trademark are, in all Language Versions listed below, the same as in the Reference Exhibit.

1. Language Version: French
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number W770-3F that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1985 - 1990 . All rights reserved.


2. Language Version: German
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number W770-3D that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1985 - 1990 . All rights reserved.

3. Language Version: Portuguese
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number W770-3P that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1985 - 1990 . All rights reserved.

[PAGE NUMBER] 7
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222276 CONFIDENTIAL
[STAMPED] MS98 0186793 CONFIDENTIAL
[HANDWRITTEN] 608-7


[Plaintiff's Exhibit 373 Page 8 of 12]


4. Language Version: Spanish
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number W770-3E that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1985 - 1990 . All rights reserved.

5. Language Version: Swedish
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number W770-3S that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1985 - 1990 . All rights reserved.

Exhibit to the License Agreement dated September 1, 1990, between MICROSOFT CORPORATION and VOBIS DATA COMPUTER GMBH.
03/14/[19]90 0975L

[PAGE NUMBER] 8
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222277 CONFIDENTIAL
[STAMPED] MS98 0186794 CONFIDENTIAL
[HANDWRITTEN] 608-8

[Plaintiff's Exhibit 373 Page 9 of 12]

EXHIBIT C7 (SYSTEM COMMITMENT)

PRODUCT: Microsoft [REG] MS-DOS [REG]
VERSION NO: 5.0 (Includes QBasic Interpreter)
VERSION NO: 5.0
LANGUAGE VERSION: ____ (Do not fill in if Domestic USA Version)
PRODUCT DELIVERABLES:
(a) Product in Object Code form.
(b) Standard Documentation in Series Number D781-5Z and D707-5Z that MS delivers with the Product.

PRODUCT SPECIFICATIONS:
The Product will have features as specified in the above-referenced Product documentation.

ROYALTY PAYMENTS AND REPORTING REQUIREMENTS:

(a) For purposes of royalty payments, this Exhibit shall become effective as follows: If COMPANY prompty provides written notice to MS of the date on which COMPANY first ships this Product to any customer, then this Exhibit shall become effective on the shipment date specified in the notice; otherwise, this Exhibit shall become effective on the first day of the calendar quarter for which COMPANY files a royalty report indicating shipment of this Product.

(b) For each Customer System identified below, COMPANY agrees to pay MS a royalty, at the applicable rate set forth below, multiplied by the greater of (i) the number of full or partial Customer System(s) shipped or placed in use by or for COMPANY during the term of this Agreement, or (ii) the number of full or parital copies of Product, including Update Releases and Version Releases, licensed or otherwise disposed of by COMPANY during the term of this Agreement for use with such Customer System, but excluding copies of the Product for which COMPANY pays royalties at the Upgrade Royalty Rate (see below).

Customer System  Royalty Rate ($US)
Exhibit M1 $7.82

(c) COMPANY's report shall specify the number of Customer System(s) shipped or placed in use by or for COMPANY during that calendar quarter and the number of copies of Product, including Update Releases, Version Releases, and Upgrades licensed or otherwise disposed of by COMPANY during that calendar quarter. COMPANY shall furnish this statement for each Customer System identified in the Exhibit M9s) and shall report for each Customer System separately by processor. In the event that no Customer System(s) are shipped or placed in use by or for COMPANY during a calendar quarter and no copies of Product are licensed or otherwise disposed of by COMPANY during such calendar quarter, COMPANY shall indicate this on the royalty report.

[PAGE NUMBER] 9
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222278 CONFIDENTIAL
[STAMPED] MS98 0186795 CONFIDENTIAL
[HANDWRITTEN] 608-9

[Plaintiff's Exhibit 373 Page 10 of 12]


ROYALTIES FOR NEW VERSION RELEASES:

MS may elect to increase the per system royalty applicable to new Version Releases, which royalty shall apply to succeeding releases until changed hereunder or by agreement of the parties. The maximum per system royalty for a new Version Release shall be determined as follows:

Maximum royalty = R + (R*N*1.5%).

where R is the initial per system royalty described above and N is the number of months (rounded to the nearest whole number) that have elapsed from the Effective Date until MS delivers the new Version Release.

ROYALTIES FOR UPGRADES:

COMPANY may elect to pay MS a royalty of fifty percent (50%) of the highest applicable royalty stated in this Exhibit C (i.e., the initial royalty stated above plus increase, if any, for new Version Releases) for each full or partial copy of the Product, including Update Releases and Version Releases, licensed or otherwise disposed of by COMPANY as an Upgrade during the term of this Agreement. Upgrade Product shall conform to the definitiion contained in Section 1 of the Agreement.

ADDITIONAL PROVISION:

Prior to distribution of product by COMPANY, COMPANY shall implement a tracking procedure (e.g., bar coding, serialization) that has been approved in writing by MS. COMPANY's tracking system shall enable COMPANY to identify its customer (i.e. distributor, dealer, end user) for ecah unit of the Product distributed. COMPANY shall not license or otherwise dispose of the Product unless and until such tracking system has been approved by MS and implemented by COMPANY.

COPYRIGHT NOTICE:

COPYRIGHT (C) MICROSOFT CORPORATIO, 1981-1991. All rights reserved.
COPYRIGHT (C) MICROSOFT CORPORATIO, 1982-1991. All rights reserved.

PRODUCT NAME AND ASSOCIATED TRADEMARK:
Microsoft [REG] MS-DOS [REG]
QBasic [TM] Interpreter

Exhibit to the License Agreement dated September 1, 1990, between MICROSOFT CORPORATION and VOBIS DATA COMPUTER GMBH.
11/29/[19]90 dos5djj

[PAGE NUMBER] 10
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222279 CONFIDENTIAL
[STAMPED] MS98 0186796 CONFIDENTIAL
[HANDWRITTEN] 608-10

[Plaintiff's Exhibit 373 Page 11 of 12]


EXHIBIT C8

ADDITIONAL LANGUAGE VERSIONS

The Product described in Exhibit C7 ("Reference Exhibit") shall include the additional language versions identified below. COMPANY shall license at most one language version of the Product for use on each applicable CUSTOMER SYSTEM. COMPANY's royalty obligations shall be as set forth in the Reference Exhibit regardless of the language version licensed with each applicable CUSTOMER SYSTEM. Notwithstanding the preceding sentence, COMPANY shall pay MS an additional language version royalty equal to fifteen percent (15%) of the royalty otherwise payable to MS for each full or partial copy of each language version of the Product identified below which is licensed or otherwise disposed of by COMPANY during the term of this Agreement. COMPANY shall report, on a a language version by language version basis, and pay such language version royalties pursuant to and as part of its royalty payment and reporting obligations under the Royalty Payments and Reporting Requirements section of the Reference Exhibit. The Product Name and Associated Trademark are, in all Language Versions listed below, the same as in the Reference Exhibit.

1. Language Version: French
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5F and D707-5F that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1981 - 1991 . All rights reserved.
COPYRIGHT (C) MICROSOFT CORPORATION, 1982 - 1991 . All rights reserved.

2. Language Version: German
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5D and D707-5D that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1981 - 1991 . All rights reserved.
COPYRIGHT (C) MICROSOFT CORPORATION, 1982 - 1991 . All rights reserved.

3. Language Version: Portuguese
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5P and D707-5P that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1981 - 1991 . All rights reserved.
COPYRIGHT (C) MICROSOFT CORPORATION, 1982 - 1991 . All rights reserved.

[PAGE NUMBER] 11
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222280 CONFIDENTIAL
[STAMPED] MS98 0186797 CONFIDENTIAL
[HANDWRITTEN] 608-11

[Plaintiff's Exhibit 373 Page 12 of 12]



4. Language Version: Spanish
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5E and D707-5E that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1981 - 1991 . All rights reserved.
COPYRIGHT (C) MICROSOFT CORPORATION, 1982 - 1991 . All rights reserved.

5. Language Version: Swedish
Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5S and D707-5S that MS delivers with the Product.
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT (C) MICROSOFT CORPORATION, 1981 - 1991 . All rights reserved.
COPYRIGHT (C) MICROSOFT CORPORATION, 1982 - 1991 . All rights reserved.

Exhibit to the License Agreement dated September 1, 1990, between MICROSOFT CORPORATION and VOBIS DATA COMPUTER GMBH.
03/14/[19]90 0975L

[PAGE NUMBER] 12
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] X 222281 CONFIDENTIAL
[STAMPED] MS98 0186798 CONFIDENTIAL
[HANDWRITTEN] 608-12


PLAINTIFF'S EXHIBIT 461 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 461 p. 1 of 3]


I doubt that's the problem. When I dial MCI on the same machine, I never have a problem. Only dialing into 867-1681. I get a connect message, but then after hitting enter a number of times, nothing happens. So 1 hang up, redial, and repeat 5-10 times until I get the login message.

No problems at all calling other services.

| >From corpop Mon Nov 26 08:18:44 1990
| To: bradsa
| Subject linelock
| Cc: tomh
| Date: Mon Nov 26 08:05:33 1990
|
|
| It's possible you are running into a hung modem. Please call us
| at 936-3367 when it happens so we can have a closer look.
|
| Thanks, Corpop.
|
|--------------------------------------------
|
| >From joero Sun Nov 25 09:33:02 1990
| To: AllenY GregP corpop
| Subject: linelock
| Date: Sun Nov 25 09:27:59 1990
|
| From bradsi Sat Nov 24 18:50:33 1990
| To: joern
| Subject: linelock
| Date: Sat Nov 24 18:48:07 1990
| Why do I have so much trouble getting through linelock
| to get a login message? I typically have to dial 867-1681
| 5-10 times before I get login.
|
|


################################################################### 101
>From bens Mon Nov 26 12:21:28 1990
To: bradsi
Cc: camcromn davidw pauima philba richab
Subject to share or not to share, that is the question
Date: Mon Nov 26 12:1934 1990

Issue:

- Should MS Apps (and other ISVs) have access to Windows source code?

Complications:

1) If MS Apps have access to the Windows sources, then all ISVs should have access, else we are subject to restraint-of-tradc complaints (to say nothing of the morality of the situation).

[Some exhibit labelling is missing from this transcript page.]

[Plaintiff's Exhibit 461 p. 2 of 3]

2) If outside ISVs have access to our sources, then we make it much easier for another company to come along and clone Windows.

I talked to some apps guys on a recent recruiting trip, and they made the following very good point;

Windows is poorly documented

The state machine that is USER.EXE is barely documented in the SDK. Any ISV that wants to write a great Windows app ends up looking at the source code (like our Apps group), unassembling the DLLs, or writing experiment code to divine the actual behavior of the system.

Arguments against giving ISVs access to Windows source code usually boil down to:

An ISV will use some undocumented feature of Windows, or directly access internal data, in such a way that future versions of Windows will be forced to support this bad behavior, restricting MS ability to innovate in Windows.

I claim that letting an ISV looking at the source code is the *best* way to avoid this problem:

1) An ISV that unassembles Windows to figure out its behavior is effectively looking at source code, but without the benefits of source comments. This approach is more work for the ISV, and gives MS no opportunity to guide the ISV. With source code, there are generally comments discussing rationale for the behavior of the system.

2) An ISV that writes test apps to divine Windows behavior is really on thin ice. Either the ISV spends a great deal of effort writing test code to be certain Windows is fully understood, or the ISV may end up making assumptions which are not correct Since tbe former approach is a great deal of work, and the ISV is never sure when to stop ("do I really understand how this works now?"), most ISVs will end up in the latter situation. These are the most dangerous apps, since they are most dependant upon the exact behavior of a specific release of Windows.

The key problem is that our documentation docs not provide sufficient depth of coverage. The key question is:

How much would it cost to provide sufficient documentation, and is that any different, really, from providing source code?

The key difficulty in writing really great documentation is anticipating all the questions an ISV might have about the behavior of the system. Presumably, there is a level of documentation which is great enough that an ISV would have to perform only a small amount of experimentation.

Solutions:

[Plaintiff's Exhibit 461 p. 3 of 3]


A. Status Quo
+ Simple
- MS Apps have (unfair) advantage over other ISVs.

B. Make Windows sources available for a fee, with a restrictive licensing agreement (only available to N trusted employees, must be kept on a secure server, no derivative of these sources may be shipped, etc.)
+ All ISVs are equal
- Cloning risk

C. Disallow all ISVs, including MS Apps, from looking at Windows sources (in practice, this would be a very hard thing to do, and certainly runs counter to the spirit of MS).
+ All ISVs are equal
- Difficult to make happen at MS
- Reduces information flow on Windows to ISVs

D. Write great documentation.
+ All ISVs are equal
+ Avoids cloning risk

Conclusion

Given that we want to make Windows programming as attractive as possible, I vote for (B) making the sources available. This gives us a little extra incentive to keep enhancing Windows, so that it does not become a stationary target for doners, but otherwise benefits the Windows ISV community.

- bens

################################################################### 102
>From ericbo Mon Nov 26 12:25:51 1990
To: jamesm
Cc: dos5beta
Subject: Re: Where is DOS 5.0???
Date: Mon Nov 26 12.24:51 1990

We do not have a current release that has been tested well enough to be used in an internal beta. We are in the process of testing a new version and it should be ready later this week. Mail will be going out when it is ready.

If you wish to be a "guinea pig" and install the build we are currently testing for an internal release, email dos5beta and we will add you to our list If you need disk images though, you will probably have to wait until the internal beta version is ready.

-Eric Hough
Dos 5 Beta Support


>From jamesm Mon Nov 26 12:11:32 1990
To: dos5beta
Subject Where is DOS 5.0???


PLAINTIFF'S EXHIBIT 462 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 462 p. 1 of 3]


Cc: bradsi gcnfac janfac
Date: Mon Nov 26 14.4X.39 1990

The Facilities Department can be contacted via e-mail on the "genfac" aliasis. The "facman" alais is no longer in use.

In regards to your request concerning the Bldg 3 2nd floor mens room, I am having a plumber out to correct a clog in the urinal. It looks like it will need to be pulled off the wall to repair correctly. For the urinal to be clogged this bad, something has physically been put down the urinal path. Please be patient for one more day while repairs are effected.

"janfac" will ensure that the area is cleaned up in a sanitary manner.

Thank you for your patience!


################################################################### 127
>From philba Mon Nov 26 14:53:50 1990
To: bens bradsi
Cc. camcronm davidw paulma richab
Subject Re: to share or not to share, that is the question
Date: Mon Nov 26 15:50:19 1990

You bring up some valid points however there is one other point that needs to be understood: ISVs that have access to our code often take advantage of internal data-structures, bugs or other quirks of the implementation. This prevents us from changing the code and adds a serious burden of compatibility beyond the API.

I favor simply doing a better job of documenting.

| From bens Mon Nov 26 12:21:28 1990
| To: bradsi
| Cc: camcromn davidw pauima philba richab
| Subject to share or not to share, that is the question
| Date: Mon Nov 26 12:1934 1990
|
| Issue:
| - Should MS Apps (and other ISVs) have access to Windows source code?
|
| Complications:
|   1) If MS Apps have access to the Windows sources, then all ISVs should
|       have access, else we are subject to restraint-of-tradc complaints
|       (to say nothing of the morality of the situation).
|
|   2) If outside ISVs have access to our sources, then we make it much easier
|       for another company to come along and clone Windows.
|
| I talked to some apps guys on a recent recruiting trip, and they made
| the following very good point;

[Plaintiff's Exhibit 462 p. 2 of 3]

|
|   Windows is poorly documented
|
| The state machine that is USER.EXE is barely documented in the
| SDK. Any ISV that wants to write a great Windows app ends up
| looking at the source code (like our Apps group), unassembling
| the DLLs, or writing experiment code to divine the actual behavior
| of the system.
|
| Arguments against giving ISVs access to Windows source code usually boil
| down to:
|
|   An ISV will use some undocumented feature of Windows, or directly
|   access internal data, in such a way that future versions of Windows
|   will be forced to support this bad behavior, restricting MS ability
|   to innovate in Windows.
|
| I claim that letting an ISV looking at the source code is the *best* way
| to avoid this problem:
|
|    1) An ISV that unassembles Windows to figure out its behavior is
|        effectively looking at source code, but without the benefits of
|        source comments. This approach is more work for the ISV, and
|        gives MS no opportunity to guide the ISV. With source code,
|        there are generally comments discussing rationale for the behavior
|        of the system.
|
|    2) An ISV that writes test apps to divine Windows behavior is really
|        on thin ice. Either the ISV spends a great deal of effort writing
|        test code to be certain Windows is fully understood, or the ISV
|        may end up making assumptions which are not correct Since tbe
|        former approach is a great deal of work, and the ISV is never sure
|        when to stop ("do I really understand how this works now?"), most
|        ISVs will end up in the latter situation. These are the most
|        dangerous apps, since they are most dependant upon the exact behavior
|        of a specific release of Windows.
|
| The key problem is that our documentation docs not provide sufficient
| depth of coverage. The key question is:
|
|    How much would it cost to provide sufficient documentation,
|    and is that any different, really, from providing source code?
|
| The key difficulty in writing really great documentation is anticipating
| all the questions an ISV might have about the behavior of the system. Presumably,
| there is a level of documentation which is great enough that an ISV would have
| to perform only a small amount of experimentation.
|
| Solutions:
|
|    A. Status Quo
|       + Simple
|       - MS Apps have (unfair) advantage over other ISVs.
|
|    B. Make Windows sources available for a fee, with a restrictive licensing

[Plaintiff's Exhibit 462 p. 3 of 3]

|        agreement (only available to N trusted employees, must be kept on
|        a secure server, no derivative of these sources may be shipped, etc.)
|       + All ISVs are equal
|       - Cloning risk
|
|    C. Disallow all ISVs, including MS Apps, from looking at Windows sources
|        (in practice, this would be a very hard thing to do, and certainly
|        runs counter to the spirit of MS).
|       + All ISVs are equal
|       - Difficult to make happen at MS
|       - Reduces information flow on Windows to ISVs
|
|    D. Write great documentation.
|       + All ISVs are equal
|       + Avoids cloning risk
|
|
| Conclusion
| Given that we want to make Windows programming as attractive as possible,
| I vote for (B) making the sources available. This gives us a little
| extra incentive to keep enhancing Windows, so that it does not become
| a stationary target for doners, but otherwise benefits the Windows
| ISV community.
|
| - bens
|


################################################################### 128
>From cliffw Mon Nov 26 14:57:51 1990
To: bradsi
Subject: XL Bugs
Date: Mon Nov 26 14:55:23 1990

The PREVIEW bugs have been forwarded to the appropriate people.

The Style Combo box display problem is probably a WIN 3 installation bug. I think this is the problem because I just checked a the beta version on a 386/25 in the configuration lab, and there were mo problems.

If you re-install Win 3 (or I will do it if you wish) the problem should go away. If it doesnt go away, please contact me.


################################################################### 129
>From korys Mon Nov 26 15:03:58 1990
To: dwgroup
Subject: Printers in 3/2165
Date: Mon Nov 26 15:01:16 1990

\\prntsrv\aplw - Apple LaserWriter plus
\\prntsrv\ljiii - HP Usenet III
[PE462 END]



PLAINTIFF'S EXHIBIT 638 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 638 p. 1 of 6]

[STAMPED] Depo. Ex. 33
[STAMPED] Depo. Ex. 607
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] CONFIDENTIAL
[STAMPED] X0599970
[STAMPED] PLAINTIFF'S EXHIBIT 291 C.A. No. 2:96CV645B
[STAMPED] MS-PCA 1177233 CONFIDENTIAL

Microsoft Memo

To: Jeremy Butler
From: Joachim Kempin
Date: March 26, 1991
Subject: Europe Trip Report
cc: Bill Gates, Mice Hallman. Bernard Vergnes. Chris Smith, Richard Fade, Ron Hosog, Jeff Lum. Tim Beard

ADI Software (Windows ISV) - Karlsruhe. Germany

Received a very positive update on the German Win and Win apps market ADI is the exclusive distributor for the Asymmetrix Toolbox and is localizing it. In addition, they used it to write multimedia apps and combined it with their own database technology. It's amazing how this improved speed. Paul Allen should look at this. On a side note - my old friend Mark Ursino, now Asymmetrix Marketing Manager, seems to be as unpopular as ever according to ADI and Internal Asymmetrix feedback.

CeBIT- Hanover. Germany

The show now consists of 20 halls and next year will be expanded to 24 halls. They had a new record this year with 570,00 visitors in attendance. MS had four well-organized booths. Most impressive was the ISV booth. The main booth, MAC and LAN Man booths were well attended. Business climate was upbeat: 12-16% unit growth to be expected, but a lot of disappointment in Eastern European business opportunities. It will take 5-6 years to see the fruits of any investments. The USSR might not be in for good business for another 10 years. East Germany will be healthy in 2-3 years. (It takes one day to make a scheduled phone call into the Eastern part of Germany, meaning no communication infrastructure). At the same time, skilled and motivated labor does not exist. Fifty years of communist government took the desire to perform out of the country. Do we need to revisit our localization strategy?

Most impressive: TFT color laptops from Sharp, Toshiba. Amstrad (7.5K$), and Aquarius.

Ultra-thin Notebook as well as Amstrad's "Mac-killer"

KEY CUSTOMER VISITS

I missed Schneider and Compass due to a delay in Frankfurt. Manfred needs to follow up with Koester (ex Schneider now part of Compass), a dealer chain association which is planning to build their own PCs.

Tandon

Tandon is still growing 20% per year, very professional and healthy. Doing 30% of all Tandon sales. The bad news: they signed up to distribute Solborn SPARC stations in Germany and if successful, they will do so in other European countries. We need to change their thinking after the MIPS announcement. They would switch if told by their US headquarters.

[Plaintiff's Exhibit 638 p. 2 of 6]

Zenith

Zenith has completely reorganized and changed all sales & marketing management in Germany. They are fighting to come back, and still being profitable, the local VP was able to survive. More important was the message that they are happy with the management changes at HQ and that they now have more input on product directions. 70% of Zenith's total business is now done in Europe.

DRI

Germany Is their best county worldwide. Lots of Taiwanese clone manufacturers and Vobis are present with DRI bundles. We will have to target them one by one. Interviewed a DRI person who wants to switch sides, will follow up with JoachenH.

Interview with Rainerk

CWedell asked for my opinion of whether he could do the German PSS manager job. I supported his recommendation after talking to him. From the interview and some discussion with the OEM managers. It was obvious that we need a unified European PSS concept. This is particularly true for large account customers when supporting mission-critical apps. These guys shop for the best prices and have started laughing at our European inconsistencies.

WORKS

Still complaints from Juttad that C8M and Schneider are violating distribution agreements. I confronted the CBM GM and he admitted that this had happened but had been corrected In Schneider's case. I asked GmbH to buy packages without hardware to give us some proof. We should terminate their WORKS license if true - being the second offense. Jochenh told me that al these OEM bundles in Germany are generally helping his WORKS sales.

Aquarius

Aquarius sold 40K units in their first year of existence. Confirmed that 386SX is the best selling PC type in Germany. They upped their commit to 60K and shook hands over a new 2.5M/per year deal giving us $40+ per system. I committed to pay for 20K of advertisements in local catalogs for making DOS 5.0 popular and fighting DRI. Because they had bought FG DOS and WIN before they started shipping royalty based products we forgave them 150K of min commits. Their ultra-thin notebooks is interesting. It contains DRI ROM-DOS from Taiwan, but they will force their vendor to change to MS-DOS.

Ronh please make sure that Taiwan OEM follows up.

Pretty bad if true:
They are working closely with Suja in the USSR and claim they have sold them 65K PCs in components to be assembled locally. The Russian company Informed them that they had signed a deal for $2 for DOS equivalent in Rubles. We denied that such a contract existed asked for a copy. I will follow up with Dalech.

Peacock

The usual "you do not love us" complaint. I promised to visit them in Paderborn in May. They are interested in partnering with us on some integration projects. Key point: IBM is hurling us. Novell is being positioned by the IBM sales force as being "it". (This was confirmed by 3-4 other OEMs). OS/2 is an IBM proprietary system etc., etc. They and others asked us to do a better PR job in the US and locally to make our story more believable. They told us we as a software company are missing the boat by not having a NOTES type product today and a low-end LAN Man solution.

[Plaintiff's Exhibit 638 p. 3 of 6]

Vobis

This is a company owned 20% by Kauhof which in turn is owned by METRO. There was talk in Hanover that an additional 20% might be sold to Kaunof by Theo Liefen, their founder. We met him twice to discuss the new contract. It contains a WIN and WORKS license for all their systems to be expected to exceed 200K units this CY making them number 2 after IBM. And it includes a DOS 5.0 "trigger" license which makes them pay for MS DOS three months after they receive the OAK. Liefen denied that he ever agreed to this. (From my DRI contact, I knew that he had extended his current deal by 50K units- thus putting him under pressure). With Manfred not being present because of his leg surgery, we agreed to meet again the next day to resolve the issue.

Interesting enough, Amstrad and other German companies have been noticing Vobis' success and its' DRI bundling. Liefen himself mentioned to us that he could influence DRI in their product development, etc. After talking to Manfred, it was obvious that Liefen was reneging on the deal. Round two: I took the opportunity to negotiate in German, sign our offer as is - this is an agreed upon package deal or if you change any component, we will too. Second option: scratch the DOS clause, pay $35 for Windows instead of $15. You have untill 04/01/91 to consider. If no contract is signed, seize shipping Windows by 05/01/91. They are shipping without a license (!) submit royalty reports for WIN and WORKS and pay up. (They have a valid WORKS license at $25). The proposal showed impact. They threatened with GeoWorks. Lotus, etc We asked if the DOS price was an issue - he answered no. We parted as friends - believe it or not, Manfred will follow up with a letter and a phone call and I have a bet with Jeff that they win sign as is. In my judgement they will hurt if they do not ship WIN and paying $35 for it is out of the question.

IPC

Very amiable meeting regarding marketing plans. They are departing from ESSEX their parent company. Told us that IPC Spain is in financial trouble, I informed Ignacio.

Compaq, Gmbh

Got to know two new GMs and confirmed the European MIPS announcement. Great response so far to their Ballpoint promo. Healthy business outlook based on top management meetings with large accounts at Hanover.

Schneider and Koch

A true disaster meeting. They are one of the leading network software sellers in Germany with a 30% market share. They still are without a 2.0 German LAN Man OAK. One reason was the delaying of the 2.0 amendment by them and second the amendment being held up in legal because of sloppy preparation by Gmbh. They did praise K.H. Breitenbach and his support, but GmbH management - having closed an exclusive distribution agreement for LAN Man (would not have happened at my time) - refused to authorize their dealers for 6 months. Meanwhile, they are selling Novell - now stupid and arrogant can we get? I promised them help by sending an OAK asap and told the German guys to allow them to ship product from their warehouse included in their box in the interim. I hope they can work this out They asked for a distribution agreement and I told them to work with Gmbh. This might not be possible locally until the exclusive deals runs out by year end.

Amstrad

We took them out for a nice lunch in a truly nice place. After the usual "lower your prices" and how well they did their WIN shell without paying 1M$ to MS development, we were able to break the ice. Again, not enough resources to serve and nurture them, let them believe we do not care. They are seriously thinking of shipping DR DOS 5.0 for their German 8086 PCs starting in October - because Vobis does it. They are designing a new low-end and high-end product line and are asking for our help. This time they will get it. We exchanged some market info and they were impressed by how well informed we were. We succeeded in interesting them in Pen-WIN

[Plaintiff's Exhibit 638 p. 4 of 6]

based PC design and recommended doing this before they start a MM PC. Tney agreed - Malcolm was actually really excited about it

I believe the DRI Issue can be put to rest if they see a true effort Malcolm and myself established a good rapport. I am going to make them my European "pet" account.

CBM

Met with their new GM. This was the third time he has worked tor CBM. Lots of energy and drive. He is turning things around 35% of total CBM sales is in Germany. They are still selling 300K 64's and more than 200K Amigas. Their CD-ROM video is catching on. It is based on an Amiga 500 chip residing in a remote control - a great player as most customers believe and very easy to operate. Tandy, you might get killed! GmbH will follow up with some educational sales opponunities.

Siemens

Feldman and Dewltd asked a lot of questions regarding OS/2, WIN positioning, why MIPS, and how real is LAN Man? I confirmed our MIPS plans and answered their questions. Again they raised the issue of a completely confused market and IBM going around trashing us. We invited them to the planned May briefing. Their first R3000 based MIPS machine will be a WS followed by a MP machine - incompatible for us. They will do a R4000 machine asap. Interesting enough they feel pretty sure that MIPS win ship "error free" CPUs in Q1 CY92 at the earliest. One year too late - by then SPARC 3, a better PA chip, and the next RS 6000 will be out. This means no free lunch for the MIPS consortium. We might have to fight harder than expected. They are developing banking apps by using the PM interface. I confirmed that OS/2 3.0 will conserve the PM interface but I think they understood clearly to move to WIN 32-bit asap to have their investment secured in the future.

They were disappointed about our dying mailserver cooperation but want to do development project with us. I said pick a small one and we can start, build trust and get it done. I will make this somebodies project. All in all a very amiable and constructive meeting.

EUROPEAN OEM REVIEWS

General Issues

European OEM sales are complaining about our LAN sales force taking the business direct. Some countries seem to have a hit list to replace OEM sales and not Novell sales, Schneider & Koch in Germany is miffed, so is Siemens. Looks like the sales force is not always following the rules of engagement - no surprise to me! Time to reinforce?

Sweden

No surprises. Victor wants to extend contract locally to secure support. Tandy will have to make a decision. Nokia had a bad quarter and the outlook is flat. Nokia is another miffed LAN Man OEM who has expressed shipping FG instead of OEM product. This could be complicated if European distribution agreements are not welcome and subs do not cooperate when small units are being purchased. We need a cheaper 400 BPI OEM Mouse to stop Logitech gaining market share (and reduce retail pricing).

UK

Seeing more Notebooks designed/imported with DRI ROM-DOS. DRI's main thrust in the UK is to sell Flex-DOS and CP/M to imbedded system vendors. Business outlook is solid, only UPB problem with Torus and Apricot - both are being addressed. Biggest issues - resources: we

[Plaintiff's Exhibit 638 p. 5 of 6]

need dedicated administrative support for SandyO's group and should add al least one additional AM for FG & DOS. We are missing opportunities. Ideally we hire a Pakistani or Indian person.

France

Rumor has it that ailing Goupil is being bought by Siemens/Nixdorf and Fujitsu is interested in Siemens/Nixdorf who is hurting financially. Expect Goupil to ask for min commit relief. ChrisA is leaving the group to work for PascalM. He's frustrated with running a shrinking business. More about this later.

Spain

Ignacio asked for resources to service Investronica who is getting PCs through Chi Leog. I would not recommend this. 30 OEM people are enough for his royalty business.

Holland

Tulip did not do a LAN Man deal because of our retail distribution. They are considering dropping OS/2 as well. (Want to buy FG instead). This would mean they can buy LAN Man for direct sales only according to our local policy. But Tulip sells through resellers. Bernard, make sure we do not cut a local deal which violates your policy.

Italy

Business is within expectations. Olivetti needs more attention as you can see from their recent letter. Maurb is going to get a fulltime AM in July - finally! He should quit dealing in MSU issues as well and I am unsure why he spends so much tone with BSA activities. Their FG-DOS volume is extremely low and time spent here might increase sales. We should move Lorip into the Redmond OEM group. 95% of her time is spent on Olivetti support.

Germany

Worse than Japan is my only comment:
In talking to CWedell and JoachenH, it is clear that they want to replace Manfred, but this is not enough. They just do not and cannot take on the number of customers they need to serve. The group's admin sells FG-DOS over the phone and does not give sufficient admin support to the rest of the group which delays contracts by 3-6 months (no kidding!) and/or delivers sloppy work. With 20M$ business, they need one fulltime manager, four people plus a dedicated admin. They have 1/2 manager for OEM royalty business. 2 AMs and .25 of an Admin at the most. During mid-year review and additional HC was rejected - why? The same resource shortage exists in the FG-DOS sales team. They cannot serve 250 (going to 350) customers with 2-3 people. It takes at least 5-6 to serve and nurture them. This is a 20M$ business on its own, why can't anybody else see this? The result is a demoralized sales force being swamped with customer request who are unable to respond timely and as a result only in Germany does DRI has a stronghold! Bernard please take note and decisive action!

OEM in Europe

A. Communication with customers needs to be improved by direct mail and an installation of a FAX server service. At the same time. I see a need for a semi-annual OEM briefing. We will start 05/31/91 - 12 months between the US meeting and a stretched sales force just does not get the job done.

B. OEM is seen as being a non-attractive group. Stock options are not used properly to attract talent and keep it. Some people complained that their increases and options were decided by local management who does not understand or appreciate what the OEM sales people are doing. Great input. I will insist for the next review not only to send you my evaluation but close the loop to review what local management proposes. Some people even thought that

[Plaintiff's Exhibit 638 p. 6 of 6]


reporting directly to Jeff would be the best solution. I tend to agree, why not follow the ITIS/MSU model?

C. Because of our resource shortage we dont stand a chance to support customers in special/key project situations. I will dedicate some HQ marketing headcount immediately to work on some key project with European customer and the product groups.


PLAINTIFF'S EXHIBIT 709 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 709 p. 1 of 4]

[Plaintiff's Exhibit 709 p. 2 of 4]

[Plaintiff's Exhibit 709 p. 3 of 4]

[Plaintiff's Exhibit 709 p. 4 of 4]

[STAMPED] MS 0084769 CONFIDENTIAL
[STAMPED] EXH  47 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER

>From bradsi Mon Apr 29 11:05:50 1991
To: bobgu davidcol philba tunbr
Subject: Re More Undocumented Windows
Date: Mon Apr 29 11:05:46 1991

totally agree, bob!

| > From bobgu Mon Apr 29 09:23:45 1991
| To: bradsi davidcol philba timbr
| Subject: Re: More Undocumented Windows
| Date: Mon Apr 29 09:18:35 1991
|
|
|
| | | Brad, one very major item I forgot in my previous list:
| | |
| | | Microsoft really need to publish the source code for all the
| | | built-in controls (EDIT, etc.), just like the source for DefWindowProc
| | | was made available. It's nearly impossible to subclass things in a
| | | rational way, because it's difficult to tell how they implement their
| | | behavior in tbe first place. This of course has been a common complaint
| | | for years
|
| Call me paranoid, but the day we publish sources to the controls is tbe day
| we can no longer enhance them. This isn't just speculation. We have has
| to add numerous work-arounds in our code to satisfy some app that has
| reverse-engineered some internal feature of a control and assumed that it
| would never change.
|
| The basic question is how hard we want to push for a common user interface
| vs the desire of ISV's to create custom controls by hacking ours. Hopefully,
| tbe AFX project will eliminate this problem.
|
|
| | | Frankly, I also think it would be useful to release in some wau
| | | the source code for various system utilities, such as Program Manager,
| | | File Manager, Task Manager, etc. Perhaps Microsoft already does this
| | | through "Microsoft University" courses? Actually, that would make a very
| | | interesting and useful book, by tbe way, sort of along the lines of tbe
| | | 'Inside Windows* book you were discussing: a detailed walkthrough of the
| | | source code for several key Windows components. Would be very enlightening.
|
| The question here is when are we going to publish our shell API's? To date,
| all tbe "juicy" API's needed to write a shell have been undocumented.
|
| Soap Box Statement
| ------------------
| This group has been MUCH to lax when it comes to adding API in the product.
| Both external and internal. We don't do any sort of API policing within
| the Group. ISV usability issues are rarely considered. For example, the
| palette support API is a total mess because the developers doing the work
| decided what the API would look like without regard to how the ISV would use
| it. For undocumented API's we add them at will without thinking about
| whether or not they should be documented. They are hacked in ugly things
| that often eventually get documented. PrestoChangoSelector is one of these.
| We need to do l better job at this.
|


################################################################### 6
>From bradsi Mon Apr 29 11:10:21 1991
To: jimall
Subject: Re: FW: Re: change in Win3.1
Date: Mon Apr 29 11:10:20 1991

thanks, i agree with ericr's viewpoint. we should do it
right, if we can.

schedules for 3.1: will ship the same time, but we do have to fix
some additional things we hadn't planned on.



################################################################### 7
>From bradsi Mon Apr 29 11:56:04 1991
To: bradc
Subject: Re: dos announce
Date: Mon Apr 29 11:56:03 1991

what do you think? i'd rather not send the whole development team.
but if they'd be very bummed then I'd reconsider, did the whole developement
team go to the win announce?



################################################################### 8
>From bradsi Mon Apr 29 11:58:47 1991
To: pattys
Cc: jodys
Subject: ISV support issues
Date: Mon Apr 29 11:58:44 1991

good ideas!

| >From bobgu Mon Apr 29 11:20:34 1991
| To: bradsi
| Subject: ISV support issues
| Date: Mon Apr 29 11:16:03 1991
|
| There needs to be SENIOR developers/program managers in the Windows group
| whose job it is to:
|
|   - Act as PSS-Development intermediaries
|   - Reduct the support overhead from developers
|   - Feedback product weaknesses to development based on ISV feedback
|   - Work closely with User-Ed to create usefull manuals
|   - Create real-world sample apps that do real things.
|   - Monitor CompuServe and other popular BB's to gleen usefull information
|
| These people need to be free from any product responsibilities and be 100%
| totally focused on driving User-Ed and PSS to provide quality ISV support.
| The reporting structure of these people has to be free and clear of the
| groups that have product responsibilities. Take the past SDK groups as
| an example of what happens when ISV support people are a sub-group of the
| retail product development group.
|
| It's great to hear that you get 5 new heads to do things like this. I hope
| they don't get mired down in the political muck of getting a product out
| the door.
|
| = = = = = = = = = = = = = = =
|
| Now, here's the real kicker - wouldn't it be a good idea to have all
| of this in place and available for Win 4 by the time ISV's started using
| the product???!!!
|
| = = = = = = = = = = = = = = =
|
|
| That's all for now....
|
| -BobGu
|


################################################################### 9
>From bradsi Moo Apr 29 11:59:31 1991
To: bobgu
Subject: Re: ISV support issue
Date: Mon Apr 29 11:59:30 1991

great ideas. I've talked to pattys (gm pss) about having you
on the developers support strike team, and she agreed,
you should hear from her shortly (maybe today).



###################################################################
10
>From bradsi Moo Apr 29 12:00:30 1991
To: davidcol
Subject: some ideas
Date: Mon Apr 29 12:00:29 1991

| >From stewlc Mon Apr 29 11:29:41 1991
| To: bradsi steveb
| Subject: some ideas
| Date: Sun Apr 28 11:26:36 1991
|
| 1. To Improve Win3.1 Robustness:
|
| * Buy "Robustness" Tools
|    Purchase a suite of Win backup/recovery tools from Symantec
|    or some other vendor and bundle with Win3.1.
|
| * Completely Eliminate the UAE Message.
|    Replace it with a "Global Protection Fault" message ala OS/2 2.0 so that we
|    appear no worse than OS/2 2.0. Plus make attempts it diagnosing
|    the problem with a help dialog box. We will still be inferior in that
|    we require rebooting. But we will signal to users that we've revamped our
|    error handling, that the scheme is similar to OS/2 2.0's, and
|    that it is in some ways better because of our help dialog.
|
|
| 2. To Discount the Better-Windows-than-Windows Claim:
|
| * Put in OS/2 2.0 and WLO-detection into Win3.1.
|    If an app is WLO, then run it. If OS/2 2.0 exists and the app is not
|    a WLO app, then put up an error message indicating that:
|    - the app is not certified by MS
|    - that it can be run at one's own risk.
|    Hard for IBM to make a Better Windows claim with such a scheme in place.
|
| Once we detect the existence of OS/2 2.0, there may be other special
| warnings we can emit that highlight legitimate problems with Windows apps
| running on top of OS/2 2.0. For example, a message, each time one
| prints, that says the the Windows and OS/2 printer models are different
| and that output will be different between the two systems.
|


###################################################################
11
>From bradsi Mon Apr 29 12:04:17 1991
To: tunya
Subject: Re: DOS/Win and IBM
Date: Mon Apr 29 12:04:13 1991

a good sign


PLAINTIFF'S EXHIBIT 805 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 805 Page 1 of 8]

[Plaintiff's Exhibit 805 Page 2 of 8]

[Plaintiff's Exhibit 805 Page 3 of 8]

[Plaintiff's Exhibit 805 Page 4 of 8]

[Plaintiff's Exhibit 805 Page 5 of 8]

[Plaintiff's Exhibit 805 Page 6 of 8]

[Plaintiff's Exhibit 805 Page 7 of 8]

[Plaintiff's Exhibit 805 Page 8 of 8]


[STAMPED] SIGNED ORIGINAL
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] AMENDMENT
[STAMPED] X 222282 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT NO 609 FOR IDENTIFICATION DATE: 5-4-98 RPTR: [initials]
[STAMPED] MS98 0186799 CONFIDENTIAL

AMENDMENT NUMBER 2

Dated July 1.1991

TO THE LICENSE AGREEMENT

Dated September 1,1990
Numbered G15-04271

for

Microsoft MS-DOS, Version 4.01
Microsoft Works, Version 2.O

Between

MICROSOFT CORPORATION,
a Delaware Corporation,

and

VOBIS DATA COMPUTER GMBH,
a German Corporation



Effective July 1.1991. the below signed parties agree that the indicated portions of the above referenced License Agreement (hereinafter the "Agreement") are hereby amended by this instrurnent (hereinafter the "Amendment"), as follows

1. Section 9, entitled "TERM OF AGREEMENT", is hereby amended and as amended shall read as follows:

9. TERM OF AGREEMENT
Provided this Agreement has been properly executed by COMPANY and by an officer of MS. the initial term of this Agreement ("Initial Term") shall run from the EFFECTIVE DATE until the earlier of: (i) two (2) years from the end of the calendar quarter in which COMPANY's first shipment to a customer of any Product for revenue occurs; or (ii) two (2) years and six (6) months from the end of the calendar quarter during when the EFFECTIVE DATE occurs. COMPANY shall give MS written notice of the first date of shipment of any Product by COMPANY to a customer for revenue.

2. Section 20. entitled "PRIOR AGREEMENT", is hereby added and as added shall read as follows:

20. PRIOR AGREEMENT
Provided that, as of the EFFECTIVE DATE of this Amendment. COMPANY has paid MS all amounts due under G150-9325 dated December 1, 1989 between COMPANY and MS (hereinafter the "PRIOR AGREEMENT"), and COMPANY has complied with all the terms and conditions of PRIOR AGREEMENT:

a) COMPANY shall receive a credit equal to: i) the minimum commitment payments made by COMPANY pursuant to Exhibit B of the PRIOR AGREEMENT, minus ii) the amount of said minimum commitment payments determined to represent earned royalties in accordance with the PRIOR AGREEMENT. If earned royalties exceed such minimum commitment payments there shallbe no credit. If such minimum commitment payments exceed earned royalties, the excess prepaid royalties shall be transferred to this Agreement and shall be recoupable against earned royalties.

b) The PRIOR AGREEMENT shal lbe deemed to be terminated as of the EFFECTIVE DATE hereof.

3. In Exhibit C3, the table in subsection (a) of the ROYALTY PAYMENTS AND REPORTING REQUIREMENTS section is hereby amended and as amended shall read as follows:

Customer System           Royalty Rate ($US)

Exhibit Ml                     $13.05

4. Exhibit C4 and C8 is hereby amended and as amended shall read as set forth in the attached Exhibit C4 and C8.

IN WITNESS THEREOF, the parties have executed this Amendment as of July ., 1991.All signed copies of theis Amendment shall deemed to be originals.

MICROSOFT CORPORATION VOBIS DATA COMPUTER GMBH

[signature]Joachim Kempin [signature] Theo Lieven

By By

Joachim Kempin Theo Lieven
Name (Print) Name (Print)

Vice President, OEM Sales General Manager
Title Title

3/9/92 Dec. 27, 1991
Date Date


EXHIBIT C4

ADDITIONAL LANGUAGE VERSIONS

The Product described in Exhibit C3 ("Reference Exhibit") shall include the additional language versions identified below. COMPANY shall license at most one language version of the Product for use on each applicable CUSTOMER SYSTEM. COMPANY'S royalty obligations shall be as set forth in the Reference Exhibit regardless of the language version licensed with each applicable CUSTOMER SYSTEM. Notwithstanding the preceding sentence, COMPANY shall pay MS an additional language version royalty equal to fifteen percent (15%) of the royalty otherwise payable to MS for each full or partial copy of each language version of the Product identified below which is licensed or otherwise disposed of by COMPANY during the term of this Agreement. COMPANY thai report, on a language version by language version basis, and pay such language version royalties pursuant to and as part of its royalty payment and reporting obligations under the Royalty Payments and Reporting Requirements section of the Reference Exhibit. The Product Name and Associated Trademark are, in all Language Versions listed below, the same as in the Reference Exhibit

1. Language Version: Danish

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2DK that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

2. Language Version: Dutch

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2NL that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

3. Language Version: Finnish

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2SF that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

4. Language Version: FRENCH

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2F that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

5. Language Version: German

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2D that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

6. Language Version: Italian

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2I that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

7. Language Version: Norwegian

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2N that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

8. Language Version: Portuguese

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2P that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

9. Language Version: Russian

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2SU that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

10. Language Version: Spanish

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2E that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

11. Language Version: Swedish

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2S that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

12. Language Version: Turkish

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D701-2TR that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

Exhibit to the License Agreement dated September 1,1990, between MICROSOFT CORPORATION and VOBIS DATA COMPUTER GMBH
02/26/91 0975L

EXHIBIT C8

ADDITIONAL LANGUAGE VERSIONS

The Product described in Exhibit C7 ("Reference Exhibit") shall include the addffioral language versions identified below. COMPANY shall license at most one language version of the Product for use on each applicable CUSTOMER SYSTEM. COMPANY'S royalty obligations shall be as set forth in the Reference Exhibit regardless of the language version licensed with each applicable CUSTOMER SYSTEM. Notwithstanding the preceding sentence, COMPANY shall pay MS an additional language version royalty equal to fifteen percent (15%) of the royalty otherwise payable to MS for each full or partial copy of each language version of the Product identified below which is licensed or otherwise disposed of by COMPANY during the term of this Agreement. COMPANY shall report, on a language version by language version basis, and pay such language version royalties pursuant to and as part of its royalty payment and reporting obligations under the Royalty Payments and Reporting Requirements section of the Reference Exhibit. The Product Name and Associated Trademark are, in all Language Versions listed below, the same as in the Reference Exhibit.

1. Language Version: Dutch

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5NL that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

2. Language Version: French

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5F that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

3. Language Version: German

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5D that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

4. Language Version: Italian

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5I that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

5. Language Version: Portuguese

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5P that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

6. Language Version: Russian

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5SU that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

7. Language Version: Spanish

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5E that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

8. Language Version: Swedish

Product Deliverables:
(a) Product in Object Code Form
(b) Standard document in Series Number D781-5S that MS delivers with the Product
Copyright Notice (if not completed, the Copyright Notice shall be as set forth in the Reference Exhibit):
COPYRIGHT MICROSOFT CORPORATION, 1987 - 1990 . All rights reserved.

Exhibit to the License Agreement dated September 1,1990. between MICROSOFT CORPORATION and VOBIS DATA COMPUTER GMBH.
03/01/90 0975L


PLAINTIFF'S EXHIBIT 869 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 869 Gordon v. Microsoft p. 1 of 2

[STAMPED] Depo. Ex. 97

>From stevetho Fri Jul 26 13:48:24 1991
To: bradsi jancl janinah theresa
Cc: dand danq davidcol karlst philba ralphl
Subject: RE: FW: Re: Digital Research
Date: Fri Jul 26 14:47:37 PDT 1991

To say that we don't support EMM for Win31 is one thing. To say that we don't support DR is quite another. They are OnLine customers. My job is to support our customers to the best of my ability.

This is the same situation we were in when IBM sent in certain unfavorable questions. We better start drawing the line as to WHAT we support, not who. As long as these companies are our customer we owe them the best support possible. If we're not going to support them, somebody ought to tell them that.

Telling DR that we won't support them because they're not an offical [sic] beta site buys us some time but when 3.1 is released we'll have to come up with another excuse. I'd like to be prepared when that time comes. Can we get a decision on this soon -- please!

Steve

>From janineh Fri Jul 26 13:53:34 1991
To: bradsi
Cc: dand danq daavidcol karlst philba ralph1 stevetho
Subject: FW: Re: Digital Research

Date: Fri Jul 26 14:01:26 1991

I've talked w/ Steve Thompson. DRI was notified by a Win 3.1 Beta tester that it wasn't working w/ DR-DOS. Steve will tell them that he can't help them because they aren't in the beta program and because we don't answer beta questions via regular Online accts. This is a consistent message we have been telling anyone asking 3.1 questions via regular online. If they ask to be in the beta, we will just tell them the standard answer, write or fax Windows Beta.

As a follow-up I'm going to work w/ both Dand and Danq about people on the black list.

Janine

>From bradsi Fri Jul 26 11:49:39 1991
To: davidcol karlst philba ralph1
Cc: Kalak [indistinct]
Subject:  Re: Digital Research

Date: Fri, 26 Jul 91 11:47:21 PDT

How does DRI have Win 3.1 in the FIRST PLACE?????????????? They are on the beta blacklist!!
Is the report from DR or a customer who is a beta tester of 3.1 who also uses dr dos?

>From ralphl Fri Jul 26 10:26:22 1991
To: bradsi
Subject: Digital Research
Date: Fri Jul 26 10:25:45 1991

What is our position wrt helping DR with problems?

>From danq Fri Jul 26 08:48:26 1991
To: greglo ralph1
Cc: marthawb [indistinct] neilsa stevetho
Subject: Re: Loadhi.VxD - Digital Research
Date: Fri Jul 26 08:46:47 PDT 1991

Greg or Ralph,

[STAMPED] MS-PCA 1178475 CONFIDENTIAL
[STAMPED] X 584927 CONFIDENTIAL
[STAMPED] EXH 38 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] PLAINTIFF'S EXHIBIT 401 C. A. No 2:96CV645E

Plaintiff's Exhibit 869 Gordon v. Microsoft p. 2 of 2

Didn't I hear a comment one day in either of your offices re this. What should we tell DR

Thanks,
Dan

>From stevetho Thu Jul 25 11:20:41 1991
To: danq neilsa
Cc: marthawb
Subject: Re: Loadhi.VxD - Digital Research

Date: Thu Jul 25 11:17:03 PDT 1991

These guys need to know what they need to do to get their memory manager to work in 3.1

Dan- Do you know where we can get this info?

Thanks,
Steve

>From neilsa Mon Jul 22 12:56:59 1991
To: danq stevetho
Cc: marthawb
Subject: Re: Loadhi.VxD - Digital Research

Date: Mon, 22 Jul 91 12:55:56 SPD

There is no loadhi VxD available for Win31 because that functionality has been built directly into VMM. The old VxD should detect this and not load on a 3.1 system, which is how it should be. So it's hard to say what problem they are having...

I do know that there were problems with OEMM in this area, but I'm not privy to that information...


>From stevetho Mon Jul 22 11:25:51 1991
To: danq
Cc: marthawb neilsa
Subject: Loadhi.VxD - Digital Research
Date: Mon Jul 22 11:25:09 PDT 1991

Hey dude,

A few months ago Neil helped DR with a Loadhi problem with Windows. Well, they got everything working for 3.0 but now they say they break in 3.1. Can you find out if there's a new Loadhi we can ship?

Thanks,
Steve

[STAMPED] X 584928 CONFIDENTIAL
[STAMPED] MS-PCA 1178476 CONFIDENTIAL

PLAINTIFF'S EXHIBIT 874 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 874 Page 1 of 1]


######################################################## 240
>From bradsi Mon Jul 29 07:29:48 1991
To: karlst philba ralphl steveb
Subject: Re: Digital Research
Date: Mon 29 Jul [19]91 07:29:40 PDT

you bet. dri is in on the dos/win blacklist.  nothing gets shipped from this BU without the blacklist being checked.

i made it clear to the online guys that windows supports ms dos 3.1 and above.  we are not to help dri.  I will also talk to pattys about this.

in this case, a 3.1 beta tester discovered that drdos doesn't work with win 3.1 so called dr for help. dr called us for help on online.  the online guys wanted to help, because "dr is a paying customer". i told them no.

From steveb Sun Jul 28 21:35:21 1991
To: bradsi karlst philba ralphl
Subject: Digital Research
Date: Sun Jul 28 21:35:18 1991

bard [sic, "brad"] pls make sure we are not supporting DRI anywhere in the company with this stuff thx

######################################################## 241
>From bradsi Mon Jul 29 07:35:30 1991
To: jancl pattys
Subject: Digital Research
Date: Mon, 29 Jul 91 07:35:27 PDT

We should not be providing Digital Research any assistance getting their os to work with our software. Our software supports ms dos, not dr dos. It's completely up to them to figure out and resolve any problems that may occur.

Just recently, a situation came up where a win 3.1 beta tester discovered that dr dos doesn't work with win 3.1 and called dr for help. DR submitted an SR to their online account, and the techs were dutifully following up very conscientiously. We are not to help them, paying customer or not. Windows supports MS-DOS and PC-DOS 3.1 and above.

Thanks.

######################################################## 242
>From bradsi Mon Jul 29 07:36:48 1991
To: philba tomle
Subject: nigelt
Date: Mon 29 Jul [19]91 07:36:45 PDT

the nt guys had big problems with nigel.  please talk to paulma to understand what happened.  I will do the same when I get back.

######################################################## 243
>From bradsi Mon Jul 29 07:43:15 1991
To: russs
Subject: Re: FW: Intelligence on Novell-DR DOS deal
Date: Mon, 29 Jul [19]91 07:43:13 PDT

thanks

the tough question in all this is whether novell's intentions are offensive or defensive.  there are signs of both.  some novell people (and market watchers) are saying defensive.  other statements by novell, including darrell miller, clearly are offensive (control of the desktop).

[STAMPED] EXH 8 DATE 2/13/[20]02 WITNESS Silverberg MARY W. MILLER
[STAMPED] MS-PCA 1178530 CONFIDENTIAL
[STAMPED]  MS 5050789 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT 407 C.A. No. 2:96CV645B

PLAINTIFF'S EXHIBIT 959A Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 959A Gordon v. Microsoft p. 1 of 2

To: winbug
Cc: bambi scottq lawren
Subject: Bambi:&DR DOS 6.0, many files turned into crosslinked mess

Windows Bug Report
Title:
Bambi:&DR DOS 6.0, many files turned into crosslinked mess
Opened By: lawren
Sev: 1
WinVer: 3.10 47
Product: Drivers
Area: Bambi
Origin: Internal
Sugg. Assign: scottq
DOSVer: 6.00
Mode: high
Name:          Phone:

Problem Statement:

Had run fine other times. Did a "DIR" and I got garbage characters. Chkdsk found well over 100 files with crosslinked files, invalid clusters, and invaild [sic] & incorrect lengths.

Machine Configuration:

Cumulus 386 2M VGA

Config.Sys:

Autoexec.Bat:

Form #; SYS003       Version #: 2.00     Revision Date: 08/22/[19]91

################################################################## 476
>From karlst Thu Sep 19 11:55:06 1991
To: steveti
Cc: bradsi philba
Subject: jeanp/vacation
Date: Thu Sep 19 11:34:54 1991

I believe we've settled the issue with jeanp regarding his vacation. Don't call him. Kudos to bradsi for making this happen

################################################################## 477
>From philba Thu Sep 19 11:58:17 1991
To: lawren
Cc: bambi bradsi lawren scottq
Subject: Re: Bambi:&DR DOS 6.0, many files turned into crosslinked mess
Date: Thu, 19 Sep [19]91 11:34:11 PDT

[STAMPED] MS 5055905 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT 479A C.A. No. 2:96CV645B
[STAMPED] MS-PCA 1195217 CONFIDENTIAL

Plaintiff's Exhibit 959A Gordon v. Microsoft p. 2 of 2

tee hee hee...thats [sic] what you get for running an incompatible OS

sorry, I couldnt' resist. Scott, let's figure this out asap - could be a problem thats [sic] lurking for real DOS.

>From lawren Thu Sep 19 11:34:15 1991
To: winbug
Cc: bambi scottq lawren
Subject: Bambi:&DR DOS 6.0, many files turned into crosslinked mess

Windows Bug Report

Title: Bambi:&DR DOS 6.0, many files turned into crosslinked mess
Opened By: lawren
Sev: 1
WinVer 3.10 47
Product: Drivers
Area: Bambi
Origin: Internal
Sugg.Assign: scottq
DOSVer: 6.00
Mode: high
Name:       Phone:

Problem Statement:  Had run fine other times. Did a "DIR" and I got garbage characters. Chkdsk found well over 100 files with crosslinked files, invalid clusters, and invalid & incorrect lengths.

Machine Configuration: Cumulus 386 2M VGA
Config.Sys:
Autoexec.Bat:
Form #: SYS 003  Version #: 2.00  Revision Date: 08/22/[19]91

####################################################### 478
>From greglo Thu Sep 19 12:10:11 1991
To: bradsi
Subject: Re: bambi
Date: Thu, 19 Sep [19]91 12:05:40 PDT

The new WzMail version (which uses the WinMail document handling routines rather than the ancient DH library) also gives you a huge perf boost. If you use WzMail you want it.

[STAMPED] MS 5055906 CONFIDENTIAL
[STAMPED] MS-PCA 1195218 CONFIDENTIAL


PLAINTIFF'S EXHIBIT 979 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 979 Gordon v. Microsoft p. 1 of 2

[HANDWRITTEN] Non-responsive material redacted

######################################################## 603
>From bradsi Sat Sep 28 10:54:53 1991
To: joachimk peterbra steveb
Cc: alexn jeffl jonro markbu richab richardf ronh teresach
Subject: Re: Compaq Windows status
Date: Sat, 28 Sep [19]91 10:54:43 PDT

you're asying [sic, prob. "saying"] that when someone buys compaq dos for $99, they also get windows for free.  but if you want windows alone, it cost you $150

and compaq wants windows for free

am I missing something why this is good for us?

don't forget that today, retail is still 61% of windows revenue.

######################################################## 604
>From bradsi Sat Sep 28 11:23:54 1991
To: philba
Subject: Re: dr
Date: Sat, 28 Sep [19]91 11:23:53 PDT

let's talk.

>From philba Sat Sep 28 10:59:13 1991
To: bradsi
Cc: karlst
Subject: Re: dr
Date: Thu 26 Sep [19]91 10:45:39 PDT

Sorry for the silence -- dont [sic] interpret it as ignoring you.

The approach that ralph and I have discussed is to use a vxd to extend' dos by patching it. In this case, we would create a subfunction in the findfirst/findnext family -- findabunch to allow filemanager to make a single call to get directory information. We would not patch unknown OSs and most likely would only patch MS DOS 5.x. The big advantage here is that it provides a legitimate performance improvement.  However, it won't prevent us from running on foriegn [sic] OSs (unless we explicitly decide to refuse to run) -- they just wont [sic] run as fast.

This hasn't been implemented yet and to be honest it sends a mixed message about project team focus. The work necessary would be 3 days of ralphl's time to implement and then a somewhat open-ended use of it by various components. We definitely want to do winfile and user file list box.  Neither of which is a huge amount of time (1-2 days each to implement and thoroughly test). The network guys may want to implement this in their redir.

[STAMPED] DEPOSITION EXHIBIT 1114
[STAMPED] MS 0098698 CONFIDENTIAL
[STAMPED] MS98 0181665 CONFIDENTIAL

Plaintiff's Exhibit 979 Gordon v. Microsoft p. 2 of 2

Is this the approach you want to take? Or would you prefer a simple check and refuse to run? That's a lot easier but clearly quite defeatable. I'll come and talk to you about it.

>From bradsi Fri Sep 27 22:44:35 1991
To: philba
Subject: dr
Date: Fri Sep 27 22:44:10 1991

can you tell me specifically what we're going to do to bind ourselves closer to ms dos?  since you haven't been replying to my messages, I do n't [sic] know how to interprety [sic] your silence. Let me emphasize the importance: ibm is going to annoucne [sic] the drdos deal at comdex (almost 100% certain).

OK?

####################################################### 605
>From bradsi Sat Sep 28 11:35:49 1991
To: joachimk peterbra steveb
Cc: alexn jeffl jonro markbu richab richardf ronh teresach
Date: Sat, 28 Sep [19]91 11:35:40 PDT

we cannot have compaq sell Win+OS product that is cheaper than what we sell windows alone for.  even if it's [sic] works just on compaq machines, we thus tell the market that we are dramatically overcharging for windows.  and our retail business goes down the tubes.  if they want to offer their dos,  which includes windows,  for $200, then we're talking.

yes, I want the compaq business but not to give to them for nothing and kill my retail business, too.

they realize they are falling behind and they have to do something to fix it.

####################################################### 606
>From bradsi Sat Sep 28 11:40:25 1991
To: karlst lisacr
Subject: Re: file open/save as changing dir behavior
Date: Sat, 28 Sep [19]91 11:40:23 PDT

ok. include 123/w. let us know.


####################################################### 607
>From bradsi Sat Sep 28 11:42:13 1991
To: jimall paulma
Subject: RE sys mgt
Date: Sat, 28 Sep [19]91 11:42:11 PDT

ok they are yours.

the winlogin schedule has always been dec-jan.  should finish dec but if it stretches into jan, you shouldn't count on them jan 1.

>From jimall Sat Sep 28 11:39:04 1991
To: bradsi paulma
Subject: RE: sys mgt
Date: Sat Sep28 11:38:38 PDT 1991

I have been assuming ericpe was coming to win4 by year end.  It sounds like this winlogin is slipping.  I will be happy to take tedst too.

jim

>From bradsi Fri Sep 27 10:10:57 1991
To: jimall paulma
Subject: sys mgt

Date: Fri, 27 Sep [19]91 10:08:19 PDT

[STAMPED] MS 0098699 CONFIDENTIAL
[STAMPED] MS98 0181666 CONFIDENTIAL



PLAINTIFF'S EXHIBIT 981 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 981 Gordon v. Microsoft p. 1 of 2


* Do we want to do a [sic] IBM only Slick for new HW? Makes sense to me. This is straight forwared using approach I've outlines [sic] for our OEM Packaged Product (and the work to do this is only a minor increment over what we're already planning on doing for the generic packaged product).

Lest there be objections, I intend to have Davebe start on OEM DOS Setup changed to support OEM Packaged Slick product as soon as some cycles free up (but lower priority than DOS 5.0a).

Comments?

Eric.

######################################################## 323
>From mackm Mon Sep 30 08:10:33 1991
To: bens raype
Cc: bradsi
Subject: FW: david bernard
Date: Mon Sep 30 08:13:27 PDT 1991

Good News!     Thanks Brad.

>From bradsi Fri Sep 27 13:59:53 1991
To: lynner mackm
Subject: david bernard

Date: Fri, 27 Sep [19]91 13:58:11 PDT


finally talked to him. he's going to try to come out end of next week. We are still a bit apart on salary, I didn't up it. I only want to do that once and close him on the spot. Hopefully that can happen next week.

######################################################## 324
>From alexn Mon Sep 30 08:11:16 1991
To: ericst
Cc: bradsi jonro
Subject: FW: oem weekly
Date: Sun Sep 29 08:10:02 PDT 1991

Can we meet this week to get the first issue defined and initiated? I am out of town tomorrow, but back in for the rest of the week.

How does Wednesday look?

>From bradsi Sat Sep 28 13:08:11 1991
To: alexn ericst
Cc: davidcol richt tomle
Subject: oem weekly

Date: Sat, 28 Sep [19]91 13:05:57 PDT

would be nice to get this started and keep it going on a weekly basis. thanks.

######################################################## 325
>From philba Mon Sep 30 08:15:02 1991
To: bradsi
Cc: davidcol
Subject: Bambi on DR-DOS 6.0
Date: Sat, 28 Sep [19]91 07:57:50 PDT

heh, heh, heh....

my proposal is to have bambi refuse on this alien OS. comments?

[STAMPED] MS-PCA 1179289 CONFIDENTIAL
[STAMPED] MS 0098785
[STAMPED] PLAINTIFF'S EXHIBIT 504 C.A. No. 2:96CV645B

Plaintiff's Exhibit 981 Gordon v. Microsoft p. 2 of 2

>From chuckst Sun Sep 29 17:16:46 1991
To: mikedr philba scottq
Subject: Bambi on DR-DOS 6.0
Date: Sun Sep 29 17:16:39 1991

I tracked down a serious incompatibility with DR-DOS 6 -- They don't use the 'normal' device driver interface for >32M partitions. Instead of setting the regular START SECTOR field to 0ffffh and then using a brand new 32-bit field the way MS-DOS has always done, they simply extended the start sector field by 16 bits.

This seems like a foolish oversight on their part and will likely result in extensive incompatibilities when they try to run with 3rd part[y] device drivers.

I've patched a version of Bambi to work with DRD6, and it seems to run Win 3.1 without difficulty. This same problem may have caused other problems with Win 3.1 and the swapfile under DRD6.

It is possible to make Bambi work, assuming we can come up with a reasonably safe method for detecting DRD6. The runtime hit would be minimal in time and space, although we would have a couple of instructions in the main code path for checking the 'special' DRD6 flag.

What do we think? Should we test further with the patched Bambi to see if there are any more incompatibilities????


####################################################### 326
>From philba Mon Sep 30 08:16:27 1991
To: bradsi davidw
Subject: Re: Intel/Dell Frame Buffer Windows Video Driver
Date: Sat, 28 Sep [19]91 08:02:47 PDT

hmm, putting 2 and 2 together. He has been around a lot lately.

Brad, FredE is Fred Einstein who wrote the original (and highly buggy) 8514 driver and is not held [in] high regard by many windows developers.

>From davidw Sun Sep 29 21:20:00 1991
To: bradsi philba
Subject: Intel/Dell Frame Buffer Windows Video Driver
Date: Sun 29 Sep [19]91 21:19:55 PDT

(i haven't gone through all of my email yet, but ...)

why does this have frede written all over it?

david

>From philba Fri Sep 27 09:47:43 191

never heard of it.

>From bradsi Fri Sep 27 09:33:22 1991

know anything about this?

>From carls Fri Sep 27 09:26:40 1991

I just saw some email that Intel is having some problems with a Windows driver for the joint Intel/Dell video frame buffer. I don't know if this is for Windows generally or just for the Multi-media extensions. This is where Intel/Dell hired MCS to do the driver.

We are having an exec meeting at 11am. It would be best if Billg and Bradsi understood the issues prior to 11am so that we can respond to either complaints or requests for help. Please send mail on the status of this.

####################################################### 327

[STAMPED] MS 0098786 CONFIDENTIAL
[STAMPED] MS-PCA 1179290 CONFIDENTIAL

PLAINTIFF'S EXHIBIT 984 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 984 Page 1 of 1]

buggy) 8514 driver and is not held high regard by many windows developers.

> From davidw Sun Sep 29 21:20:00 1991
To: bradsi philba
Subject: Intel/Dell Frame Buffer Windows Video Driver
Date: Sun 29 Sep [19]91 21:19:55 PDT

(i haven't gone through all of my email yet, but ...)

why does this have [indistinct. "freda"?] written all over it?

david

>From philba Fri Sep 27 09:47:43 1991

never heard of it.

>From bradsi Fri Sep 27 09:33:22 1991

know anything about this?

>From carts Fri Sep 27 09:26:40 1991

I just saw some email that Intel [indistinct] is having some problems with a Windows driver for the joint Intel [indistinct]/Dell video frame buffer. I don't know if this is for Windows generally or just for the Multi-media extensions. This is where Intel/Dell hired MCS to do the driver.

We are having an exec meeting at 11 am. I twould be best if Billg and Bradsi understood the Issues prior to 11 am so that we can respond to either complaints or requests for help. Please send mail on the status of this.

######################################################## 327
>From davidcol Mon Sep 30 08:16:28 1991
To: bradsi philba
Subject: Bambi on DR-DOS 6.0
Date: Mon Sep 30 08:15:57 1991

It should say unsupported version of DOS.

######################################################## 328
From: karlst Mon Sep 30 08:20:17 1991 [Hour of day is indistinct, might be 06 instead of 08]
To: win31dev winprog
Cc: bradsi jobren [indistinct] philba
Subject: HOTLIST SUSPENDED
Date: Mon Sep 30 08:18:03 1991 [Hour of day is indistinct, might be 06 instead of 08]

[STAMPED] MS 5054012 CONFIDENTIAL
[STAMPED] MS-PCA 1143027 CONFIDENTIAL
[STAMPED] EXH 13 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER


PLAINTIFF'S EXHIBIT 985 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 985 Page 1 of 1]

[Printed text.]

"We are suppose[d] to give the user the option of continuing after the warning. However, we should surely crash at some point shortly later." - and yet Opera continues to spread their FUD about MSN sending bad data to Opera browsers...
------------------------------------
Depo. Ex. 1722

    message during copy files   marcw, andyhi
    swapfile rework             craigc
    machine config. tuning      paulke
    lim compatibility work      mikecole
    dlg for setup from windows  marcw

USER work:

    SendMessage revalidation    mikesch
    Log error conditions        sankar
    gp fault cleanup            neilk
    layer size, speed opts      neilk
    Fix edit ctrl msgs          sankar
    change api's - hInst/hMod   jeffbog
    Local memory validation     jont

GDI/T2 work

    VDMX support                gunterz, jeanp
    Font file preload           amitc
    Cleanup font cache code     gunterz
    Finish remove glbl lock     kensy, davidw, amitc
    Complete Fontmapper work    davidw

Other Critical Work

    Bambi in build 50           philba

p.s. By now, everyone should have seen rickem's mail regarding PUCUS.  We'll need to spend some time today fixing our source tree after Friday's crash.

####################################################### 329
>From davidcol Mon Sep 30 08:24:05 1991
To: karlst philba
Subject: supported DOSes
Cc: bradsi
Date: Mon Sep 30 06:23:47 1991

It's pretty clear we need to make sure Windows 3.1 only runs on top of MS DOS or an OEM version of it.  I checked with legal, and they are working up some text we are suppose[sic] to display if someone tries to setup or run Windows on a[sic] alien operating system.  We are suppose[sic] to give the user the option of continuing after the warning.  However, we should surely crash at some point shortly later.

Now to the point of this mail. How shall we proceed on the issue of making sure Win 3.1 requires MS DOS.  We need to have some pretty fancy internal checks to make sure we are on the right one.  Maybe there are several very sophisticated checks so the competitors get put on a treadmill.  Aaronr had
some pretty wild ideas after 3 or so beers, earleh has some too.  We need to make sure this doesn't distract the team for a couple of reasons 1) the pure distraction factor  2) the less people know about exactly what gets done, the better.

Please advise.

####################################################### 330
>From paulma Mon Sep 30 08:29:45 1991
To: carriet
Subject: Harel Kodesh
Cc: andyp bradsi daveth, karenh(?) nathanm(?)
Date: Mon Sep 30 08:29:22 1991

This is a good guy whom I know who works at Motorola Mobile Data in Bothell. He is a D12/D13 developer/development manager - smart and
lots of energy.

He is tired of "six sigma" bureaucracy at Motorola, and [sic:in] fact that Moto as a company is strapped for cash to invest in its[sic] products - so he is looking at offers from Slate and Rational in the Bay Area. He

_Gates_

[STAMPED] DEPOSITION EXHIBIT _84_ 2/28/02
[STAMPED] MS-PCA 1179299 CONFIDENTIAL
[STAMPED] MS 0098788 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT 510 C.A. No. 2:96CV645B

PLAINTIFF'S EXHIBIT 990 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 990 Gordon v. Microsoft p. 1 of 2

time and effort needed to become productive with their computers.

"Microsoft is committed to supporting our users who choose the Macintosh and CD-ROM as part of their computing environment," said Mike Maples, senior vice president of applications for Microsoft. "With this updated CD-ROM version of The Microsoft Office, we are providing users with an efficient way of getting the applications they need for professional and personal use."

CD-ROM Version Provides Extra Benefits for Users

Because these four applications are included on one compact disc, installing them on the machines becomes a very easy task -- simply run the installation programs right from the CD; no more switching floppy disks in and out.

In addition, the storage capabilities of CD-ROM allow the documentation to be included on the disc. This means users do not have to take the time to search their offices for product manuals, then scan through the manuals looking for the answers. They can access all of the information online -- as they need it. (Hard copies of all documentation are available from Microsoft upon request for $15.50 plus tax.) Another benefit of CD-ROM is the audio segments that are included throughout the online documentation. These spoken tutorials talk users through a process while they watch what is happening on the computer screen.

System Requirements, Pricing and Availability

The Microsoft Office on CD-ROM will be available in early October 1991. Version 1.5 has a suggested retail price of $750 from now until December 31, 1991. Registered users of version 1.0 may upgrade to version 1.5 for $129. (All prices listed are U.S. suggested retail prices.)

The CD-ROM version of The Microsoft Office for the Macintosh requires a Macintosh Plus, Classic, LC, SE, or II-family with an Apple or compatible CD-ROM drive; at least 2 MB of RAM; and System 6.0.4 or higher. Microsoft Mail is compatible with AppleShareR and EtherTalkR, and requires an AppleTalk or compatible local area network. Additional Microsoft Mail server software must be purchased seperately to use the workstation version included on the CD.

Microsoft Corporation (NASDAQ "MSFT") develops, markets and supports a wide range of microcomputer software for business and professional use, including operating systems, network products, languages and applications, as well as books, hardware and CD-ROM products for the microcomputer marketplace.

                        ########

Microsoft and PowerPoint are registered trademarks of Microsoft Corporation. Apple, AppleTalk, AppleShare, EtherTalk and Macintosh are registered
trademarks of Apple Computer, Inc.

####################################################### 441
>From scottq Mon Sep 30 17:48:06 1991
To: bradsi
Subject: bambi
Date: Mon Sep 30 17:48:00 1991

How do you power-down your machine at night? If you are running windows and you power down, you will probably get lost clusters with or without bambi loaded.

Also, you will get lost clusters if you use local reboot.

Scott

####################################################### 442
>From markp Mon Sep 30 17:53:55 1991
To: a-long bradsi brentk chrisg davidds davidtry jodyg johnen jont
    neilk richp richsa sandeeps(?) steveja terrib(?)
Cc: a-janj georga(?)
Subject: Re: bambi v.35
Date: Mon, 30 Sep 91 17:53:27 PDT

Where is it, Jody?

[STAMPED] MS 5049397 CONFIDENTIAL
[STAMPED] MS-PCA 1143037 CONFIDENTIAL

Plaintiff's Exhibit 990 Gordon v. Microsoft p. 2 of 2

>From jodyg Mon Sep 30 17:39:11 1991
To: a-long bradsi brentk chrisg davidds davidtry johnen jontmarkp neilk richp richsa sandeeps steveja terrib
Cc: a-janj georga
Subject: bambi v.35
Date: Mon Sep 30 17:38:53 1991

You should update to the new version.

>From scottq Mon Sep 30 17:27:56 1991
To: bambi georgef
Subject: bambi v.35
Date: Mon Sep 30 17:27:02 1991

Bambi v.35 has passed developer testing. The primary change fixes a major problem with accessing logical units on external hard disks. Also, DR DOS is detected (needs testing!) and bambi refuses to load. Finally, ignore was removed from the pop-up dialog box.

Scott

[handwritten:] Privileged material redacted

[STAMPED] MS 5049398 CONFIDENTIAL
[STAMPED] MS-PCA 1143038 CONFIDENTIAL

PLAINTIFF'S EXHIBIT 1048 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 1048 Gordon v. Microsoft p. 1 of 1

>From bradsi Tue Oct 29 13:40:01 1991
To: davidcol(?) philba
Subject: DR and Win 3.1
Date: Tue, 29 Oct 91 13:40:01 PST

looks like dr-dos works with win 3.1.

>From richf Tue Oct 29 12:21:15 1991
To: bradc bradsi
Subject: DR and Win 3.1
Date: Tue Oct 29 13:20:30 PDT 1991

I loaded DR 6 on my own system about a week ago and have been using it since. Although I was unable to get Windows to run in enhanced mode using DR's EMM386.SYS, I do run it using HIDOS.SYS, their equivalent to himem. So while I have no UMB support, it doesn't matter much since I use all Win apps. I was also able to run standard mode using HIDOS.SYS.

I haven't had any UAE or crashing problems of note. The only difference, which I haven't tested so I can't confirm, seems to be that printing from Winword 2 is much slower than it was under MS-DOS 5.

In short, I haven't seen any basic kernel incompatibilities.

####################################################### 49
>From bradsi Tue Oct 29 13:40:21 1991
To: bradc richf
Subject: Re: DR and Win 3.1
Date: Tue, 29 Oct 91 13:40:21 PST

which version of win 3.1?

>From richf Tue Oct 29 12:21:15 1991
To: bradc bradsi
Subject: DR and Win 3.1
Date: Tue Oct 29 13:20:30 PDT 1991

I loaded DR 6 on my own system about a week ago and have been using it since. Although I was unable to get Windows to run in enhanced mode using DR's EMM386.SYS, I do run it using HIDOS.SYS, their equivalent to himem. So while I have no UMB support, it doesn't matter much since I use all Win apps. I was also able to run standard mode using HIDOS.SYS.

I haven't had any UAE or crashing problems of note. The only difference, which I haven't tested so I can't confirm, seems to be that printing from Winword 2 is much slower than it was under MS-DOS 5.

In short, I haven't seen any basic kernel incompatibilities.

####################################################### 50
>From bradsi Tue Oct 29 13:45:26 1991
To: steveb
Subject: 6.167
Date: Tue, 29 Oct 91 13:45:26 PST

i played with it for about 45 min this morning. it's quite easy to crash, even hard hang that requires a power cycle. lots of "uae"s. still has a long way to go stability wise. though they have clearly made a lot of progress since the last drop we had. the workplace shell is useble[sic] and win apps do run, mostly. easy to run out of memory doing simple things, and it was on a 9M machine.

one nice thing is that way to config a dos vm. i'll make sure the dos guys have a look.

####################################################### 50
>From bradsi Tue Oct 29 13:46:48 1991
To: bradc mackm(?) mikedr tomle
Subject: os2 2.0 6.167
Date: Tue, 29 Oct 91 13:46:47 PST

it's running in chrissh's perf lab. check it out. the nice thing is the way the[sic] let you config a dos vm, essentially managing the config.sys in a dialog box.

[STAMPED] Depo. Ex. 970
[STAMPED] EXH 26 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS 5065665 CONFIDENTIAL
[STAMPED] MS-PCA 1179419 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT 543 C.A. No. 2:96CV645B

PLAINTIFF'S EXHIBIT 1077 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 1077 Page 1 of 3]

[Plaintiff's Exhibit 1077 Page 2 of 3]

[Plaintiff's Exhibit 1077 Page 3 of 3]


[STAMPED] CONFIDENTIAL
[hand-written] Non-responsive Material Redacted
[STAMPED] PLAINTIFF'S EXHIBIT 563 CA. No. 2:96CV645B
[STAMPED] X0592196
[Page Footer] WinMail 1.21         philba          Tue Feb 23 16:49:36 1993         Page: 160
[STAMPED] MS-PCA 1179462 CONFIDENTIAL

>From chuckst Wed Nov 6 18:38:46 1991
To: philba
Cc: scottq
Subject: Scott's 9/30/31 mail re: DRDOS
Date: Tue Feb 23 17:07:06 PDT 1993

| >From scottq Mon Sep 30 13:57:38 1991
| To: dosdev
| Subject: detect dr dos 6.0
| Date: Mon Sep 30 13:50:35 1991
|
| Does anybody know bow to detect dr dos 6.0? Bambi will not
| run properly on dr dos 6.0 because of a quirk in their device driver
| handling, so we need to detect them.
|
| Scott
|

Prom chuckst Wed Nov 6 18:40:05 1991
To: philba
Cc: scottq
Subject: More Bambi/DR-DOS mail
Date: Tue Feb 23 17:07:11 PDT 1993

| >From chuckst Sun Sep 29 17:IS:46 1991
| To: mikedr philba scottq
| Subject: Bambi on DR-DOS 6.0
| Date: Sun Sep 29 17:16:39 1991
|
| I tracked down a serious incompatibility with DR-DOS 6 -- They don't use
| the "normal" device driver interface for >32M partitions. Instead of
| setting the regular START SECTOR field to Offffh and then using a brand
| new 32-bit field the way MS-DOS has always done, they simply extended

| the start sector field by 16 bits.
|
| This seems like a foolish oversight on their part and will likely result
| in extensive incompatibilities when they try to run with 3rd part device
| drivers.
|
| I've patched a version of Bambi to work with DRD6, and it seems to run
| Win 3.1 without difficulty. This same problem may have caused other
| problems with Win 3.1 and the swapfile under DRD6.
|
| It is possible to make Bambi work, assuming we can come up with a
| reasonably Bafe method for detecting DRD6. The runtime hit would
| be minimal in time and space, although we would have a couple of
| instructions in the main code path for checking the 'special' DRD6
| flag.
|
| What do we think? Should we test further with the patched Bambi to
| see if there are any more incompatibilities????
|


>From cfauckst Wed Nov 6 18:55:00 1391
To: philba scottq
Subject.: Bambi/DRDS mail from 10/16/91
Date: Tue Feb 23 17:07:22 PDT 1993

This is the mail which announced the fixing of the 'bug' which had previously prevented DRD6 form working, As of this point, the DRD6 check could have been removed.

I recall pointing this out, but can't seem to find any email to that effect. I'm sure I pointed it out verbally, and also made a notation in the RAID entry on the 3.31 bug. It is still possible that there was a piece of email which I sent through. XENIX and failed to cc myself on. | >From scrottq Wed Oct 15 19:47:12 1991
| To: bambi georgef sankaz
| Subject: smartdrv 4.0.052
| Date: Wed Oct 16 16:54:23 1991
|
|
| Smartdrv 4.0.052 has passed developer resting and can be found on
| \\triskal\scratch\bamabi\smartdrv.exe
|
| Changes include fixes for dos 3.1 floppy handling, big foot packets
| on dos 3.x, and various and sundry bug fixes.
|
| Scott
|



PLAINTIFF'S EXHIBIT 1130 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 1130 Page 1 of 1]


To: winwar
Subject: 286
Date: Mon Dec 2 22:36:29 1991
?: 20434 512/Comments, etc.
30-Nov-91 10:38:12
Sb: Build 58 is RAM-hungry !
Fm: Tanj Bennett 70642,1211
To: sysop (X)

I fired up build 58 on a 286 and noticed that it eats a lot more RAM than 3.0 used to. 3.0 gave around 1.5M available memory on a 2M machine, while build 58 is down to 1.07M. How representative is this of the final shipping version? Your docs indicate this build is supposed to be a trial run at the real thing.

I`m surprised to see a 400kb hit, roughly 100% growth when you consider DOS accounts for some of the memory. What features of 3.1 caused the growth in RAM usage ?

####################################################### 220
>From bradsi Tue Dec 3 10:25:49 1991
To: donc winwar
Cc: bobgu bobt jont timbr
Subject: Re: Borland and Toolhelp
Date: Tue, 03 Dec 91 10:25:55 PST

if we break their apps when we install, it will serve them right. guess they took the approach of shoot first, explain later.

####################################################### 221
>From philba Tue Dec 3 10:28:29 1991
To: bradsi tomle(?)
Cc: mackm(?)
Subject: Re: slick
Date: Tue, 03 Dec 91 10:25:26 PST

Current plan shows development in Jan/Feb, testing in Feb/Mar and beta test in Mar/Apr so its[sic] got to be available before mid feb for the beta test.

Davebe is looking into the resource issue now. I told him that he can evaluate the schedule and tell me what makes sense. It may make sense to add another body and do some preliminary work for 6.0 to be more efficient.

Diversion or intermediate point? Probably a little of both.

The objective of msdos/windows merge is to combine win 3.1 and msdos 5.0(b?) into a single package with a seamless install. I believe that ("(b?)" is part of actual text.) there will be a fair amount of code that could be reused and certainly the team will learn a lot that will transfer to 6.0. Come on up and I'll be glad to further expound on these topics.

>From tomle Tue Dec 3 08:26:40 1991
To: bradsi philba
Subject: Re: slick
Date: Wed, 11 Dec 91 08:27:46 PST

Eric is evaluating one for me but I don't feel we have our hands on a great one yet. I really need to be clear on what the expectations for slick completion is. Depending on the time frame I am not as confident that slick is a technology on the path to MS-Dos 6 as much as it is a diversion. Phil, can you help me understand what the objectives for your project are?

Tom

>From bradsi Mon Dec 2 20:53:54 1991
To: philba
Subject: slick
Cc: tomle
Date: Mon Dec 2 20:53:40 1991

[STAMPED] EXH 60 DATE 5/17/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS-PCA 1112804 CONFIDENTIAL
[STAMPED] MSC 00730268
[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] DEFENDANT'S EXHIBIT 5614


PLAINTIFF'S EXHIBIT 1133 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 1133 Page 1 of 2]

[STAMPED] EXHIBIT 133
[STAMPED] CONFIDENTIAL

[handwritten] Non-responsive Material Redacted
[PAGE FOOTER]WinMail 1.21          philba      Wed Feb 24 08:11:14 1993               Page: 4

[STAMPED] X0592177
[STAMPED] Gates DEPOSITION EXHIBIT 85 2/28/[20]02
[STAMPED] MS-PCA 1164868 CONFIDENTIAL

>From aaronr Fri Dec 6 10:25:45 1991
To: karlst philba sandeeps
Cc: marcw
Subject: Re: msdos detection - hot job for you
Data: Wed Feb 24 08:ll:30 PDT 1991

Yesterday it was decided to do this detection and display the message if appropriate we run windows. Aaron has some code that you add to win.com

One question that you need to answer for sandeeps:

Is this a Prompt message or a Banner message?

In HIMEM and SMARTDRV the message is a banner message, they don't stop, they just display the message and go on. In SETUP and MSD the message is a prompt message:

Non-fatal error detected: error xxxx
(Please contact Windows 3.l beta aupport)
Press Enter to exit or C to continue ... <<<<<*************

Which form for WIN.COM? Also, what is the exact message test? For MSD the exact message is:

[Plaintiff's Exhibit 1133 Page 2 of 2]

Non-fatal error detected* error 5278 (Please contact Windows 3.1 beta support) Press ENTER to exit or C to continue ...

Sandeeps here is the "spec"

This is a NEAR proc that you call, so inside your code segment you do:

 Extrn        IsMsDos:near

This routine takes no arguments, it modifies flags and ONE BIT of the AX register:

    call      IsMsDos
test ax,2000h
jnx This_Is_MSD0S

NOTES:

Do not try and trace into this routine, you will die.

Try to obscure as much as possible!

The code that makes this call and the code which prints the message based en the result are 'well separated'.

The bit that is set, and the bit that is looked at by the message guy are different bits.

The bit that controls the nessage is stored in a variable that is touched, read written, ALL OVER THE PLACE.



PLAINTIFF'S EXHIBIT 1146 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 1146 Page 1 of 19]

[Plaintiff's Exhibit 1146 Page 2 of 19]

[Plaintiff's Exhibit 1146 Page 3 of 19]

[Plaintiff's Exhibit 1146 Page 4 of 19]

[Plaintiff's Exhibit 1146 Page 5 of 19]

[Plaintiff's Exhibit 1146 Page 6 of 19]

[Plaintiff's Exhibit 1146 Page 7 of 19]

[Plaintiff's Exhibit 1146 Page 8 of 19]

[Plaintiff's Exhibit 1146 Page 9 of 19]

[Plaintiff's Exhibit 1146 Page 10 of 19]

[Plaintiff's Exhibit 1146 Page 11 of 19]

[Plaintiff's Exhibit 1146 Page 12 of 19]

[Plaintiff's Exhibit 1146 Page 13 of 19]

[Plaintiff's Exhibit 1146 Page 14 of 19]

[Plaintiff's Exhibit 1146 Page 15 of 19]

[Plaintiff's Exhibit 1146 Page 16 of 19]

[Plaintiff's Exhibit 1146 Page 17 of 19]

[Plaintiff's Exhibit 1146 Page 18 of 19]

[Plaintiff's Exhibit 1146 Page 19 of 19]


[STAMPED] EXH 77 DATE 5/31/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS 5055650 CONFIDENTIAL

###################################################### 532
>From peterhey Tue Dec 17 16:43:09 1991
To: bradsi
Subject:
Date: Tue, 17 Dec 91 17:41:19 PST
this is the letter referred to in my other mail requesting
OK of advance release of Win 3.1 launch date to OEMs.

| >From peterhey Tue Dec 17 15:05:45 1991
| To: alexn billmi jonro josephk oemman richt
| cc: jonl peterhey wddirect
| Subject: 1st Win 3.1 Launch Letter to OEMs
| Date: Tue, 17 Dec 91 16:03:20 PST
|
|
| Below is the letter created for mailing to our OEM Windows
| licensees--the first in a series to build support for the
| Win 3.1 launch. Thank you for submitting your conments
| asap. We aim to drop later this week.
| ***********************************************************
| \024December 18, 1991
| Dear Windows Hardware Vendor,
|
| Windows 3.1 will be launched at Windows World on April 6.
| Microsoft is pouring unprecented resources into the Windows
| 3.1 launch to make it an unqualified success. The main
| focus of this letter is to help you leverage Microsoft's
| enormous investment.
|
| FACT: We will be investing more promotional,
| marketing, and support dollars in the Win 3.1
| launch than any other product launch in Microsoft
| history.
|
| In many respects, Windows World 1992 vill be the 3.1 Launch
| Event. This Windows release will be the most comprehensive
| and best supported ever. We believe all the conditions are
| right for a massive shift to Windows. A complete foundation
| of Windows Applications is in place, 70% of all PCs sold are
| Windows capable, and Windows 3.1 is the right product.
|
| FACT: The Windows 3.1 beta program is the largest and
| most exhaustive ever in the PC industry.
|
| Feedback from our 15,000 beta testers has been
| overwhelmingly positive. The product's improved fit and
| finish, substantial ease of use, reliability, and
| performance improvements deliver on the Windows promise to
| make using PCs easier for everyone.
|
| FACT: In developing Windows 3.1, over 1,100
| modifications were made to Windows 3.0. These
| changes were focused in four general areas:
| improved usability, greater reliability, enhanced
| applications support and new technologies (such as
| Windows for Pen Computing).
|
| Windows 3.1 represents a dramatic evolution to an already
| highly successful product. We are committed to working
| closely with you to make the Windows 3.1 launch a very
| successful launch for you as well as Microsoft.
| Specifically, we would like to work with you on the
| following:
|
| o Windows 3.1 Compatibility
| o Windows logo Program
| o Pre-launch Technical Training
| o Windows World Participation
| o Windows Upgrade Program
| o Windows hardware design optimization
|
| Below you will find descriptions of our activities in these
| areas and how you can benefit by participating. Details on
| how to reach us via the 3.1 Launch Hotline will be in our
| next letter to you. In the interim, your Microsoft Account
| Representative, or his/her designee, will be contacting you
| soon to help answer any questions you have as well as learn
| more about your plans for this important event.
|
|
| Sincerely,
|
|
| OEM Launch Conmunications
|
|
|
| Windows 3.1 Launch Programs
|
|
| Windows 3.1 Compatibility
| We strongly encourage you to test all of your PC lines for
| Windows compatibility. Clearly communicating compatibility
| in marketing communications and collaterals will enhance the
| appeal of your systems to purchasers of Windows
| applications. Microsoft developed the Hardware
| Compatibility Test (HCT) to measure and certify a PC's
| ability to compatibly run Windows. With only minimal manual
| intervention, an OEM can execute the HCT on all models
| suitable to run Windows and submit the results (which are
| automatically compiled onto a floppy) to Microsoft for
| analysis.
|
| FACT: In the first nine months of 1991, sales of
| Windows applications by the top vendors totalled
| $711 million, an increase of 85% over the full-
| year 1990 sales total of $385 million.
|
| If the results indicate compatibility, we will provide to
| you the new Microsoft Windows logo to use freely in your
| marketing and manufacturing efforts.
|
| FACT: The name of each PC model that successfully
| passes the test will be entered on the Hardware
| Compatibility List (HCL) distributed with each
| copy of Windows 3.1.
|
| The HCT is part of the beta OAK and ODK products. It can
| also be obtained by contacting your Microsoft Account
| Representative. A application to license the Windows Logo
| will be sent automatically to all OEMs submitting successful
| HCT results.
|
|
| Windows Logo Program
|
| Microsoft has created a new logo for Windows to provide
| useful information to customers shopping for Windows
| hardware like your own. Customers will look for the logo as
| a statement that the product on which it appears supports or
| in some way explicitly contributes to the Microsoft Windows
| Operating System.
|
| FACT: Sales of Windows applications in 1991 will
| approximate $1 billion.
|
| This logo was prominently displayed at Fall COMDEX '91 and
| will be an integral part of the packaging for Windows
| beginning with the 3.1 release. A black-and-white rendition
| is pictured below:
|
| (graphic of logo inserted here)
|
| We strongly encourage you to license use of this new logo
| (at no charge) and include it on your product packaging and
| in your promotional materials.
|
| FACT: Microsoft is encouraging independent software
| vendors to assimilate the new Windows logo into
| their packaging and promotional materials.
|
| Given the importance of compatibility between personal
| computers and the new 3.1 Windows release, we are
| requesting all hardware system vendors to complete the HCT
| before receiving rights to use the logo.
|
|
| Pre-Launch Technical Training
|
| As vendors of Windows products, we know that the 3.1 upgrade
| will trigger calls to your support lines as well as
| Microsoft's. We want to help you prepare for these calls.
| Accordingly, we are planning Windows 3.1 Product Support
| Training Seminars to be held around the U.S. in the month
| before the launch.
|
| Please note that because of capacity constraints attendance
| at these seminars will be restricted to Product Support
| professionals only. Pre-registration will be requested to
| guarantee a spot. Cities, dates and times are still being
| determined and will be cannunicated to you in our next
| mailing.
|
|
| Windows World Participation
|
| As the largest Windows-focused tradeshow in the country,
| Windows World is a major marketing opportunity for any
| vendor of Windows-related products.
|
| FACT: This year Interface expects between 60,000 and
| 65,000 attendees to the combined Windows World and
| Spring COMDEX shows in Chicago.
|
| FACT: Over 55% of last year's attendees were Corporate
| End Users and Buyers.
|
| Enclosed you should find an exhibitor brochure for the 1992
| edition of the Windows World Show. Interface expects around
| 1,000 exhibitors to sign up for the combined Windows World
| and COMDEX shows, making Chicago the second largest computer
| industry gathering in the U.S.
|
| FACT: By this past November, exhibitor registrations
| for Windows World '92 had already passed the total
| number of exhibitors that appeared at Windows
| World '91.
|
| Our plans to launch. Windows 3.1 at the show only add to an
| already exciting event. We expect Windows World to be a
| forum for the launch of a number of other important hardware
| and software products from a range of vendors. Don't miss
| the chance to be there and demonstrate the strengths of your
| Windows PCs. Contact The Interface Group at (XXX) XXX-XXXX
| Ext 4023 to sign up as an exhibitor.
|
|
| Windows 3.1 Upgrade Offer
|
| We strongly encourage you to address your users' upgrade
| needs. We expect most Windows 3.0 users will upgrade to
| Windows 3.1 because it offers dramatic improvements over 3.0
| in a number of areas, including performance, reliability,
| and usabilility. Upgrading end-users represents a
| substantial revenue opportunity for OEMs, particularly those
| that have bundled Windows in the past and thereby possess a
| large end-user registration database.
|
| As a service to OEMs who wish to make the Windows 3.1
| upgrade available to customers without assuming
| responsibility for fulfillment, we would like to provide
| you, at our expense, upgrade coupons that Microsoft will
| fulfill directly through the mail. These coupons will be
| Business Reply Cards that should fit in most mailings and
| provide your customers the convenience of ordering the new
| release through the mail. Coupons will be delivered to OEMs
| requesting them by March 15. You can request these coupons
| through our Launch Hotline, or through .your Microsoft
| Account Representative.
|
|
| Windows Hardware Engineering Conference
|
| On March 1-3 at the San Francisco Harriot, Microsoft will
| host a conference and mini-exhibition aimed at helping your
| best engineers build better Windows PCs.
|
| Why focus on building Windows PCs?
|
| FACT- The Windows software standard virtualizes the
| hardware interface for application vendors,
| thereby liberating PC vendors from the need to
| conform to a rigid hardware standard--providing
| room for greater differentiation and
| profitability.
|
| The Windows Hardware Engineering Conference (WinHEC) will
| brinq OEMs together vith Microsoft's Windows device driver
| engineers and a wide range of independent hardware vendors
| active in display, audio, networking, storage, processor and
| other technologies. The object; three days of technical
| discussions reviewing the widening range of creative
| alternatives for PC engineering.
|
| WinHEC's general sessions will consider current and upcoming
| evolutions in the Windows device driver interfaces
| (including updates on Windows NT and multimedia), flesh out
| the new technical directions disclosed at the OEM briefing
| and provide detailed insights into the function of the
| Windows execution performance tests now being developed by
| leading industry analysts. Participants in the session on
| upcoming Windows products will be required to sign non-
| disclosure agreements prior to entry.
|
| WinHEC is being timed to allow you to begin to formulate
| Windows product plans prior to WinWorld so that you can
| address customer needs and top-of-mind concerns at that
| forum.
|
| For more information on the Windows Hardware Engineering
| Conference, contact your Microsoft Account Representative.
| After January 5, you may call directly for an invitation and
| conference brochure (7:30a.m. - 5:30p.m. PST, Mon. - Fri.):
|
| XXX-XXX-XXXX in US/Canada.
| XXX_XXX_XXXX other International
| XXX_XXX_XXXX FAX
|
| Summary
|
| Windows is becoming huge--bigger than we ever anticipated.
| We strongly encourage you to leverage Microsoft's revised
| and expanding investment plans by:
|
| o Executing the HCT on all appropriate PCs and
| returning results to Microsoft for analysis
| o Licensing and leveraging the new Windows logo.
| o Preparing for the Windows 3.1 Launch by attending
| pre-launch training and contracting for a booth at Windows World
| o Aggressively exploiting Windows 3.1 upgrade opportunities
| o Attending the Windows Hardware Engineering Conference
| in San Francisco on March 1-3, 1992 to obtain
| ideas for new, creative hardware designs.
|
| Catch the wave.
|
| You'll be hearing from us again in January. Until then,
| have a happy holiday.
|

###################################################### 533
Fran adamt Tue Dec 17 16:43:31 1991
To: richt timbre
Cc: adamt bradsi chriswo davesm deniser gerardz gregg leighj maroc marked
Subject: PSS Text Issue

Standard Mode: Bad Fault in MS-DOS Extender.
Fault: 000D Stack Dump: 0000 0000 0070
Raw fault frame: EC=0000 IP=5DlD CS=0397 FL=3006 SP=000A SS=02F7

if I installed HIMEM.SYS instead of 386MAX CS=037F was the only change this always happend after windows copied WIN386.PS2 I tryes this about 12 times I had NO autoexec.bat my config only had what STACKER needed
dsk# #fil SIZE   Build 61b
1 39 1,195,812
2 46 1,202,566
3 129 1,182,728
4 72 1,195,448
5 43 1,204,718
6 68 1,173,026
7 80 1,192,092


###################################################### 223
>From andyhi Mon Dec 16 09:30:31 1991
To: kalak winbeta
Cc: bradsi
Subject: HOT: please sign up
Date: Mon Dec 16 09:27:33 PDT 1991

Please sign up and see if we can get on the first wave of final beta shipments.

Thanks
Andy

>From bradsi Mon Dec 16 08:28:38 1991
To: andyhi
Subject: please sign up

Date: Mon, 16 Dec 91 08:27:59 PST


Ken Ashbaugh
Network Systems and Services
[address]
[phone number] FAX [phone number]

as a beta tester, he's running a panel at winworld on "msdos apps under win 3.0". i've been telling him how much better win 3.1 is than win3, and I'm trying to get him to change the focus of the panel to more win3.1. but he said he was previously closed out of the beta, thanks.

###################################################### 224
>From davidcol Mon Dec 16 09:44:29 1991
To: davidw dennisad tiamcc winwar
Subject: Re: bug 8847
Cc: rickem
Date: Man Dec 16 09:43:11 1991

Sorry for the back seat driving here dennis, but has this been thought through very well. For example, 1 assume there's some hack in the font mapper to map courier new 8pt on EGA to Courier 8pt bitmap. What happens if it's not there?

What about other testing implications? What's gonna break without that 8pt font that's been there since June and been through all our testing?

let's get together and chat about this stuff, I haven't seen any email that makes me feel good about the thought we've put into this one way or the other.

###################################################### 225
>From georgem Mon Dec 16 09:45:13 1991
To: bradsi


This sounds familiar doesn't it? If you think we already know about this, I won't follow up.
15-Dec-91 16-.21:05
Sb: Build 61b problems
Em: Mark Aronson 71167,2470
To: Andy Thomas (V3.l Sysop) 73650,50

I upgrading fran build 58 (that worked great) to build 61b I would always get:
'Win Setup caused a general Protection fault in module setup.exe at 0015:096b'
then windows would force me to close at the DOS prompt I saw

Standard Mode: Bad Fault in MS-DOS Extender.
Fault: 000D Stack Dump: 0000 0000 0070
Raw fault frame: EC=0000 IP=5DlD CS=0397 FL=3006 SP=000A SS=02F7

if I installed HIMEM.SYS instead of 386MAX CS=037F was the only change this always happend after windows copied WIN386.PS2 I tryes this about 12 times I had NO autoexec.bat my config only had what STACKER needed
dsk# #fil SIZE   Build 61b
1 39 1,195,812
2 46 1,202,566
3 129 1,182,728
4 72 1,195,448
5 43 1,204,718
6 68 1,173,026
7 80 1,192,092

###################################################### 228
>From andyhi Mon Dec 16 09:47:43 1991
To: bradsi davidcol
Cc: a-andyt a-stevef
Subject; winbtb update
Date: Man Dec 16 09:45:28 PDT 1991

CIS has found and fixed the problem.

They will also be able to credit the accounts of people that got charged. They think they'll be able to have this done by mid-week.

I'm still working on a solution so people can use CIM and TAPCIS to automate the download process.

Andy

###################################################### 229
Fron bradc Mon Dec 16 09:48:04 1991
To: lizsi martyta w-carrin w-clairl w-pamed
Cc: bradc bradsi
Subject: MS-DOS 5 On-going PR plan
Date: Mon Dec 16 09:48:01 PDT 1991

I just read this thank you for getting it to me - lots of good ideas but we are not there yet. I have the following comments:

Situation Analysis
In the DR DOS section i think you are close but do not have the right spin. The editorial comnunity perceives DR DOS 6 as having lots of momentum and as the underdog and they go out of their way to give dr the benefit of the doubt, this happens with even the best mags like pc mag. for example, pc mag two weeks ago mentioned in the comment part of their best seller list that dr was moving up -even though they had not even reached the top 15. this month they mention again that dr is number 14 when one might argue that the more interesting item with at ProGomm Plus moved up to #3! the implications is that we need to target the dr lovers as much as the ms-dos 5 lovers as you suggest later on

you also need to point out that most Of the press incorrectly perceives dr dos to be technically superior

note that though inertia is slowing down Upgrade sales it is still one of the top 5 best sellers in the industry and has been one or two almost every month since ship

Key Messages
I would phrase these differently 1) MS-DOS is the standard - Only MS-DOS is DOS
2 MS-DOS defines the compatibility standard - we care alot about compatibility and are religous about it b/c our users are
3) MS-DOS 5 is technically superior
4) We are driving the MS-DOS standard forward - portable computing initiative and later ms-dos 6

another message is that dr dos 6 is incompatible and buggy, as i said in an earliex mail i bet it ranks with 123 for windows and probably pc tools as the buggiest products to be released this year.

Objectives
Please do not say we are spreading DR DOS FUD. that implies that the data we provide people who ask, like the press is tmtcue - it isn't. the top two pr objectives are to 1) Ensure the press gets the true story on our superiority and dr's inferiority - we have the better product; 2) derail the dr dos train - as discussed in the situation analysis the press is generally very kind to dr and the press is helping to create momentum for the product even though it is lousy.

communicating our momentum is a good objective as is cammunicating our technical leadership and how we are moving the standard forward aggressively.

Tactics

use third parties
i like the idea alot but i'm not sure about implementation, building ms-dos 5 user/company profiles is good, who are those third parties you want to solicit? i need more details.

ibm could still go with dr so i want to hold the oem release for now. it is much more powerful if ibm announces anything with dr. it might be interesting to consider a release that indicates that the top x oems (100, 200?) are now shipping ms-dos 5 with their systems, but if our only coverage is page 108 of pc week (like the last release) then there is not enough value in releasing now.

influence coverage
aggressively following reviews is great, but is not enough - i have learned this mistake and will not make it again, we need to proactively provide information to key people in the press as an ongoing part of our business.

we need to track the people positive towards dr just as much as those who are negative, we have a legitmate goal to provide these people our side of the story.

the backgrounder on what people should look for when evaluating an os is a good idea but don't you think it is too late for ms-dos/dr dos. who will read it now? you need one for windows

dr dos comparison doc if you mean feature comparison, NO! this misses the point, we do not want to get into a feature war or legitimize dr's efforts to say the two products are equal with dr having more features, we need to take a step above and win the battle on compatibility, technical superiority and our vision.

ms-dos resource kit is done, see randym

share data with folks that shows we are better - good

editor buddy program - excellent how do educate the "buddies" on ms-dos?

leverage upcoming news - much more than ms-dos 5 ran or apm; as i have discussed this is the portable computing iniative - a commitment, a position of leadership.

ms-dos terminology reminder - how will we implement so that we don't just legitimize dr? ya know when people say pc-compatible they really mean ms-dos compatible

i'll let bradsi make the call on using gordon letwin more. i'm not sure.

leverage ms-dos marketing
i like on-going direct mail/postcards etc. ms-dos tech workshops are done for now. we need to discuss the upcoming marketing efforts and see where pr can help. whatever happen to the local columnists list?

we should sit down and discuss - pis set-up. i'm sure that you will have additional ideas with my revised objectives. what is the status of getting more resources on ms-dos?

###################################################### 230
>From stephl Mon Dec 16 09:48:35 1991
To: bradsi
Cc: nataliey
Subject: Yogen Dalai
Date: Mon, Dec 16, 1991 9:46 AM

Yogen will be checking into the Woodmark Hotel (in Kirkland) this evening. You can meet him there or leave him a message and be will ccme to you. Thanks,
Stephanie


###################################################### 231
>From greglo Mon Dec 16 09:53:41 1991
To: bradsi davidcol tomle
Cc: mackm
Subject: Re: Novell
Date: Mon, 16 Dec 91 09:52:50 PST

1. the purpose is to map Fail into a reasonable error code that will be correctly handled by windows apps

2. it would ship only with Win31 retail; no one could distribute separately (although we'll probably make it available earlier to a few key accounts such as American Airlines)

3. testing will be done by our test group as well as Novell and a few corporate accounts such as American Airlines.

4. maintenance could potentially be done by our group (AaronR; he could do the initial work except that he is booked solid). I don't know if MSDOS6 will make such changes as to require major changes (like lots of new functions) which could require some help from your group.

| >From tomle Sat Dec 14 10:58:32 1991
| To: bradsi davidcol greglo
| Cc: mackm
| Subject: Re: Novell
| Date: Sat, 14 Dec 91 10:58:03 PST
|
|
| I am still confused about the solution. How do we use this library.
| How will this be used to circumvent Novell's panic on a return from
| retry,fail. Who uses the library. Is this something your sending in
| the SDK for use with Windows apps? Once we get past undestanding how this
| solution works then I need to know vho is going to test this library
| and who will support it in the future? Do I have to update it every time
| I rev the Dos?
|
| I am not against helping out here, this is a serious problem, I just
| want to understand what I am getting myself into.
|
| Tom
|
| | Here is a summary:
| |
| | This measure is to address the critical-error problem that American
| | Airlines is up in arms about. Today, users think they've hung their
| | machines when a server goes down. Novell wants us to crash individual
| | apps instead, but we can do better.
| | It would take about 3-4 days for one of the MS-DOS developers to take
| | the code out of the MS-DOS Kernel and build it into an independent
| | library routine. We could turn that into an installable driver which
| | would allow apps to continue after encountering such an error.
| |
| | We would then ship this driver with Win31: we need not give this code
| | to Novell/DR nor give them permission to redistribute it.
| |
| | FYI: Here are the gory technical details of the problem:
| |
| | A typical scenario would be where WinWord is saving a file out to a
| | NetWare server when the server goes down or the net cable gets
| | pulled, etc. You get a Window critical-error dialog giving you the
| | choices of Retry or Fail. Retry will always just give the same error
| | again, so you can't escape that way. Fail will also bring up the
| | same error, but if you keep hitting it again and again long enough,
| | eventually you should get back to the application and be able to save
| | your work. Most of the time, however, the user will give up and
| | reboot before then. Also, end users won't have any way of knowing
| | which seemingly endless chain will end and which will not.
| |
| | The problem is a result of two "design deficiencies", one for Novell
| | and one for Windows. What NetWare is trying to do is have MS-DOS abort
| | the application: bango, no chance to save your work. That is what they
| | do for non-Windows apps. Windows won't let them, because Kernel
| | can't survive having MS-DOS terminate an app behind its back.
| |
| | NetWare generates an int24 (critical-error) with Retry and Abort being
| | the only available choices: it never expects that to return. Windows
| | won't allow Abort, we offer the user Retry and Fail. When the user
| | chooses Fail we return to NetWare. They say "Whoa! Someone actually
| | returned, this is not kosher!" but they try to handle it as best they
| | can without, you know, really getting involved. They return an error
| | code of -1 to the original calling application. That sounds reasonable,
| | but it turns out that, since -1 is not a valid error return from most
| | MS-DOS functions, apps aren't checking for it or handling it reasonably.
| | For apps like WinWord, they will just go on writing out more and more
| | file, generating more and more errors which they ignore. It can go on
| | a long time.
| |
| | Novell's idea of the correct solution is for us to modify Kernel so that
| | they can Abort a Windows app. This would take two weeks of design and
| | coding before it could be testable, followed by goodness knows how much
| | debugging and fixing. These are potentially very destabilizing changes.
| | It is far too late to make these changes now.
| |
| | Our idea of the correct solution is for NetWare to handle these
| | critical errors the same way MS-DOS does: when the user chooses the
| | Fail option, MS-DOS figures out a reasonable error code to return to
| | the app, based upon the actual internal error and the MS-DOS function
| | being called by the application. This is very complicated mapping,
| | involving eight tables and lots of code in the MS-DOS kernel. It
| | would take Novell about three months to reverse engineer this and
| | implement it themselves.
| |
| | But again, it would only take about 3-4 days for one of the MS-DOS
| | developers to take the code out of the MS-DOS Kernel and build it
| | into an independent library routine. We could turn that into an
| | installable driver and ship it with Win31. We need not give this
| | code to Novell/DR nor give them permission to redistribute it.
|
|



###################################################### 232
Fran bradc Mon Dec 16 10:07:48 1991
To: bradsi
Subject: FW: Stacker Update
Date: Mon Dec 16 10:08:06 PDT 1991

any comments?

>From bradc Thu Dec 12 17:57:30 1991
To: bradsi mackm richf
Cc: bradc
Subject: Stacker Update

Date: Thu Dec 12 17:57:47 PDT 1991



Subject: EMAIL auto_notify

Re: RE: FW: Windows 3.10.060 Problems & Updates
Received OK on Mon Dec 16 11:43

User Message follows
Thank you for submitting your bug report to the Windows Beta Program.

Because of the large number of beta sites, and the complexity of the program itself, we will be unable to respond to each of your reports, though we review each one, and will be contacting you in the event that we need more information to narrow down the bug for our development staff.

If you experience critical problems, such as difficulty reading or writing to your hard drive, please notify us as soon as possible, so that we may respond immediately.

In your reports, please make sure to use the Systems EFORM template called "Windows 3.1 Bug Report" and include a copy of your AUTOEXEC.BAT, CONFIG.SYS, SYSTEM.INI, and WIN.INI (when necessary) files, in addition to the steps that must be followed to reproduce the problem.

Your participation in the Beta program is important to us, and we appreciate your efforts in helping us make Windows 3.1 an excellent product.

###################################################### 271
Fran lauraeu Mon Dec 16 11:48:02 1991
To: bradsi jonl sallyn
Subject: MSJ Jan Ed note
Cc: ericm joannes
Date: Mon Dec 16 14:47;21 1991

Please read and OK ASAP. It needs to go out tonight. Thanks.

Windows is stronger than ever. Near the end of 1991, the installed base of Windows had grown to 7.9 million copies. Windows is now the world's most popular graphical computer environment, with about half (46 percent) of those users outside the United States. The forecast for 1992 is even brighter; sales of 9.2 million copies are predicted. Preinstalled or bundled copies of Windows are new standard with many brands of PCs, including AST, Dell, Everex, Grid and Zenith.

Windows not only sells extremely well, it gets high marks from purchasers. In a survey conducted by Field Research Corporation, randomly selected users gave Windows an average rating of 7.6 on a scale from 1 to 10. The survey also found that users like Windows better the longer they work with it. Users aren't skimping on the hardware they use with Windows. 70 percent are running on at least a 386SX, and 80 percent work with a display of VGA resolution or better. More than three quarters of the users ran Windows on machines equipped with between 2MB and 4MB of RAM.

Microsoft is going to great lengths to ensure that Windows version 3.1 will enhance Windows acceptance still further. A beta test program with over 15,000 sites is underway to guarantee that 3.1 is very compatible with 3.0 vhile delivering significant new benefits.

The most important improvement of 3.1 is speed: faster printing, faster application startup, faster screen updates, faster disk I/O, and a faster MS-DOS box. Comparing Windows 3.1 beta 1.55 and OS/2 2.0 beta 6.167, testers at Microsoft found that Windows applications load twice as fast under 3.1 than do PM apps load under OS/2 2.0. Perhaps more importantly, they determined that Windows applications running native under Windows 3.1 loaded 30 to 50 percent faster and painted their displays 20 to 60 percent faster than they did running in OS/2 2.0's "Windows box." So much for cockamamie slogans like Ra better Windows than Windows.S

The retail release of Windows 3.1 is being augmented in a number of ways. It will support the Multimedia Windows API, and include MIDI and waveform drivers as well as several sound applets. The TrueType APIs will be improved to support font rotation. Drivers for printing high quality Truetype documents, along with a supplementary font library, will be bundled with the retail product. Robustness is also improved with 3.1. Systems developers in Redmond have declared war on UAEs caused by Windows and by misbehaving applications as well. A strongly typed WINDOWS.H, parameter validation for every function call to Windows, and improved error reporting all help in exorcising these demons from your code.

As developers ready their applications for Windows 3.1, their most important job is to test for compatibility. The next issue of MSJ will include a comprehensive checklist for ensuring that your application will run better than ever under Windows 3.1.

###################################################### 272
Frcm torahe Mon Dec 16 11:50:37 1991
To: bobt theresas
Cc: bradsi drg tomhe
Subject: Windows User Survey
Date: Mon Dec 16 11:47:26 pdt 1991

BradSi mentioned something about the Windows User Survey info we have possibly being available for distribution (maybe in summary form?), so Symantec is now excited and wants it ASAP. They also want the customer support tools SteveB mentioned in his speech (not sure what this is exactly).

Any help would be appreciated.

###################################################### 273
Fran andyhi Mon Dec 16 11:57:20 1991
To: marcw winwar
Cc: a-richh korys vlads
Subject: RE: Upgrading OEM display drivers bug
Date: Mon Dec 16 11:54:58 PDT 1991

I think that this is going to generate a lot of calls to the tech team. A good number of internal and external users have OEM displays.

But I haven't heard many complaints yet, is this a fairly new bug? What happens if we upgrade over a prev. version of 3.1?

>From marcw Mon Dec 16 11:15:38 1991
To: winvrar
Cc: a-richh korys vlads
Subject: Upgrading OEM display drivers bug

Date: Mon, 16 Dec 91 11:08:49 PST

After upgrading an OEM display that uses the 3.0 internal VGA VDD (most OEM VGA displays do), the user will get an error whenever they try to run and non-windows application, telling them to run Setup again.

The problem is that whenever we upgrade an OEM device over 3.0 Windows, we do not do the correct translation. For displays, we do not change the WIN386 VDD from *VDDVGA to VDDVGA30.386. The change is very straightforward: if we are upgrading an OEM device, we always put the file installation through our existing translation code.

Work around: specify OTHER for display and use the OEM setup disks to specifically install the display.

I am sitting on the fence on this one. The change will ONLY AFFECT OEM DEVICE UPGRADES. Other upgrades or new installs are not affected, so this seems pretty safe. However, there is a simple work around and this change has not been thoroughly tested on all the various OEM device drivers (it affects not just OEM displays, but all OEM devices).

###################################################### 274
>From jeffpar Mon Dec 16 12:04:12 1991
To: bens mikem raype
Cc: bradsi macKm tomle
Subject: Re: cougar command.com
Date: Mon Dec 16 12:03:41 1991

Negative, it is not true.

| >From raype Mon Dec 16 09:05:03 1991
| To: bens jeffpar mikem
| Cc: bradsi mackm tomle
| Subject: Re: cougar command.oom


| installation, I restarted windows, i no longer have net access from
| the file manager, no net menus under Disk, apparently my winnet
| driver is no longer loader, it was fine in build 61b. I exited
| windows and rebooted the machine, same problem -- no net access from
| file manager.
|
| running maintenance mode install shows that it detected my
| net correctly : lan man 2.1 basic.


###################################################### 285
Fran nathanm Mon Dec 16 12:51:50 1991
To: billg bottnu bradsi cameronm carls darrylr davec davidcol davidw dennisad edwardj gaben jimall jonl karenh mikeraur paulma paulo philba raleighr rashid robg steveb stevesh tonyw w-pamed
Subject: Winstone suggestions
Date: Tue Dec 17 12:50:30 PDT 1991

JonL had a meeting a month or so ago on the "new world" we face with OEMs. One idea which came out of that meeting is the idea of having a benchmark suite which we called "Winstones". There hasn't been a lot of general discussion since then, so I thought I would send some ideas on the topic, both to people in the original meeting and to others that may be effected.

I think that Winstones are an INCREDIBLY important marketing move for us, which can benefit just about every aspect of our systems strategy.

The basic idea is simple:

- Create a benchmark suite analagous to SPECmarks, but specialized to Windows based systems (both Win 32 and Win 16, and on x86 and MIPS).

- The suite will include measurments of a set of different activities - screen graphics, printing, text, disk I/O, CPU etc.

- We would create the suite, get it out to magazines and others, and generally publicize the hell out of it.

The general motivation to do this is to provide a focal point for activity to improve the hardware that Windows runs on (accelerator boards, , and at the same time give a firm quantitative basis for many of our present challenges - improving value for our customers, selling Windows vs OS/2a selling Windows NT, promoting Jumbo, promoting TrueType, showing the value of Win32, showing the value of scalability, demonstrating the value of MIPS... There are a LOT of potential benefits.

Note that this is a MARKETING activity. There are a number of technical aspects, but this is first and formost a technical marketing activity.

The Winstone suite would contain a number of different test suites. This is not a small joke benchmark like Dhrystone - it is more like SPEC, but probably even bigger.

There are several requirements placed on Winstones by marketing factors:

- There must be between 5 and 15 different tests. This way you can make a nice graph of system performance.

- We want to have one focussed test in each area that is going to be important for somebody to improve. As an example, if we want to encourage graphics accelerators, then there must be a separate graphics test.

- There should be an overall number - the "WINmark" which is a harmonic or geometric average of the separate tests. We should also define the particular subsets - "Graphics WINmark", "I/O WINmark" etc.

- We must be able to give the source code of the test away. This is probably not "public domain" in a strict sense (see below) but close to it.

- We want some of the tests to measure the entire machine configuration specific. As an example, some tests will run faster if you have more RAM, so that you can do more cacheing. Although that may not seem fair, it actually is JUST what we want. This is discussed more below. - The tests must be able to run automatically and then return an answer without human intervention.

There would be two categories - system level benchmarks and application level benchmarks. The system benchmarks would primarily exercise Windows and the underlying hardware. The application benchmarks would measure what kind of performance people oould expect in their apps. Note that SPECmarks are purely an application benchmark in this terminology.

The system benchmarks would be created largely by using the "artificial app" technology that has been in use in the NT group. Basically this lets you run a real app, trap all of its calls and arguments, and then be able to replay them in a script. This is a terrific way to get "real" tests easily. We can append the scripts from several different runs of an app or different apps to make a single test for each sub category.

Here is a sample list of the system benchmarks:

  • Screen graphics & Text
    • Fonts and rich text
    • General GUI
    • Presentation
    • Draw programs
    • Paint programs
    • CAD programs
  • Printing
    • Rich Text
    • Graphics
    • Bitmaps
  • Disk I/O
    • General file read/write
    • Database access
    • Multitasking disk access
  • Virtual memory
    • Memory allocation/freeing
    • Large memory access
  • Multitasking
    • Running many of these tests in parallel
    • Background communications at 9600 baud
  • Multiprocessor/thread test
    • CPU bound process with threads suitable for MP
    • CPU and I/O " " " " " " "
  • Messaging
    • Windows message passing
    • OLE performance???
  • Multimedia
    • CD ROM input speed ?
    • Animation test?
    • Sound card performance?

In most cases the "artificial app* will give us a very good benchmark pretty easily. To give an example in more detail, here is one way to break down screen graphics and text.

Screen graphics & Text

  • Rich text & fonts - combined script from:
    • Word for Windows
    • WordPerfect far Windows
    • Aldus Pagemaker
  • General GUI - combined script of dialog/menus from many Window apps
  • Draw programs - combined script from
    • Corel Draw
    • Micxografix
    • Aldus Freehand
  • Presentation - combined script from
    • PowerPoint
    • Persuasion
    • Freelance
  • CAD programs - combined script from
    • AutoCAD
    • AutoSHADE
    • Alias Upfront
    • other' windows CAD program
  • Paint programs - combined script from
    • Aldus PhotoStyler
    • other serious 24 bit paint program

There are thus 6 separate tests done within the screen graphics & test section. Each one should have a running time of between 2-5 minutes so we can get good accuracy, and so that we are future proofing ourselves for a factor of 4X or so speed improvement in the next several years.

Note that we would want to use both ISV apps AND Microsoft apps. I do NOT think that we need to spend a lot of time or effort actually creating the benchmarks "with" the ISVs in a serious way. We don't want this bogged down with politics, and there is no reason to do so. We probably should get permission from them and should sanity check the data file that we use. If this is done properly the ISVs should love this.

Some of the tests would need to be created by hand, or substantial modifications must be done to the recorded scripts. In the case of virtual memory performance we would want to allocate a ton of memory (say 16 meg) and then touch it to test paging performance. This may be better done with a synthetic program than a recorded script. The multitasking test and Windows messaging tests are other examples which may have to be written largely by hand.

This is almost certainly true of the multiprocessor/threading test. This will have to be some quasi-real example of a parallel algorithm with good scalability out to at least 16 processors. It must be able to run on a uniprocessor. There are many examples we colud use for this.

The goal is that almost all tests must run on Win 3.1 and on Windows NT (in BOTH x86 and MIPS) and on Win32s. This will not be true of a couple of the tests, but in general it must be the case.

Note that we will have to review each of the benchmarks to make sure that it says something reasonable. The existing state of benchmarks in the PC industry is so poor that nearly anything we do will be better than what exists today. Nevertheless we should try to do as good a job as we can.

The application benchmarks are similar in spirit to the SPEC benchmarks, or the larger set of programs which MIPS uses for their benchmarks. The goal is to get sane real programs which truly exercise the cache, memory system, CPU etc.

The ideal thing would be code samples from real products. The obvious problem with this is that we really need to distribute source code to the tests. Another idea would be to use the SPEC set. Unfortunately the SPEC benchmarks run on UNIX systems, and they have a number of problems. SPEC is the best set of benchmarks available, but even so there are a couple of bad programs (particularly matrix3OO) and the set is too oriented toward FORTRAN and numerical stuff.

The best solution from a practical standpoint is to get some public domain code (which may need to be ported to our OS) to create the benchmark. There should be a set of 5-10 different integer programs and again as many floating point programs. We should make sure that the programs are quite different in their composition. There is a place called the Austin Code Works which sells tons of PD software, and that is a good place to look first.
------------------------------------
The goal in creating the benchmark suite is to provide a cannon standard of reference for the Windows computing community to use. The idea of doing benchmarks in each of these areas is not exactly new - PC magazine and other reviewers regularly run ad hoc benchmarks. The problem is that these are of uneven quality, and no single one of them has enough of a following that people use it as a common reference point.

The "brand identity" of the benchmark is just as important as its contents for this particular purpose. We want people to have a universal metric which can be used in advertising, product reviews etc. We want to get this to be a very common way to describe a system. The SPECmark rating has become a very sucessful effort in the UNIX workstation world because it is famous enough for everybody to quote. We want a similar phenonema to occur in the Windows world. The concept of how many "Winstones per dollar" and other direct comparisons will directly follow.

Microsoft should be the direct sponser of Winstones, and we should support that with an active promotional campaign. It may also help to have some publications champion them in their reviews, but we would like this to be "neutral" enough that they become quoted in ALL major reviews, and not just in those of a single magazine. We probably want to announce the benchmark with one or a couple of magazines, but then have them spread to other areas as well.

-------------------------------------
There are many uses of Winstones in our marketing strategy, each of which may have a special implication for our system strategy. In general, winstones can be used in the following ways:

Supporting our product line strategy
We will have a number of different Windows implemenations - Win 3.1, Win32s on Win 3.1, Win NT on x86, Win NT on MIPS. Which one should people buy? Winstones should help explain this. The reason is that many of the tests are designed to test advanced features of wirxiaws, and to scale with the available memory on the machine.

We would like to be able to make a chart like the following.

System             | Win 3.1/16     Win32s   Win NT
---------------------------------------------------
386SX/20, 2 meg |
386DX/33, 4 meg |
486/25, 8 meg |
486/33, 16 meg |
R4000, 8 meg |
...

Ideally speaking, this would allow us to directly DERIVE our system strategy from "objective" empirical results. Of course, I would not leave this to chance - we would adjust the benchmarks to make sure this is the case. This should not really require any cheating - the fundamental truth is that if we have reasonable benchmarks, we should be able to demonstrate precisely this effect. Win NT might be slower at graphics than Win 3.1 on a small display because the Win 3.1 code will be tuned assembler, but NT should make better use of large memory (8 meg and above), multitasking and it can do some of the advanced tests (multithreading etc). Win32s will be slower that Win 3.1/16 on the system benchmarks because of the overhead of the thunk layer, but it should show an improvement in the application benchmarks which will exercise large memory in 32 bit mode.

Note that one powerful reason for us to take a leadership position in creating the Winstone benchmarks is that we want to have a BALANCED and COMPLETE set of benchmarks out there otherwise the opposite will happen - people will TRASH our systems strategy. If you just measure a single niche, such as graphics performance, then you could possibly find that our higher end systems do not do as well. The default way that people approach benchmarking is rather naieve and this will HURT us. If this occurs, then we will be fighting a rearguard action in a defensive posture. Coming out FIRST with a benchmark which does support our systems strategy avoids all of this.

Note also that this issue becomes critical to the MIPS platform. Its primary reason for existence is performance, and the Winstone figures for it will be essential for making a case for MIPS. The central marketing message for the companies producing MIPS based Windows machines is that they can deliver more absolute Winstones than any other platform, and hopefully more Winstones per dollar too.

Selling against OS/2
The claim that they will be a "better Windows than Windows" is put directly to the test by comparing the Winstone rating for a given machine configuration. I believe that this can be a very powerful tool for embarrasing them. This can occur in two ways.

First, many of the scripts would be fully Win 3.1 exploitive. It is entirely possible that the suite will not run under OS/2.0. Even if they support the correct feature set, the tests which allocate large amounts of memory will not operate in real mode, and will be at a strong disadvantage in standard mode, so any attempt to run Windows in a restricted mode will be exposed.

Second, even if they can run the scripts, it will be hard to look good. General performance problems will crop up in many ways. Having OS/2 in memory will consume several megabytes, even if there are no other problems.

Windows accelerators
Dozens of companies are creating add on graphics boards to accelerate Windows. They have little technical guidance, and little way of comparing their efforts. Winstones are ideal because manufacturers can directly advertise how their boards effect the "Graphics WINmark", "Disk I/O WINmark" etc. The existence of a standard metric for this will will help to focus their activities. Competition will increase, and end users will have a better way to judge what they're getting. To date, most of the "accelerator" performance story has centered on display cards, but we also would like to encourage low cost RAID disks, better system caches, better printing solutions and a variety of other hardware improvements.

PC price/performance
The emphasis on "Winstones per dollar" highlights the price performance aspects of a system, and it is an excellent way to demonstrate the value delivered to the customer. This does not directly help Microsoft, but it does help get the Windows computing world focussed on deliering the maximum bang for the buck to our customers, so it is useful in that context.

Selling Jumbo
The printing benchmark focusses attention on the best Windows printing solution. This is an excellent opporunity to highlight Jumbo and the advantages it brings.

Selling TrueType
The font and text benchmarks are a good opportunity to highlight the benefits that TrueType brings to Windows customers. Alternative approaches would look pretty silly if the decrease the over all Winstone rating of a system.

Taking the technical high ground
Finally, this is an opportunity for Microsoft to take a strong leadership position in nurturing the community of suppliers to the world of Windows computing. Establishing a standard benchmark of this sort is a good move - it helps both customers and IHVs develop better products.

----------------------------------------------------------
My take on this is that it is a very important opportunity which we should captialize on ASAP. Comments are welcolm.

Nathan

###################################################### 286
>From georgem Mon Dec 16 12:51:55 1991
To: bradsi
Subject: Re: bullet
Cc: dennisad, elik

I talked to Chuck Bigelow about this so that I could get the story from the horse's mouth, so to speak.

His answer was that bullets really aren't standardized by various typographers, and so it is really left up to the "random tastes of the artist". They had always hated the large bullets that seem to go with every other font on the planet. He said it seemed to them that those bullets were almost too big and that they seemed to jump out of the page when you were trying to read it, thereby distracting you. They brought too much attention to themselves, in other words.

Since Lucida Bright and Sans were designed as text faces to be used in books, they wanted a more discreet looking bullet that would blend in better. Since they already had a zillion bullets in the Lucida Stars font, and they knew you could always use one of them if you were unhappy, they did the smaller bullet.

For presentations and such, it would probably be better to use something from Lucida Stars. If you were writing a book, on the other hand, he likes the smaller bullets since they aren't as garish.


###################################################### 287
Fran greglo Mon Dec 16 12:54:00 1991
To: bradsi davidcol tomle
Cc: mackm
Subject: Re: Novell
Date: Mon, 16 Dec 91 12:52:58 PST

It is an installable driver which is loaded at boot time by the a system.ini entry. It hooks Kernel's MS-DOS function handler and critical error and munges the registers on the critical error based on the current dos call. (It could also be a simple DLL loaded by load= rather them an installable driver; doesn't really matter.)

As for Mack's suggestion I don't think it is reasonable for us to provide the workaround and not make it available to customers. If we relegate it to a PSS fix we'd still be sending it out: it's just that more users will crash and not report it and not realize that a solution exists. It would make no sense to make Novell reimplement it-over again, either. The testing is an additional burden but we are already required to do fairly massive post-beta3 distribution of


what is the msdos 5 experience been like?

###################################################### 345
>From jnetter Mon Dec 16 15:42:40 1991
To: lorisi winsquad
Cc: debbieh
Subject: RE: Windows Focus Squad Meeting Reminder
Date: Mon Dec 16 16:41:52 POT 1991

RussW is in Zurich and will not be available to attend...

>From lorisi Mon Dec 16 14:36:58 1991
To: wingguad
Cc: debbieh
Subject: Windows Focus Squad Meeting Reminder

Date: Mon, 16 Dec 91 14:34:45 PST

Next meeting is scheduled for tomorrow, 12/17 at 8:30 am in 2/2103.



###################################################### 346
>From joachimk Mon Dec 16 15:43:46 1991
To: billg bradsi steveb
Subject: EMI
Date: Mon Dec 16 16:20:06 PDT 1991

Mike signed a 20 M$ commit license today for all their 386 sx and higher end systems/per pocessor for WIN and DOS. This was a tough one against DRI. THey will continue to ship DRI on 286 and 8086 systems honoring an old license agreement. DR was at $3-4 with DOS, ve are getting more than $35 for WIN and DOS. The slim pack DOS helped to close the deal.

###################################################### 347
>From debbieh Mon Dec 16 15:53:16 1991
To: jnetter lorisi winsquad
Subject: RE: Windows Focus Squad Meeting Reminder
Date-. Mon Dec 16 17:48:47 PDT 1991

steveb will not be attending as well but the meeting should still take place ..... thanks

###################################################### 348
Fran carls Mon Dec 16 16:00:05 1991
To: billg bradsi paulma tcmle
Subject: ULSI
Date: Mon, 16 Dec 91 15:58:22 PST

fyi...

| >From intelca!mipos3!ccm!Mike_Bruck@intelhf.hf Mon Dec 16 15:27:25 1991
| Return-Path: <mipos3 !ccm!mike_bruck@intelhf.hf="">
| Received: by mipos3 (5.57/10.0i); Mon, 16 Dec 91 13:53:32 PST
| Received: by intelhf .hf.intel.oom (A=~A Smail3.1.17.5 #17.19); Mon, 16 Dec 91 13:50 PST
| Received: by oatt.hf.intel.com (ccmgate) Mon, 16 Dec 91 13:50:10 PST
| Date: Mon, 16 Dec 91 13:50:10 PST
| >From: Mike Bruck <mipos3 !mike="" bruck@com.hf="">
| Message-Id: <911216135010_2@ccm.hf.intel.com>
| To: Robert_Sullivan[S]_at_ccssw fm@ccm.hf, Cindy_L_Thomas@ccm.hf,
| mipos3!intelca!microsoft!carls
| Subject: ULSI
| TO: ESM, PR CONTACTS
| THE FOLLOWING RELEASE WAS ISSUED OVER BUSINESS WIRE AT
| 12:00 NOON (PACIFIC STANDARD TIME.).
| Contact: Pam Pollace Intel Corp.
| (XXX) XXX-XXXX
| OREGON JUDGE GRANTS INTEL'S REQUEST
| FOR PRELIMINARY INJUNCTION ON KEY MATH COPROCESSOR PATENT
| Key Licensing Issue Also Decided in Intel's Favor
| SANTA CLARA, Calif., December 16, 1991 - Intel Corp. said it has



Thanks. Alessandro


###################################################### 365
Fran bradc Mon Dec 16 16:54:21 1991
To: bradsi
Subject: RE: EMI
Date: Mon Dec 16 16:54:1l PDT 1991

yup, was there when they signed today, sergio helped vith this a bit too. big win | From joachimk Mon Dec 16 15:43:46 1991
| To: billg bradsi steveb
| Subject: EMI
| Date: Mon Dec 16 16:20:06 PDT 1991
|
| Mike signed a 20 M$ commit license today for all their 386 sx and higher end
| systems/per pocessor for WIN and DOS. This was a tough one against DRI. THey
| will continue to ship DRI on 286 and 8086 systems honoring an old license
| agreement. DR was at $3-4 with DOS, ve are getting more than $35 for WIN and
| DOS. The slim pack DOS helped to close the deal.


###################################################### 366
Frcm gaben Mon Dec 16 16:55:03 1991
To: davidcol
Cc: a-kelm davidtry noladr richsa stevesh valorieo vinwar
Subject: Re: WPG Dropping the ball on the PCT
Date: Mon, 16 Dec 91 16:53:44 PST

Ok. We're just going to keep working with DavidTry on the PCT testing (see his mail below on Valorie's buglist) . If there is anything else we need to do, let me know.

Thanks. | >From davidool Mon Dec 16 16:12:28 1991
| To: gaben
| Subject: Re: WPG Dropping the ball on the PCT
| Cc: a-kelm davidtry Doladr richsa stevesh valorieo winwar
| Date: Mon Dec 16 16:10:59 1991
|
| Clearly something is screwed up, Windows folks
| don't usually flame other groups like this
| without good reason. However, we are deep in the middle
| of getting our final beta out and thus won't be able to
| resolve until after. Fran richsa's original mail, it
| looks like ve can deal with the gap for final beta.
|

| >From davidtxy Mon Dec 16 14:29:42 1991
| To: valorieo | Cc: a-kelm gaben noladr richsa
| Subject: Results of Valorie's testing
| Date: Mon Dec 16 14:29:16 PDT 1991
|
| Here's the most recent info I've received fran Valorie. This
| is exactly the sort of data I was looking for. My comments are
| prefaced with ">>>"


| CSD is installed on the machine when you install IBM PCLP NET on it.
| I believe that the PCLP NET install docs state which DOS 4.00 CSD is
| required for it. I believe that different versions of the PCLP NET
| require different DOS 4.00 CSD versions. I also believe that it is
| very difficult to tell which DOS 4.00 CSD is installed on a particular
| machine, you simply have to KNOW which one is installed. You might be
| able to tell by looking at the date on the system files (IBMBI0.COM/
| IBMD0S.COM), but I am not at all sure about this.
|
| If you are running an IFSFUNC.EXE that does not match the IBM DOS 4.00
| CSD that you are running, you will get no warnings, the PCLP NET will
| be perfectly happy to start. Windows Enhanced Mode will NOT be happy
| to run correctly on such a machine however! All sorts of very very
| peculiar stuff will start happening. The one case I know of is that
| wierd stuff, breakage, hangs, reboots, bugs will start occuring.
| Either in windows, or in the DOS applications, or in both. I do
| not know the specific cases of all possible combonations of PCLP
| and DOS 4.00 CDSs.
|
| I have been struggling with a machine in the network lab downstairs
| 1041 (NETLAB 22, IBM PS/2 model 55sx) for three days!! This problem
| is what was wrong with it. The IFSFUNC.EXE with PCLP NET that was
| installed on this machine did not match the IBM DOS 4.00 CSD that was
| installed on this machine.
|
| I am inclined to say the following things:
|
| All PCLP NET test machines are possibly broken.
|
| The validity of all PCLP testing that we have done is at risk.
|
| I know several factual things:
|
| All PCLP test machines need to be checked.
|
| All NET testers need to be educated how to not INVALIDATE
| both the testing and the machines.
|
| All development folks who may be involved need to be educated.
| I for one, am completly unwilling to look at anything having to
| do with PCLP until it has been explained, al least to my satisfaction,
| what the scope of the problem is. And when, or if, it is going to be
| fixed.
|
| We have not been careful. We are now going to pay the price for not being
| careful. It remains to be seen what that price is..........


###################################################### 192
>From spanky!operdst
Sun Dec 15 17:02:41 1991
Received: by darla.UUCP (DBCUS UUCP w/Smail); Sun, 15 Dec 91 17:00;39 PDT
Date: Sun, 15 Dec 91 17:00:39 PDT
From: <operdst @darla.uucp="">
To: microsoft!bradc sergiop adamt richf dianet bradsi johncon richba
Subject: DOS 5 mtd Shipments...


PLAINTIFF'S EXHIBIT 1243 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 1243 Page 1 of 16

To: Joachim Kempin. Distribution
From: Jeff Lum
Date: Februrary 25, 1992
Re: January Europe OEM Sales Status Report

Attached is the European OEM Sales status report and US OEM sales status reports for January. Please provide me with any feedback that you deem appropriate. Copies of more detailed reports from each subsidiary or account managers are available upon request.

Distribution:

Reif(Ralf?) Skoglund(?) MSAB
Bengt Akerlind(?) MSAB
Peter Slum(?) MSAG
Petslek(?) Da(?) Smedt MSBV
Hans Rensal?(?) MSBV
Jochen Heinik(?) MSGMBH
Juergen Hu?h MSGMBH
Egon Salmutter(?) MSGESMBH
David Svendson(?) MSLTO
Sandy Duncan MSLTO
Mohol(?) L?oomb?(?) MSSARL
Peacel(?) Martin MSSARL
Umberto(?) P?olucci(?) MS SPA
Mauri?o(?) Bodino(?) MSSPA
Ign???(?) F?u(?) MSSRL
Sunir(?) K?poor(?) MS Europe
David Britton MS Europe
Robbie(?) Beck MS Europe
Hans(?) Apel MS Europe
Bernard(?) Vergnes MS Europe
Tim B??rd(?) 10N/1261(?)
Peter Bramen(?) 10N/1314
Brad(?) Chass(?) 3/2046(?)
Mark Chestnut(?) 10N/14??(?)
Dougl?s(?) J?ckson ?S/1077
Arne(?) Josefeberg(?) BP/8181(?)
Dwight Krouse(?) 1/1080
Paul Maritz(?) 2/2045
Nell(?) Milter(?) 8M/2258(?)
Lorl(?) Morrison(?) 10N/124?(?)
Sergle(?) Fineds(?) ?/205?(?)
Frenz Reu(?) ?5/2122
Brer?????????rg(Smudged on purpose?) '???4??(Smudged on purpose?)
Carl(?) Stertz(?) 2/2100(?)
Adam(?) Tayl?r(?) 3/2052(?)
John Williams 6/2061

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] **MICROSOFT SECRET**
[STAMPED] MS7035050 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT 1343[?] [illegible]
[STAMPED] MS-PCA 1164632 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 2 of 16]

January Status Report


Europe OEM Sales

Jeff Lum. Director


Revenue (see attachment for details)
Area January
Actual
January
Budget
FYQ3
Forecast
FYQ3
Budget
Ovr % of
Budget
Europe OEMs 1,126,791 0 17,500,000 16,692.125 105%
US OEMs 7.067,782 6,776,780 20,000,000 19,184,844 104%
Europe Sales 8,194,573 6,776,780 37,500,000 35,889,043 104%

Comments on revenue: Europe OEM: Over a million dollars in pure miscellaneous revenue throughout the subs on a budget of $0 for the first month will give us a nice boost for the quarter. I expect to finish close to $1M over budget for the quarter as we do some catch up billings as well as invoice several due-on-signings for license renewals. US OEM: Breman Team Ravenue for January was $2.7 million against a plan of S2.6 million 104% of plan. This indudes all the new accounts now assigned to Peter's group. Tandon came In at $1.9 million against-a plan of $1 millon which more than offset the $700,00 in lost revenue, bacause of tha Tandon/Positive merger. Compaq had a decent quarter in FYQ2 shipping ovar 2S0K units, and we should be at or near plan for February revenue. Chestnut Team Revenue for January was $4.3 million vs. plan of $4.3 million. NCR had a surprisingty strong second quarter, and also reported some previously unreportad net revenues, which resulted in $1.9 million in January revenues. The revenue outlook for Q3 is looking much bettar - $11.4 million vs. budgat of $10.S million. The addition of Logitech to the group is a major reason for this as they are expected to finish $1 million over budget in 03. AST shipped a record 120K systems in O2 and is also expected to finish well above budget for Q3.

New Business Signed
Account Product Comments
European OEMs    
Facit AB DOS 5.0; Win 3.0 New M/C = $2S0,OO0/year
Detevsrehuset A.S DOS 5.0; Win 3.0; OS/2 1.X New M/C = $562,000/$562,000
Prill Computer GmbH DOS 5.0; Win3.0 New M/C = $116,000/year.
BSC Computer GmbH DOS 5.0; Win3.0 Amendment M/C = $125,400/$169,800
Sender Computer Systeme GmbH DOS 5.0; Win 3.0 Amendment M/C = $113,200/year
C & S Comp & Service Vertriebe GmbH DOS 5.0; Win 3.0 New M/C = $97,500/year
J & S Speth Software Systeme GmbH DOS 5.0; Win 3.0 New M/C = $64,800/year
Geosoft Brender & Kreft GmbH DOS 5.0; Win 3.0 New M/C = $43,500/year
Pyramid Computre GmbH DOS 5.0; Win 3.0; OS/2 1.X Amendment M/C = $54,000/year
Lehmann und Partner DOS 5.0 Amendment M/C = $240.000/year
Rain Electronik GmbH DOS ROM; WIN ROM; Flash New M/C = $22.000/year.
Ti'Ke Computer Corporation Ltd. DOS 5.0; Win 1.0; PPack New M/C = $383,500/year
Psion PLC DOS 1.11; Flash 1.0 Amendment M/C = $87,500/year
Matavidooler S.A. DOS 5.0: DOS ROM 5.O New M/C = $67,500/year
Unidata SRL DOS 5.0; Win 3.0: MMW New M/C = $182,000/$200,000
Microsy Electronics SRL DOS 5.0 New M/C = $720,000/year
US OEMs    
NCR OEM Premier Suppport $15,000/year
HP Corvallis MS-DOS 5/Windows 3.X Per processor for two additional years
HP Corvallis Excel; Word: Mail: Schedule+ LOI for Lion Project

** MICROSOFT SECRET **
2

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035051 CONFIDENTIAL
[STAMPED]MS-PCA 1164633 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 3 of 16]

General Issues

AB ICL willl start working with Novell. The market demand for Novell is there, and with the general reorientation from a pure technical company to a more market oriented, ICL appears to have taken a decision to complement LAN Managar with Novell. ICL is very upset bacause we did not carry out what was promised at the LM Open Council in September regarding one upgrade price to 2.1. Instead they have at least 5 different RR depending on the customer situation. This in contrast to Retail side. This is really a general problem that the Retail offering and OEM offering generally don't overlap on work group products, creating situations of discrimination against tha OEM.
IBM is very active selling OS/2 2.0. Victor remains cool, but ICL is leaning over to IBM. Tha major issue is that Micrsoft is not really interested in supporting OS/2 (as Steve Ballmer made very clear). GmbH The situation with Schneider is still unclear. We proposed a solution where Schneider has to write off half of their $3.4m PPB. Schneider may step out totaly of the PC-business. For IPC we couldn't find a solution up to now. because they felt blackmailed by our proposal, to terminate the contract within the new agreement. ObS is threatening us by several proposals from their side were they want to deduct a high amount from their invoices because of losses the mada caused by not getting the delivery partner contract. Escom and Peacock started a negotiation with us to combine their purchasing. They even told us that it is possible that both companies merge. Robotron Ascots and Robotron Soemtron are liquidated. Open Invoices win be paid.
LTD News that we are cutting a deal with Phoenix to allow them to sell ROM DOS with their BIOS and motherboard design is BAD ! We've been working hard in the UK to establish a ROM DOS 'distribution' channel. Contracts will be signed in early Feb. The development of this opportunity is a delicate one that has already taken up a lot of our time and effort. The distributor needs a dear run at tha market - the Phoenix deal will provide totally unnecessary competition. Apricot have asked to re-negotiate their agreement. They're over committed by at least 50%. Wa've tentatively agreed to reduce their second year's min. commit to $1.3M from S2.2M in return for adding a third year to the agreement.
The FG DOS business is an area for growing concern. In December we sent out 26 licenses and only saw 7 of these returned. In January we sent out 38 new agreements and only saw four of these returned !!!! In addition five of December's agreements appeared in January bringing our total for the two months to less than 30%. It's becoming s critical issue as we need to look for new business to replace existing pack DOS OEMs as we sign the top guys up to royalty. We desperately need two FG-DOS account managers.
We are also suffering from an explosion of pirate DOS. Products bearing the Minta and AcBEL labels are foremost but at the and of Jan. we saw tha worst yet, with our US EZ DOS appearing on the open market st less than £25 per copy. Our strategy is ideally to put this to bed at source (typically these 'pirate' products are rogue products produced under valid MS agreements but shipped without Hardware). If we can't do this then we have been attempting to get to the UK distributor and simply sell him our genuine package product at price list. As I write we have received an order for 1,000 units from one of these "distributors" in Stonehaven, Scotland.
SPA the "old" OEMs are twisting our arms to lower their numbers to match those of new deals. Trying to reach good tradeoffs, and not to make enemies out of old friends. Andreab is pushing CDC to cover all their systems, eliminating the risk of a DR revamping.

Compaq marketing group continues to grapple with tactical issues on Windows opportunity. Compaq concerned with number of corporate agreements MS has closed for Windows business. IBU division closed to further discussion of EBU and Mouse products. Compaq requests per-system pricing for DOS license.

** MICROSOFT SECRET **
2

[Plaintiff's Exhibit 1243 Page 4 of 16]

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035052 CONFIDENTIAL
[STAMPED]MS-PCA 1164634 CONFIDENTIAL

Ungermann-Bass: UB is concerned about our presenting a clear strategy for LM to the marketplace as we approach NT implementation and move away from OS/2. This strategy needs to be put into a format that the UB sales force can sell into their large accounts. We continue to release deals to the retail channel without considering tha OEM channel. UB would like to be able to compete on a level ground with the resellers (i.e., if we give the channel a free or reduced royalty for upgrades we should offer the same deal to our OEM's).
Positive wants to bundle applications on their hard drives. This would enable the user to evaluate the applications on their own system in their own home for a fixed number of use . They would than have to call Positive with their credit card number to purchasa the key code which would unlock the applications so that they could run. Mike Maples does not buy into this. Phillips: We are attempting to license the rights to PhotoCD from Philips to be usad on all Windows based Products.
ZDS - has informed us that our best price for Ballpoint is over $11 higher than Trackman. ZDS is very likely to choose Trackman ovar Ballpoint.
ZDS - informs us that Spinnaker is offering lower pricing, free documantation and no foreign uplift. We are uncompative from a price standpoint.
AST - AST's Tammy Watanaba is recommending IBM's JDOS and even DR's JDOS ovar MS's product due mostly to MS Japan's expressed attituda towards AST. AST is angry bacause thay believe MS Japan is continuously quoting misleading delivery dates and AST is watching MS Japan 'court Compaq' while MS Japan basicaly ignored AST's afforts to enter the Japan market.

General News

AB Tandy will open a superstore in Copenhagen, selling Victor, IBM, and Compaq. It is the familiar Computer City concept from the US. Observe that it is Tandy that is opening, not Victor Preparations are in full swing. Victor will continue operations as before. Both Computar City and Victor will source from Victor's newly opened factory in Scotland. It is expected that certain changes in the management structure will occur, but this is not confirmed publicly. BV At a seminar organised by Intel IBM was giving a presentation on OS/2 2.0. Tha product looked very stable. Thay used tha latest build, and the presenters were very confidant that IBM was going to make the target date of end march. At the presentation they positionad OS/2 as the DOS task switcher. Windows compatibility was shown. IBM states that thay have full DDE between OS/2 apps and Windows apps that was not shown. The system performence (a P75) was very acceptable in DOS, Windows and OS/2. At the demo they explicitly showed also DR-DOS in ths OS/2 DOS box. The story was that users are using Windows mainly as task switchar and OS/2 can do this much better. On top of this you get full Windows and OS/2 that is indispensible for IBM mainframe communication. The Apple/IBM deal will not generate products within five years(!) And in the mean time OS/2 is the best solution.
GmbH started a Win3.1 program together with 52 OEM's for CeBit. Almost all of them agreed to show Windows 3.1 on their machines at CeBit. The program is very well accepted. We started to work, on this together with tha local systems business unit. All OEM's were informed about our plans for the Windows 3.1 launch and all OEMs answered a questionnaire about thair plans for Windows 3.1.
SPA Good MYR, no change to OEM dept Retail (in lira) at 69% MTD, 86% YTD. SW market worsened in January. Clone market healthy. Super deal closed with Microsys, see Other OEMs section.
SRL 2 finalists for the LA Rep position are ready for final interview with you. Other highlights of the month were the MYR and Systems Strategy Briefing to 3 OEMs and 3 LAs.

** MICROSOFT SECRET **
4

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035053 CONFIDENTIAL
[STAMPED]MS-PCA 1164635 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 5 of 16]

Account News

Compaq announces sales of $873 million for O4CY91 showing improvement over Q2 and Q3. Net income was $131 million. Compaq announces agreement to resell Novell Netware. Compaq ands SGI joint venture agreement.
Sofboume has fired their president and laid off 30% of their employees within this last month. Philips: Rob Hamersma, the Managing Director of Philips Consumer Electronics, has bean replaced by Lars Nyberg. Mr. Nyberg comes from Philips Information Systems. CompUSA: Artisoft shipping free software, diske, and documentation. Introducing new Slim line machines in March.
Creative Labs: Will introduce a 16 bit sound card in June.
Media Vision: Will introduce new portable sound board that plugs into printer port in March. Headland Technology: Getting out of the sound board business.
NCR - The WPD reorganization is complete and Clemson now handles all softwara managament. Representatives from all of the key NCR organizations will be in Redmond on Febuary 10th to learn about MS OS/2 to NT Migration plan so they can assist their ISVs and customers in the transition.
NCR - Received good news from 3140 group via Stevab conversation with Tom Mays that the 3140 will not be Hobbit (RISC) but will be most likely SL based architacture.
ZDS - ZDS will offer 7 out of the 49 bids for DT4. During the first round of bidding, ZDS bid MS on all Windows bids. However ZDS bid Windows on fewer than half of their 7 bids. We are working to convince ZDS to bid MS and Windows on all bids.
ZDS - announces further layoffs and a restructuring of field sales. Over 60 sales and support people are cut.
ZDS - will reportedly be supplying IBM with a large quantity of portable machines for IBM resale via Bull. Rumor has it that ZDS will supply 150k machine per year to IBM. If this it true, it will double the number of portables shipped by ZDS.
AST - Genelle Trader, Director of AST notebook products, resigned and accepted a job as Everex's VP of Marketing.
AST - MS made significant progress with AST on DTIV. After our meeting with AST's Bob Becker, AST is expected to ship MS products only and abide by the Option 3 terms.
HP - HP Vancouver's DeskJet 500C announced last Fill at $1095 SRP, is getting good reviews from the PC press. Nearly every review notes the product's fundamental flaw of being modal (either good monochrome or good color, but not both at the same time) while endorsing this product as the best low cost color option available today.
DG - DG announced that they started fiscal year 1992 in the black, but just barely. DG earned $4M on sales of $294.8M in Q1 of fiscal year '92. This represents the fifth consecutive profitable quarter for DG after sustaining two and one-half years of losses.
DG - DG announced that Robert Mars has assumed the position of VP of the PC Business Unit. Mars comes to DG from Tandon where he was VP of Sales. Mara fils the position vacated by David Ruberg, who left in August.
** MICROSOFT SECRET **
5

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035054 CONFIDENTIAL
[STAMPED]MS-PCA 1164636 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 6 of 16]

Area Summaries


Finished Goods DOS - Europe

Subsidiary January
Actual
January
Budget
FYQ3
Forecast
FYQ3
Budget
MS AB 8,497 1,982 13,792 6.792
MS AQ 2,038 1,744 8,000 5,323
MS BV 4,033 2,100 8,873 7,200
MS GESMBH 429 984 1,300 3,002
MS GMBH 44,589 29,343 102,688 86,477
MS LTD 16,110 8,500 44,075 35,500
MS NV 1,074 1,842 4,993 5,629
MS SARL. 8,473 10,800 69,141 60,100
MS SPA 4,285 2,200 9,274 5,900
MS SRL 2,525 3,345 3,500 10,600
TOTAL 92,033 62,444 263,836 217,523

January was the strongest month for PackDOS in the history of MSAB. The figures speak for themselves. In reality January was a super month because of shipment problems in December. This has triggered several PackDOS accounts to consider licensing.
BV The pessimistic outlook we had for this month did not come through. Despite the fact that Sydec signed a licensed agreement they still placed an order for FG M5-DOS. So this month BV is on 180% of budget. The Belgium sale continues below budget. However the application sales in Belgium is at record levels. Clearly this indicates that the dedicated FG-DOS sales person is
urgently required.
GmbH Revenues are far over budget. With $5.9M we are at 153% of budget. The forecast for January was 41,800 units of package MS-DOS Incl. DOS/Windows combo packages. Actually we achieved 44,100 units which is 155% of our budget with 28.400 units. According our February forecast we expect to ship in total 29,000 units. Mice business went well. too. We delivered 9.000 units as forecasted. D+S raised their forecast for the rest of the FY'92 to 12,000 units each month.
LTD ANOTHER RECORD MONTH !!!!! Tht subsidiary beat $10M for the first time ever and our FG DOS sales contributed by beating the 16,000 barrier for the first time. We almost doubled our revised budget of 8,500 units. This is even better news when you take into account the 'loss' of Ti'Ko to a royalty contract. They were ordering at least 1,000 units per month previously. Our windows penetration has continued above the 50% mark for the third month in a row. However, this is slightly worrying as we try to move these customers to full royalty licenses due to the high cost of producing DOS and Windows manuals. Dearly EZ DOS isn't enough without EZ Windows. Meantime we've persuaded Phoenix to do OEM manuals at under $35 which will help but it isn't perfect. This combo needs to cost the OEM less than $20 to be effective. The news that we will be bring the EZ packages to Europe was very encouraging but I have serious reservations regarding Ireland's ability to deliver small quantities in good time to multiple delivery points. We should do a local deal for the UK/Europe. I will be investigating the possibility of using Amstrads printer who are based in England and provide VERY competitive pricing for Amstrad. The outlook for Pack DOS sales is extremely positive. Feb. looks quieter than January but we'll still easily exceed our revised budget Of 11,000 units. Msrch, which is year end for the majority of UK companies, looks like being a MASSIVE month.
SPA Sales of PackDOS keep great. at 194% YTD. In units, 21.726, we have beaten this FY budget. Revised budget is 31,000. Doing more active prospecting, and planned a new edition of the packDOS ad we did in November.

** MICROSOFT SECRET **
6

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035055 CONFIDENTIAL
[STAMPED]MS-PCA 1164637 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 7 of 16]

SRL Pack DOS sales were 76% vs. MTD budget (98% vs. YTO budget). CSEI and Comelta were the first customers converted to royalty business and several others like ADL (5K units) and Sieesa (10K units) may go straight to royalty business next month. Expectation is we will be way over royalty business budget ($8OOK vs. $150K budgeted] and somewhat short on Pack DOS budget (35K vs. 40K units).

Market Trends

MS AB (Bengt Akerlind)
Interest for Win 3.1 is gearing up. It is common knowledge when it will be released. We will ride this wave and convert as many PackDOS accounts as possible from "naked MS-DOS" over to MS- ODS/Win. Rumor has it that IBM will launch OS/2 2.0 on March 25. a few days after our local Win3.1 introduction.
MS GMBH (Jurgen Huels) MurtiMedia is taking off in Germany. Escom and Vobis are advertising MM machines on their front pages of the *Extrablatt" and "Denkzeell",
We see a high demand for hard disks which is currently not satisfied by any hard disk vendor. This is especially slowing down the notebook business.
MS LTD (Sandy Duncan)
Prices of new machines is getting sillier and sillier. Latest record broker of note is from Opus. 386SX/20. 40MB HD. 1MB memory, color monitor. VGA and DR-DOS at a list price of £599. Viglen and Elonex have pledged to follow! This compares with the PC World (superstore) price of £1,599 for a similarly configured Dell machine and £999 for an Amstrsd with the same spec. Lotus have appointed a full-time "Director of Copyright Protection" to defeat piracy. Lotus believe that 50% of all software in US is illegally copied.
IBM UK announced that they lost £124M in FY '91. This compares with a profit of £420M in FY '90. Total UK employees is down by 2,639 to 14,909 and turnover declined by 14%. Total UK IT market in 1991 was estimeted at £27 Billion, up 5.1% on 1990. Projections for 1992 predict similar growth to £28.5 Billion. Largest sector is Finance, followed by manufacturing. Construction was the smallest sector with Telecom showing fastest growth.
Novell is reported to be signing Merisel-Softsel as a disty on a pan-European basis. Could be the start of a significant trend?
The only superstore in the UK "PC World* hasn't had a lot of press recently but they are reported to be running a 'mass demo' of OS/2 2.0 on Feb. 15. They either believe that there's a significant home/hobbyist market for OS/2 or that corporates make purchasing decisions on Saturdays ! - Very strange.
February looks like being s VERY busy month. Highlight will be the Windows Show in London. OEM people are heavily involved with a total of 12 man days. We've helped create a separate technology stand with OEMs including Victor, Zenith , Compaq. NCR and Apricot showcasing their various windows based technologies including NT. 3.1, Pen and MultiMedia.

Key Account Summaries

European OEMs

Actebis Computer GmbH
Oh January 16th, we had a meeting and with Harr Urban and Herr Puhrsch to discuss the final details for an amendment to their contract. Agreed upon was a two year amendment to their contract for 1OOK units of MS-DOS per system at $16.00 with Shell (and this would be for the German version and the $2.00 uplift for all other versions), and $25.00 for 50K Windows (also for German version). We are also working with them on getting approval for their packaging of MS- DOS and Windows.

** MICROSOFT SECRET **
7

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035056 CONFIDENTIAL
[STAMPED]MS-PCA 1164638 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 8 of 16]

Amstrad (Richard Barrie)
Amstrad have been reasonably quiet after thrreatening to erupt last month ! Major incident took place in the Amstrad car park. Richard Barrie was last seen travelling out of Brentwood on the M2S at 150mph after bashing Sugar's Rolls Royece in the Amstrad car park. Apparently left his name, address and Digital Research's phone number on the windscreen.

Apricot (Dale Borland)
Dale and I met with Mike Jordan , who is rssponsible for the contract. Put simply - they're massively over-committed. From April their min. commits go up to 650K per quarter and they're currantly only fortcasting half of this. We have agreed tentatively with them to extend the contract by one more year at a commit of £1.3M rather than one year at £2.2M. Currentiy we're waiting for Apricot to confirm this is OK. Meantime Peter Horne will be meeting with Paul Maritz and Pam Goldschroldt when he visits Redmond on Feb. 18. Dale will attend. Horne has some notion that Multi Processing is fundammentally flawed - hopefully Paul can convince him otherwise.

Brother
They have started talking about putting DOS in their typewriters. This is the pot of gold that we've always been looking out for ! - they build 10.000 tvpewriters every week 111 - We have a LONG way to go but this could be very lucrative business for us.

ESCOM (Michael John)
The meetings with our marketing BU'S have been successfully finished. ESCOM will advertise Works for Windows stand alone on one page of his April magazine issue. They still want Microsoft to take over a part of the advertisement cstn for that issue. but we won't.
ESCOM finally agreed on the application deal. They will hard bundle Excel 3.0 and Word for Windows 2.0. They get a discount of 60% and DM 150.000.00 for advertisement which they will use for their magazine issue. ESCOM will preinstall all application demo's and test software from Microsoft on every PC.

G2 (Hans Ranselaar)
We finally agreed on the new amendment for G2. Peter Breedijk will sign it on short notice. This amendment will be effective Oct.. 1 .1992 and will include Works for Windows. Where is the Dutch version of Work for Windows anyway? Susan Boeschen told G2 that they could expect this early December 1931.

ICL/Nokia LTD (Lars Ahlgren/Dale Borland)
ICL came in with a very strong Q4 report. The Nordic figures showed over 27,000 units shipped, and from the UK we hear that they shipped over 20.000. This indicates that ICL is getting close to a 50,000 units quarter. License negotioation is closed. We sidestepped Norokorpi and went straight to Mills who gave a verbal OK on (DOS + Windows) $32 PP/PI/3Yrs + $3 for localization. This was a good win for us.

Microsys
Excellent deal closed with Microsys, in a rush to cut off DR from negotiations. A nice royalty for DOS, and a commitment of $720/year, and most of all a bad wound for DR. This gives us almost free way to try and attack Olidata. the last big customer of DR in Italy.

Olivetti (Maurizio Bedinal)
General. Olivetti slowly starting up back again, people is still quite de motivated, in spite of a very good quarterly shipments result: 221.000 systems. Too bad, this only means they pushed a lot down in the channel in December, that implies, Q3 will be bad. IBM approached Olivetti with an offer similar to Bull's, which Olivetti politely turned down.

** MICROSOFT SECRET **
8

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035057 CONFIDENTIAL
[STAMPED]MS-PCA 1164639 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 9 of 16]

LM2. The news they got re our mktg. plans for LM CTP ara no big deals, and -worse- we know there is practically no plans. Trying to push Mktg. to endorse this program & push it. if this won't happen. CTP will be a flop and LM will be even more at risk in Olivetti. They are seriously talking to Novell.

Comms server for NT. Lots of activity in this araa (this is one of the hot banking topics, together with NT). They met us, then DCL, conf called us again, and will meet DCL again. Basically we are trying to sell them the 1.1 for OS/2 as a product now, and its evolution to NT as a perspective Quoted std royalty rates (ind. commitments) and $100k for the source down from $2SOk (as thay will do significant enhancaments that in turn they'll license to us for free). We may be able to close this deal, prob. at 50% now. Bad news is Dondolini accepted a job for Olivetti Italy and will be gone in one month.

MultiMedia. Olivetti is whining for tha $15 royalty with no commit, which they consider high, and still messing up with MediaVision.

NT-OS/2. Thsa issue is becoming hotter and hotter. On the banking front, Brandi, Mensi!, and developers just had a mtg. with Davidwo, Perttir etc. In Redmond about the issue of porting Olivetti offer to NT. According to a preliminary report by Lorim & Maurz (who was thare too) the mtg. went well and bad. Well because anyway Olivetti is convincad that it a worth the affort to consider NT the target platform for the PB system, and therefore a cooparation will start with us to ensure tools and apps are ported smoothly. Bad bacause Olivetti made clear that thay don't consider Win 3.1 a viable intermediate step betwean OS/2 1.3 and NT as a replacement of OS/2 2.0. Thay confirmad instaad that if a customar asks/needs OS/2 2.0 they will make this possible, not only shipping OS/2 2.0 per se, but also investing to 'dress' OS/2 with suitable apps. This meant thay will spend energy in porting (or ready up to port) their apps, and even LM if we don't do it. This obviously implies Olivetti widening up their contacts with IBM. On a different front during an interesting mtg. with Piol (that Umbertoo called up), he requested officially that MS accepts to sponsor tha Olivetti R4000 machine launch in Paris in April. This sponsorship is letting them show NT to the audience (only Olivetti top VARis wwide), having a MS executive give a speech, and extending the NT for ISV program to Europe, pushing Olivetti MIPS machines. Waiting for answers from the US.

OPUS (David. Bradley) Bradley and I met with Adam Harris in late January. Basically a great meeting. We're sure that we can get the DR DOS issue resolved and a Windows deal signed before mid-March. (Their contract expires end March).
One of thee issues that has arisen from this is an issue which effects all of our channels at MS Ltd. Opus' biggest customer (British Rail) want to put Windows on all of their new and existing PC's. The new ones are fairly straightforward, but the existing base of c. 8,000 windows capable machines is more of an issua as our large accounts guys as well as one of our dealers are bidding for this business. The situation has been resolved - we will make a proposal to allow Opus to upgrade BR using White Box windows. We'll also give the same proposal to the dealer involved (MSL) and they can make their pitch against Opus on a fair and equal basis. MS retail, large accounts. Opus, the dealer and ultimately the customer are all happy with this.

Schneider Rundfunkwerks AG (Andreas Niegel) We proposed a solution to Schneider Rundfunkwerks AG to write off 50 % of their prepaids. Mr. Rusniok will present this proposal to the board of directors. We assume that a decision will be taken until February 15th. In this new amendment they want to commit to 50k units a year for the period from Jan. 1 to Dec. 31, 1992. Given that we take standard pricing on this deal it will be a $2.65m commitment. We assume that we have to take pan of their prepaids into that contract.

** MICROSOFT SECRET **
9

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035058 CONFIDENTIAL
[STAMPED]MS-PCA 1164640 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 10 of 16]

Siemens Nixdorf lnformationsystems AG (SNI) (Jaap VanArkel)
Meeting with Decision Makers in Redmond date Feb. 20th. In preparation. Objective: Convinca SNI that NT it on schedule and the complete NT platform can be offered within our promised time frame.
Windows. Working to get the documentation at SNI before their deadline: February 14th. They recieved the documentation files without artwork. SNI wants to co-operate at the WIN 3.1 Launch. Will work with Windows Business Unit to get maximum involvement.
Windows/CeBIT. SNI will show Windows 3.1 at the CeBIT. About 80 notebook (introduced at CeBit) with Windows installed will be on several places at the CeBIT. The Windows Business Unit will get one or more SNI NT systems in our Booths. Their first pre-installed Windows machine will be introduced at the CeBIT. 386sx price 3000OM.
NT: SNI it pushing NT strategy to their Large Accounts. In all presentation slides NT is positioned as a strategic product. SNI has postponed the multi processor development due to technical problems with Intel.
NT-SNA/COM Server - System Management This month Vesa Suomalaimam (MS) and Clive Partridge (DCL) presented SNI Open Systems development department the MS Comm Server status and strategy and possible co-operation between MS and DCL. SNI was not willing to give out requirements and status on their SNA product for UNIX and OS/2. This due to a bad experience this group had with Microsoft in co-operation on these themes more then a year ago. We agreed on a new start. February 21 we will have a meeting in Redmond about co-operation on system management and get SNI committed to SNA server for NT.
LANMAN: SNI received the LAN Manager for UNIX source code this month and started porting. Expected release date shortly after the CeBIT (April 92). The LAN Manager amendment is approved by both LAN Manager for OS/2 and LM for UNIX planning departments at Siemens and is approved. Amendments are now to the SNI management to be signed. Waiting for language for the TCP/IP utilities.
Mouse: SNI is selling the COMFO package (Logitech Mouse (12$) MS Windows and Comfodesk) as a retail product. They signed up for 10,000 units of Windows (finished goods) for the coming year. MultiMedia: No change, the system is in development and will be introduce at the CeBIT

Tulip (Hans Ranselaar)
The BUN group is working on the LM 2.1. Tulip is offering a free upgrade to customers who purchased after 1 dec. 91. The LM 2.1 package will be made up with the Microsoft manuals and the FG OS/2 1.3 package. This package is much cheaper than the license and the manuals. Our technical people are working with Tulip R&D to fix some LM bugs. We have the attention from Redmond now and this is helping. However there Is not much progress in solving the problem, but there is nothing we can do more.
Tulip had an excellent Q2FY92. they sold about 30K units which 80% more then the previous quarter, making this the best ever. Our royalties amount to over $1000K. The processor split for this quarter is 32% 286, 63% 386Sx 5% others. In particular the sales of Window machines are up.
Roland Diets Marketing manager of TCI will leave April 1, he has accepted a position as marketing manager with the Elsevier publishers.
Tulip is further delaying payments of our latest invoice until all amendments are processed. There is a PPB of about $150K, but the royalties in the last quarter are much higher then the MC. This means that after processing all the amendments Tulip will have to pay us an additional $100K on the old contract. So after the two years of the contract, the difference between the MC and the actual royalty is less then 2%.

** MICROSOFT SECRET **
10

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035059 CONFIDENTIAL
[STAMPED]MS-PCA 1164641 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 11 of 16]

Victor (Magnus Larsson) During the spring Victor will launch the new series of Grid Pen Computers. The development is handled by Grid. Tandy will establish the Computer City concept in Europe. Tha first store will be locatad in Glostrup outside Copenhagen. It will ba a 3,000 square meters suparstore containing hardware from IBM, Apple. HP, Compaq and Victor. There are plans to establish three more stores in Sweden and one in Norway during 1992. Info in italic it NOT OFFICIAL information yet, even if there has been a flash about it in a Danish newspaper.

Vobis (Stafanie Reichal) During the Mid-Year Review, Bill Gates visited the Vobis store in the Arabella-center in Munich. Bill has now seen their efforts for DR-DOS, unfortunately, but we are working on a long term plan to dramaticaly improve the presenee of MS-DOS in Vobis stores. On January 28th, we had a meeting with Vobis together with Dahmen to discuss the relationship and to officially introduce Stafania as their contact. Issues that ware discussed were whether they would be interested in licensing Win Works and they said they might be interested in talking to us more about it especially at a price of $15.00 or $30.00 bundled with Windows. He is also interested in discussing possible PCWorks 2.0 to WinWorks updates. We think that once he starts feelind stiffer competition from Escom that he will want to talk more to us about this subject of WinWorks. In this metting we were also informed that thay shippad 103,000 systems in Q2FY92 and we will be getting signed and finalized royalty reports from them in the next meeting on February 6, 1992. 68K units want out with MS-DOS S.0. and about 50% of all units went out with Windows.

US OEMs

Compaq (teresach)
The primary activity this month was Windows 3.1 negotiations. Jeffl and Teresach traveled to Houston to begin negotiations. This was followed up by numerous conference calls and a written proposal, counter-praposal, and response. The current status is that Compaq will license Windows for products due out in the June time frame. While Teresach continues to push for the April time frame. Compaq has been unable to move quickly enough to take advantage of the launch opportunity. Peterbra visited the top three Compaq subs in Europa to discuss joint marketing opportunities between the MS and Compaq subs. Billg and Eckhard Pfaiffer spoke briefly regarding Compaq announcements and Windows licensing status. Additional activities were a Pen Windows briefing, NBU and printer meetings.

Ungermann-Bess, Inc. (markbu)
UB is still dedicated to a LM strategy and is interested in raising the level of the strategic relationship between our two companiess. We met with UB to understand the new organization there and to present the MS strategy to them. It is still unclear on how tha new relationship will eventually be structured. It is clear that it will change over the next months as UB's new rganization settles in.

tomhen
Commodore: Negotiations for the per-system Windows and Works for Windows license is going very slowly. although we did receive a $42.000 PO. for Works for Windows document films. Tomhen made MS DOS 5.0 Portable Computing Initiative Preeentation to their technical staff, which was well received. This is the continuation of our effort to replace DR DOS on their palmtop machine. Tomhen spoke with Lou Egglebrecht (their palmtop architect), he agrees MS DOS belongs on the palmtop.

** MICROSOFT SECRET **
11

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035060 CONFIDENTIAL
[STAMPED]MS-PCA 1164642 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 12 of 16]

Tandon: Tomhen attended their biyearly executive level Product Planning meeting with Joachimk, and Peterbra. They presented the MS System Software Strategy and Trends in the Intel PC Market.

maricba
Philiphs: Most of January was spent obtaining final signature for the new Master Contract and pushing the contract through the channels at MS. Additionally, Markba spent a great deal of time preparing for the Executive Briefing which is scheduled for February 12th. Philips is presently establishing a factory in Europe that will receive basic PC configurations from Montreal and prepere these PCs for distribution in Europe. Most of the Montreal staff has been in Europe helping to setup this new organization.

We have also had a great deal of activity surrounding tha licensing of PhotoCD directly from Philips to MS. This license issue is very complex and requires the cooperation of the highest levels of management from both our companies. It appears that we will be successful with this license and thus MS will be able to support PhotoCO on all Windows based devices. Part of the Feb. 12th agenda will be to come to agreement on this issue.

Northgate: NG was involved in the budgeting process for almost the entire month of January. As a result things were fairly quiet. We did accomplish the reconciliation of the outstanding moneys MS owed to NG and that did take quite a bit of time. Resolution to this issue was greatly appreciated by NG management. NG is also in the process of re-evaluating its marketing strategics and considering vertical marketing approach to bundling software.

debbiefl
CompUSA: CompUSA has agreed to add a third year to their MS-DOS license and commit to 100.000 units. They have also agreed to license 50.000 copies of Windows for three years. They will preinstall Windows on their machines. They would not commit to a per-processor deal, as they do not want their costs burdened at the low and where they have so few margin dollars. They will be attending WinHEC and using our logos. They have completed the HCT. There is substantial opportunity for both high and low end applications business.

Creative Labs: The Bookshelf license was finally completed. There had been several snags, as MS had told them originally that they could sub license Bookshelf. Our agreements with the third parties, like Houghton Mifflin, do not allow us to sub license. Unfortunately, MM Pub wasn't aware of that. We also found out that CL cannot distribute Bookshelf in the UK because of a naming dispute MS is having with another company. An amendment allowing them to act at a replicator for OEM's with MMWin license was drafted and delivered. Pamelago and Debbiefl visited them this month to discuss MM marketing plans and the testing procedure we are implementing for MPC compliant components and systems. CL is extremely interested in being our supplier for Foghorn.

Media Vision: The issue that has been open this month is whether or not we can license them for Windows 3.1 to bundle with their Thunderport board at a price they find acceptable. They have been vary unhappy with what we have quoted them. There are several issues we are pursuing from a legal standpoint, so we can understand exactly what we have to provide them under their current contract. They have an amendment in their hands for signature that will extend their ability to upgrade their installed base until March 30. 1992 and allow them to act as a replicator for OEM's who have MMWin licenses with MS.

** MICROSOFT SECRET **
12

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035061 CONFIDENTIAL
[STAMPED]MS-PCA 1164643 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 13 of 16]

Pattyl
NCR - NCR signed up for OEM Premier Online support for $15,000. which went into effect on Monday, February 3rd. The new Mouse Distribution Agreement and the Master License Amendment are on Alok's desk for signature and will be signed whan he returns from Germany. Introduced Mark Buick as the new NCR Account Manager via the MS-DOS and Windows road show meeting at E&M Dayton. Steve Ballmer and Tom Mays had an excellent (last) call regarding Windows NT and Steve has been invited to Dayton for a strategy update meeting this Spring (Paulma will probably attend). E&M Atlanta came to Redmond on January 9th to understand our pland for MS-DOS 6 and it looks like we'll be able to accommodate their needs before MS-DOS 6 is available. E&M Columbia has been on site porting NT to the 3550 and there are plans to work out a program to ship NCR MP systems and NT code to selected (important) ISVs. The E&M Clemson NT enoineers were in Redmond for a metting with Carls and Mackm to discuss potential joint development of 'Win32 on DOS' (Panther) for their high-end workstations.

Johnmc
AST - Account turnover from Jeffd began. Traveled to AST and presented to 12 manager and director level personnel a presentation on MS-DOS and Wlndows plans for launch and future directions. Met with engineering group and listened to concerns of NT development, getting a better communication network up and running (direct email via gateway). Discussed AST's custom Configuration program for direct sales through the resellers. Met with Jim Schraith, VP of channel marketing , Julia Irvine, Marketing manager of mass merchant channel and had Mike Negrin present and outline MS marketing channels and strategies.

AT&T/NCR - Came to verbal agreement with Safari systems group to amend the Rhapsody agreement for MS-0ffice to utilize existing pre-paid from Rhapsody agreement for Safari systems. Met with Naperville Multimedia group and provided information on Win 3.1 and MultiMedia extensions. Naperville group outlined new product plans for a MultiMedia system due in Q193. Confirmed the Naperville group is now responsible for the upgrade MultiMedia kit developed by AT&T. Began formulation of royalty pricing for the PC mail groups Easy link project. Finally scheduled technical review meeting with 3140 group to review MS and NCR future products. Account turn over to markbu began and will be completed in February

Wyse Technology - Met with Rich Wesson and Jim Munro to address outstanding issues. The meeting was very encouraging. Discussed alternative arrangements for EBU products for the Wyse Decision laptop. Wyse will most likely license Works for Windows and Money on a Per System basis for a 6 month period. Discussed the license for Windows and came to a verbal agreement for Windows Per processor if issues surrounding corporate licensing programs can be resolved. Wyse also wants to present to MS executives the business case for the Windows NT Terminal proposal. Wyse sent a programmer and system to NT development group to begin the porting process of NT to the Wyse 700i. Met with NT technical staff to outline projected milestones for SMP port of NT.

Tomda
Zenith - This has been another month of high activity in the ZDS account. ZDS entered the mass market at the January '92's CES show and announced that Spinnaker Works will be shipped with all mass market machines. We have lost this business. ZDS is soon to announce a portable pointing device available for all ZDS portables. Logitech's Trackman is the likely winner. ZDS will ship in excess of 100k units of this product per year. We also met with ZDS and gained closure on Winball. Upon availability of Winball. ZDS will ship this product with all ZDS network ready machines. Groupe Bull and IBM announced that IBM will purchase a stake in Groupe Bull and that IBM will resell ZDS portable machines.

** MICROSOFT SECRET **
13

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035062 CONFIDENTIAL
[STAMPED]MS-PCA 1164644 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 14 of 16]

Jeffd
AST - AST's VP of Channel Marketing. Jim Schraith, visited MS's Mike Negrin. Alex Nadar, and John Williams to discuss co-marketing activities for Windows World, the mass-market channel, and MS's Fall Promotions. These discussions want well and will be followed up with Mike Negrin visiting Tom Yuen on March 4th. AST announced its new AST Office Advantage product line at the CES show in Las Vegas. Their press announcement stated that Lotus was on Office Advantage PCs which was not true (see issues). AST's account transition began. Johnmc introduced himself well presenting MS's system software update on MS DOS 5's APM and Windows 3.0 momentum. John and I will work together on the outstanding issues and the transition will be complete by February 13th.
AST continues to grow selling 120K systems in Q2.

Darcyh
HP Corvallis -Signed the LOI for Lion. Additionally, signed the MS-DOS 5.0 and Windows 3.X Amendment per processor. Remaining issues are what will be used for the PIM on Lion. MS Mail group presented a spec, for Address Book, as part of Schedule Plus, but did not fully meet HP's requirements. Chrisp has assigned Roberd full time as Project Mgr. replacing Josephr. Roberd will work to define best PIM solution for HP and to be key technical interface for Lion for MS. Delivered first draft of tha Agreement for Lion.
HP Grenoble
Met with Jacques Clay, GM, Grenoble, at dinner with HP Sunnyvale to discuss proposed Windows bundling proposal. Discussed Windows bundling options and how Windows could benefit Grenoble's planned network ready/capable desktops, planned for release this spring and Fall. Grenoble is interested in pursuing an exclusive bundling option similar to Corvallis' Lion. Would like to investigate joint development opportunities. Will be meeting with Jacques in February to investigate how Grenoble/MS can work together. Jimko, TAM, will be visiting Grenoble in Feb./Mar. to review support for HP, and how MS can better support Grenoble.
HP Sunnyvale
Held a meeting with Sunnyvale to hear about plans for 'Edison' a pen-based 386 PC, with email and fax capabilities. Planned release Q1 '93. Key opportunities for MS are Pen Win, MS Mail and Schedule Plus. Met with Duane Zitzner, GM, Sunnyvale, at dinner to discuss Windows bundling proposal. Discussed Windows bundling options and how Windows and Window applications could benefit Sunnyvale's planned 486/25 CPU upgrade able machine planned for spring '92. Submitted revised Windows proposal, and will review with Sunnyvale in effort to name Windows system by 2/28, which is required in contract in order to recover $100K of current PPB.
HP Colorado Networks Division (CND)/lnformstion Networks Division (IND)
Bobkr and Darcyh met with Information Networks Division (IND) and Colorado Networks Division (CND) in joint meeting on 1/30. IND is now responsible for all mktg. of LM for UNIX Systems. Discussed ways to promote joint networking strategy and reviewed strategy for LM for UNIX Systems. IND continues to defend Novell announcement as a simple 'balancing' message, as they have been leaning too far towards LM. Darcyh will be scheduling meeting with mktg. groups from MS and IND to further educate IND on LM strategy and to investigate joint PR opportunities.

Richab
Hewlett - Packard - The new year brought with it some good progress on the HP front. Highlighting the month was a successful executive meeting where HP gave us their final commitment to build TrueType into the Payette printer. In addition, agreements were drafted to put to rest concerns relating to ownership and disclosure plans for two interface standards.

** MICROSOFT SECRET **
14

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7035063 CONFIDENTIAL
[STAMPED]MS-PCA 1164645 CONFIDENTIAL

[Plaintiff's Exhibit 1243 Page 15 of 16]



                                                   +--------------+
| Jeff Lum |
| Director |
| Dept 8806 |
+--------------+
|
+ --- Debby Johnson
| Admin Asst.
+---------------------------+--------------------+----+------ -- -- -- -- ----+
| | | |
+---------------------+ +----------------------+ | +--------------------+
| Braman Account Team | | Chesnut Account Team | | | |
| Peter Braman | | Mark Chesnut | | | European OEM Sales |
| Group Manager | | Group Manager | | | |
| Dept. 1501 | | Dept. 1505 | | | |
+---------------------+ +----------------------+ | +--------------------+
Mark Baber Rich Abel Lorl Morrison Bengt Akerlind
Northgate, Philips Compaq Printer Bus Database Arministrator MS AB (Scandinavia)
Hewlett-Packard Printer Bus. European Sales Support
Teresa Chapman Meurtzio Bedina
Compaq Mark Buick MS SPA (Italy)
NCR/AT&T
Debbie Flynn Sandy Duncan
Activision, Headland Jeff Daniels MS LTD (England)
Compudyne, Media Vision Data General, EMI
Logitech, MAS Ignacio Feu
Tom Henningsgard MS SRL (Spain)
Austin Computer, Commodore Tom Davis
Swan, Tandon Sun, Zenith Jurgen Huels
MS GmbH (Germany)
TBH Darcy Holle
Atari, PC Innovations Hewlett-Packard, Momenta Pascal Martin
Tandem/UB MS SARL (France)
Patty Lazarus
NCR Hans Raneelaar
MS BV (Holland)
John McLauchlan
AST, Wyse


Microsoft Confidential
Rev. 1.31.92 Christu

[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7038064 CONFIDENTIAL
[STAMPED]MS-PCA 1164646 CONFIDENTIAL


[Plaintiff's Exhibit 1243 Page 16 of 16]

[The last page shows the members of the OEM sales team]
[Unreadable]


[STAMPED]HIGHLY CONFIDENTIAL
[STAMPED]MS7036065 CONFIDENTIAL
[STAMPED]MS-PCA 1164647 CONFIDENTIAL




PLAINTIFF'S EXHIBIT 1394 Gordon v. Microsoft   [ Top ]

[STAMPED "PLAINTIFF'S EXHIBIT 1394" BUT OCCURS AMONG RESPONDENT'S FILES ON THE COURT WEB SITE.]

Plaintiff's Exhibit 1394 Page 1 of 4

>From stefanir Thu Aug 13 21:41:20 1992
To: baerblb bengta bernardv carstens cwedell jeffl joachmk jochenh
Cc: berndk gerdab reginaw stefanir wolfe
Subject: URGENT-VOBIS Meeting with Lieven Aug 13, 1992/Impt. long email
Date: Tue Sep 01 09:10:45 PDT 1992

Date: Thu Aug 13 21:38:28 PDT 1992

You are all probably going to demand that my email rights be taking [sic] away after the length of this email, but I felt it was very important to communicate all the issues that happened today.....

Today I went to Vobis to discuss current business issues and to prepare for the agenda of items that Lieven wishes to discuss in next weeks meeting with Billg in London. Also attending the meeting with Lieven was Gerda Beining, Channel Marketing Specialist from the OBU. I invited Gerda along with me in order to approach Lieven on the Marketing Plan targeted towards working together with Vobis that we developed together (This plan was distributed to all of you on July 6th). We also wanted to present a proposal for working together on a Christmas Promotion and WinWord campaign in Germany.

Overall, the meeeting went very well and included some very positive reactions on many issues, which I need to tell you about and get your feedback as quickly as possible.

Lieven is looking forward to the meeting with Billg next week. He is excited to be able to discuss with him where MS's strategy is going and his own. Currently, Lieven is very positive towards MS and is very committed to the contract which was signed on July 2nd. The negative problem areas that exist in the relationship right now are the problems that have happened in the past and currently in the receipt of product tapes and documentation.

Lieven plans on mentioning this issue because he sees it as an area that continues to stand in the way of us working together smoothly and for him to be able to bring to market the products. Otherwise he is very positive.  I was also able to present to him in this meeting information on MS-DOS 6.0 and WFW.  He was very excited to hear about our intentions for MS-DOS 6.0 (for example H/D compression and our anti-virus product.) It turns out that he was lately in contract discussions with Central Point to license from them their anti-virus product, but they couldn't agree on price.  Needless to say, he was impressed and excited.

I also discussed with Lieven our plans for WFW.  Many of you are aware that our competitors in the market are aggressively targeting our OEMs, and they are not forgetting Vobis. LOTUS has been actively targeting Vobis in licensing and selling Lotus 1-2-3 and AMI-Pro for DM  88.00. BORLAND has also been approaching them lately as well with offeres and WordStar Germany has offered WordStar for a license of DM 2.00. NOVELL is also aggressively targeting Vobis with selling NetWare Lite and DR-DOS, especially since they are their top OEM in Europe.

The Sum of All of this is as follows:

Lieven has been giving thought to these offeres, but he prefers our product palette and feels that we are the market leaders. Price is ofcourse [sic] important to Lieven, but he would rather pay a slightly higher price to us knowing that he can offer the customer the leading market standard and high quality products.  What he is willing to do, is commit entirely in a "Strategic Alliance"....He even said he would be willing to no longer offer DR-DOS, LOTUS, or BORLAND etc.    In exchange for such

[Plaintiff's Exhibit 1394 Page 2 of 4]

a committment [sic] and investment from him, he would like to see an equal committment [sic] and investment from Microsoft. Along these lines, he would like to have a commitment from us for better customer support to him in areas such as documentation and in joint marketing and strategically working together in the marketplace.

On this topic, I would then like to refer to our discussion together with him on the topic of our Marketing Plan and the proposed Christmas and WinWord promotion. Call it fate, but it just so happens today he introduced a gentleman by the name of Marc Pastuere. He is a new member of the Board and is Lieven's right hand guy in getting Vobis and in particular the "HIGHSCREEN" name recognized in the market. The timing of our presentation and his coming on board couldn't have been more perfect. They invited us to Lunch to discuss the marketing proposals more in depth. After several hours of discussion, they loved the concept and are basically ready to do the following:

- After next weeks [sic] meeting with Billg, he wants to issue together with us a press release stating the "Strategic Alliance" to us and committment [sic] to Microsoft's strategy. (This would send a very direct statement to the public and the media that they have chosen MS, and not DRI, NOVELL, Lotus or Borland). [HANDWRITTEN UNDERLINE]
- They will invite us in early Octoboer to present and train their Store Managers and sales people from all of their stores in Europe on MS-DOS 6.0, Win 3.1, NT and WFW as well as applications.
- They will also commit to atleast [sic] initially to a WFW agreement for 25K WFW licenses with the anticipation of much more since they have plans to break into the networking market and target small businesses. They would NOT offer NetWare Lite and instead push us and they are planning on having a stand at the PC-Windows Show in Frankfurt Oct 14-17th, where they would issue jointly with us a press release stating that they are committed to MS and our strategy with WFW in networking. (NOT NOVELL's NetWare Lite)
- Lieven wants to do ACTIVE marketing with us.... He wants to actively on a Europe-wide base aadvertise and promote software in all 130 stores, including information displays, auto-demos, show window displays and stickers in windows etc.... He is willing to share costs with us and at NO CHARGE to MS bring us into his flyer "Denkzettel" that has a current distribution in Germany alone of over 6 Million, and around Christmas he estimates it to be at over 10 Million. He also wants to bring us into his new Advertising campaign for Highscreen computers which is appearing on Billboards and major business magazines (ie. Stern, Spiegel etc).
- Lieven also wants to start actively selling retail packages of software, in particular WinWord and Excel to his "Power user" that needs more than the Works for Windows product.

Many of you who attended the meeting at CEBIT92 where Lieven stated "That he doesn't care about software and is a "Hardware Seller only" and is not interested in marketing or selling software", are probably wondering now why Lieven had a 180 degree turn in his viewpoint. This is because quite simply, Lieven sees what is happening in the Marketplace and that he needs to react if he wants to continue to grow and be successful in the marketplace. He now understands and "Has seen the Light" regarding the importance of software helping sell his hardware. He knows that he cannot offer every operating system, every spreadsheet or every word processing product.

[Plaintiff's Exhibit 1394 Page 3 of 4]

He therefore feels that as the leading OEM in the Market he should partner with the leading software company....which he feels is Microsoft. In exchange for the above items which he wants to do with us he proposed the following committment [sic] from us on the Marketing side:

- To participate in the Microsoft Christmas Promotion in all of his 100 stores in Germany. He will only do this if we give him exclusivity and not if we are proposing on doing this with ESCOM or Shadt etc.
- He would then like to work together with us very aggressively along with a Decoration Service for Window display where he will also pay for a share of the costs.
 - He will put atleast [sic] 1 or 2 pages of free advertising of this promotion in all of the flyers/Denkzettels that go to approx. 10 million recipients per month.
 - He wants to plan to do joint datasheets that would then be displayed in info racks throughout his stores along with coupons to mail back for more information.
- He wants to do joint mailings on this to both the Vobis and MS database.
- He wants to be able to bundle WinWord and Excel with his computers during the time of this promotion. He wants them on commission and sell them as follows:
Machine bundled with Works for Windows: Price-DM2999.00
Same machine, but with option to choose WinWord or Excel: Price-DM3499.00
Same machine with both WinWord and Excel: Price-DM3949.00

(I realize that these are ridiculously low....Lieven knows it, but its [sic] what he proposed and he basically wants a counter proposal that is attractive and competitive)

- He also wants to sell all of our app products such as Office, Winword, Excel, Powerpoint, WIndows, PC-Word, Mouse, Publisher Money, Golf, and Flight Simulator. He wants to be able to sell them as closely as possible to the prices that mail order houses sell for and get them from us on commission and at a price where he can earn a little on it.

(All of the above is waiting for a counter proposal from us....even if we do not agree to all, he basically needs something to differentiate and would only want the special prices during the promotion period. Afterwords, [sic] he would go to standard prices and buy from distributors. He sees our benefits as being able to achieve in this short period time a high market penetration and get the update business. He says at the prices that our competitors are offering that we need to be aware that alot [sic] of OEMs may just sell it to get the business in the store.)

- After this promotion, he would want to continue to work ith us on a regular basis doing in-store promotions and havingnormal non-christmas information displays etc. as we had proposed in our original marketing proposal for Vobis. They also would like us to work together at the different sub-levels doing promotions in each country.


[Plaintiff's Exhibit 1394 Page 4 of 4]


He needs to know that if we do this project, that we really are committed and support them in making a success, just as they will.

He says that unless we can offer him something which is for a short period of time exclusive for 3 months and can help differentiate himself in the Market and from his competitors, then he is not interested in giving us exclusivity or any of the above mentioned items. Quite simply said, he needs to decide in the next month which partner with whom to work and he is asking us first. He wants us, but may have to choose another partner if we are not interested in this type of strategic alliance. I ask that you consider his proposal for working together and I am basically just reporting to you what Lieven and Vobis are interested in doing. The strategic decision for such lies with you, the management in how we should proceed. Lieven wants to act on this as soon as possible, especially since any Christmas promotions etc. need to be planned to take place as soon as November.

As follow up and to emphasize his interest, he has requested a meeting with Jochen Haink on Tuesday, August 18th in Munich.

Please respond as quickly as possible on your thoughts regarding this email and let me know how we should proceed.

Thanks

Stefanie


PLAINTIFF'S EXHIBIT 1397 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 1397 p. 1 of 4


[This is a fax cover sheet. Printed letterhead addresses at top and bottom impossible to read, mostly hand-filled out form.]
Microsoft·[fancy logo]

Microsoft GmbH
[ADDRESS / TELEPHONE /FAX]

Nordrrlassung(?) Brrlm:(?)
[ADDRESS / TELEPHONE /FAX]

Nieferlasmng(?) Kad(?) Hambury;(?)
[ADDRESS / TELEPHONE /FAX]

Microsoft·(fancy logo) GmbH

Telefax - Mitteilung
An : Bill Gates
Fax Nr.: [FAX]
CC:
Von: Stefanie_Reichel
Datum : August 17, 1992

Gesamte Seitenzahl: 14

Pre-meeting briefing notes + VOBIS Account Profile

Bill -

Please find attached(?) the requested(?) information for the meeting with Vobis in London this week. I will be briefing you in more detail on Tuesday at the Hyatt Carlton Towers.

Regards,

Stefanie Reichel

[LETTERHEAD FOOTER WITH ADDRESS / TELEPHONE ]

[STAMPED] SR00015 CONFIDENTIAL
[STAMPED] MS-PCA 1164660 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT 1357(?) Reichel(?) 8/26/98 CER(?)

[Plaintiff's Exhibit 1397 Page 2 of 4]

Microsoft - Vobis

Pre-Meeting Briefing Paper

DATE: August 17th, 1992
 
TO: Bill Gates
 
CC:
  Bengt Akerlind Jochen Haink
  Jurgen Huls Joachim Kempin
  Jeff Lum Christian Wedell
 
FROM: Stefanie Reichel
 
Re: MICROSOFT - VOBIS Executive Meeting in London
August 19th, 1992

Attached are pre-meeting briefing notes and the Vobis Account Profile for your reference in preparation for the meeting with Vobis's Founder and President, Theo Lieven. The meeting will take place on the day before the Annual Board of Directors Meeting being held in London, England.

The meeting will be a lunch meeting where current and future strategic issues related to our two companies working together will be discussed and an opportunity for you and the President and Founder of Vobis to be able to meet. I intentionally did not include a formal agenda because I would like to have the meeting be more informal. The pre-meeting briefing has been scheduled for Tuesday, August 18th at the Hyatt Carlton Towers hotel in London at 2PM where I will review the issues and agenda with you in more detail.

[STAMPED] MS-PCA 1164661 CONFIDENTIAL
[STAMPED] SR 00016 CONFIDENTIAL

[Plaintiff's Exhibit 1397 Page 3 of 4]

Microsoft - Vobis

Pre-Meeting Briefing Paper

Meeting Date: August 19th, 1992
 
Time: 1PM
 
Location: Restaurant "Chez Nico"
90 Parklane (by Hyde Park)
London, England

Tel.#: 44-71-409-1290

Vobis Attendees:

Theo Lieven Founder and President

Microsoft Attendees:

Bill Gates CEO and Chairman
Jurgen Huls OEM Sales Manager - MS GmbH
Stefanie Reichel OEM Account Manager-Vobis

Purpose of Meeting/Overall Meeting Objectives:

1) To demonstrate to Lieven/Vobis that Microsoft is committed to them as a partner and that we want to build a successful and mutually profitable business relationship with them.

2) Discuss Microsoft's future product and systems strategy. (MS-DOS 6.0, Windows NT and Workgroups for Windows.)

3) Discuss Vobis's plans for the future and how they plan on responding to the quickly changing market.

4) Get commitment for a "Strategic Alliance" between Microsoft and Vobis to work together closely in the market.

5) Get a commitment to get DRI/Novell out of the Account.

6) Get a commitment for Windows for Workgroups.

[STAMPED]MS-PCA 1164662 CONFIDENTIAL
[STAMPED]SRO00I7 CONFIDENTIAL

[Plaintiff's Exhibit 1397 Page 4 of 4]

Microsoft - Vobis

Pre-Meeting Briefing Paper

Key Issues to Address:

- Windows for Workgroups

- Windows NT

- MS-DOS 6.0

- New Contract and Marketing Actions

Hot Issues:

- Late Delivery of Documentation and Products

- Problems of Past

- Pricing

NOTE:
Please review the attached Vobis Account Profile which contains up-to-date information on the status of the account and an Executive Summary and personal profile on Theo Lieven.

[STAMPED]MS-PCA 1164663 CONFIDENTIAL
[STAMPED]SR000I8 CONFIDENTIAL



PLAINTIFF'S EXHIBIT 1513 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 1513 Page 1 of 10

VOBIS Microcomputer AG
Account Profile for Q1FY93
Microsoft·(fancy logo)
&
VOBIS

PLAINTIFF'S EXHIBIT 1339 Reichel(?) 8/26,6g(??) CER(?)
Stefanie Reichel
OEM Account Manager - Germany
SR 00034 HIGHLY CONFIDENTIAL
EXHIBIT
vergnes(?)
45
6.2207(? - half of line cut of due to end of fax or something)

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195373 CONFIDENTIAL

[Plaintiff's Exhibit 1513 Page 2 of 10]

Executive Summary

Vobis Microcomputer AG (Vobis) represents one of the most strategic accounts to Microsoft both in terms of the revenue they bring us, but also because of their fast growth and strong presence in the German and European market.

From a market share standpoint, Vobis is by far the largest manufacturer and seller of IBM compatible computers in Germany with 1991 annual shipments of 270,000 at a sales revenue estimated at over DM1 Billion ($650 million - the figures have not been officially reported yet). In a recently published IDC report. Vobis was placed as having 15.3% of the Intel-based PC market share in Germany, leading by a significant margin against IBM which had 10.5%, Siemens-Nixdorf with 52%, Escom with 5.1% and Compaq with 5% respectively.

Vobis is also projected to continue to grow rapidly and dominate the German market and eventually expand into the rest of the European market with as much momentum. In the last three quarters alone, they have shipped over 255,000 systems, and expect to ship a record 350,000 systems this CY92. This will place Vobis as the largest shipper of PC's in the German market

From a royalty revenue standpoint for Microsoft, Vobis is by far our largest OEM in Germany, and one of our largest in Europe. Since they started shipping our products with our systems in Q2FY91, they have brought in $2,506,980.00 in royalty revenue for FY91. Actual revenue for FY92 was $7,449,632.00 although they had been forecasted at $5,174,336.00. Effective July 1st, 1992, Vobis committed to a new I8 month contract agreement for an annual commitment of $11,880,000.00. This commitment includes new licenses for Works for Windows and Entertainment Pack. In addition, we believe there is potential for additional revenue for licensing products such as Windows for Workgroups. This could mean royalty revenue for Microsoft exceeding $12MM in FY93.

Vobis's strengths and reasons for such success lie in their ability to offer the lowest and most competitive prices for PCs in the market In addition, they have more visibility and presence than any other PC reseller due to their network of 100 sales outlets in Germany and 30 others throughout Europe. This has allowed them to offer customers more convenient locations and easier "one stop shopping" than is readily available elsewhere. They also actively advertise their aggressive pricing for PCs in a monthly collateral piece which is inserted in the top 10 newspapers and magazines in Germany with an estimated circulation of over, 6 million. They also have fast logistics and a quick central operations group in their headquarters.

Vobis's weaknesses lie in their growing reputation for poor quality control in their products and in customer support. This is where many of their competitors are trying to steal market share from Vobis, even if they cannot compete with them on price.

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00035 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195374 HIGHLY CONFIDENTIAL

[Plaintiff's Exhibit 1513 Page 3 of 10]

Our greatest challenge and threat in this account has been DIGITAL RESEARCH (DRI). Vobis is still DRI's largest OEM in Europe and where they have had their strongest foothold in an account. There are many reasons that DRI was able to get this position with them. One of them is that Vobis does not want to feel dependent on one vendor and this means Microsoft. Another reason is that internally at Vobis there are several "DRI Disciples" who preach within and to the Vobis customer base that DR-DOS 6.0 is better than MS-DOS 5.0.

Our challenge in countering this DRI threat is slowly being achieved and in the last three quaners alone, the percentage of DR-DOS sold to MS-DOS has dropped from 50% to 18%. In addition, Windows is taking off. In Ql and Q2FY92, our percentage of Windows penetration was 50%. Since Q3FY92, the Windows penetration on their systems was increased to 63%. It is currently after Q4FY92 at 77%. By Microsoft continuing to work together more closely and supporting Vobis with a dedicated Account Management team, by dedicating marketing monies for projects such as training and seminars for their store managers and promotions in their stores (such as with Windows), we will be able to create needed visibility and continue to chip away at DRI's stronghold in this account as well as any other competitcrs.

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00036 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195375 HIGHLY CONFIDENTIAL

[Plaintiff's Exhibit 1513 Page 4 of 10]

Company Business Profile

Vobis Microcomputer AG (Vobis), a manufacturer of IBM Compatible personal computers and systems, was officially founded in 1975 in Aachen, Germany by Rainer Filing and Theo Lieven. The company originated as a purchasing club at Aachen University of Technology (Germany). Within two years it had become a full-time mail- order business operating out of the garage and apartment of the two founders with a turnover of DM 2 million ($1.15 million) in their first year of business. By 1980, obis had reached a turnover of more than DM 10 million (S5.7 million). The company has since then grown rapidly at an average rate of almost 100% each year.

In 1989, in order to finance and maintain their tremendous growth rate, Vobis sold a 50% share of their business to KAUFHOF Holding Group, the largest retail store chain in Germany and a division of the Metro International Group. By establishing this new distribution channel and partnership with KAUFHOF, Vobis was able to expand its market from the private sector to the semi-professional and professional market Currently, Vobis markets and sells direct to business enterprises and home users with their own brand name called 'Highscreen'. They accomplish this in Europe through their own network of 130 sales outlets and the network of their partner's retail outlets (Kaufhof and MediaMarkt) and small business wholesale centers (Metro) which is estimated at being over 400 locations.

The projected annual sales for this CY92 are 350,000 systems at over DM 2 Billion in revenue (S1.2 Billion). Vobis's primary business and sales are in Germany and they continue to dominate and grow at a strong rate in the German market. However, Vobis's main strategic goal is to continue to aggressively expand with additional stores throughout Europe (mainly Spain, Italy, Belgium, Holland, later France, UK and Poland), and to grow these markets as successfully as they have grown the German market.

Vobis currently manufacturers their PC's in locations such as Berlin, Vienna/Austria, the former Philips Computer Plant and through the hardware manufacturer Aquarius Computer. In addition, they just completed building a Warehouse and Assembly Plant near their headquarters in Aachen, Germany where they will be able to centrally store and ship their products to all stores throughout Europe with an immediate response time.

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00035 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195376 HIGHLY CONFIDENTIAL

[Plaintiff's Exhibit 1513 Page 5 of 10]

Quick Personal Profile on Theo Lieven:

- Founder, President and Member of Board. Age 40.

- Attended Aachen University of Technology in Germany, receiving a degree in Mathematics.

- Lieven is the main contact for Microsoft. He is a very complex character who needs special attention to his success and ego-stroking. He is an intelligent, very eccentric man with a great sense of humor. He can be sensitive and if he is teased in a negative manner, he shuts off completely.

Examples of his extraordinary lifestyle and personality are:

  • He is still a registered student in the 40th semester in mathematics
  • He is a concert pianist with over 12 years of training and lessons
  • He is a certified pilot and flight instructor
  • He drives only 12 cylinder cars (Jaguar, Mercedes, BMW) and collects eld sports cars
  • He indulges daily on gourmet food and expensive old wines

Previously, Lieven did not see the importance of the Operating Systems or software applications to his company's success. He was aggressively marketing several software products and three (3) operating systems (DR-DOS, MS-DOS, and OS/2). Up until recently, he favored DR-DOS despite the fact that he had to pay for these licenses extra, since he has a per processor agreement on MS-DOS. He is most likely paying about $5 for DR-DOS on a per copy basis.

However, with the signing of the new contract agreement. Lieven has been showing signs of wanting to work closer with Microsoft in a "Strategic Alliance". He has recently come to realize that the market is becoming increasingly competitive and he needs to be able to differentiate himself from his competitors. He also has had the problem of a growing reputation for poor product quality and service.

In order for Lieven to continue to keep Vobis growing and be successful, he feels that as the leading OEM in the Market that he should partner with the leading software company ...Microsoft. He is even willing to discuss no longer offering DR-DOS if we are willing to work together with him in marketing in the European market and helping him sell our software with his hardware. The Marketing Plan which was written towards working together with Vobis was presented to him and he is willing to allow us to include in his stores information racks, auto-demos, window displays, posters, stickers in windows and joint advertising in his flyers. We are currently discussing with the GmbH how we can agree on a joint strategy to work together and present a proposal to Lieven.

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00035 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195377 HIGHLY CONFIDENTIAL

[Plaintiff's Exhibit 1513 Page 6 of 10]

Vobis at a Glance*


Name and Address;

Vobis Microcomputer AG
Postfach 1778
Rotter Bruch 32-34
D-5100 Aachen
Germany


Business:

Original Equipment Manufacturer (OEM) of IBM Compatible personal computer systems.

# of retail stores: # of employees:
 
Germany: 100
Austria: 10
Switzerland: 4
Italy: 5
Belgium: 2
Holland: 9
1283
 
Annual Sales: Channels:
 
1989: DM 280 million Vobis Stores: 90%
1990: DM 540 million Distributors: 9%
1991: DM 1 Billion + (Figures not reported yet) Mail Order: 1%
 
Markets: Sales by Territory:
 
Individuals/Semi- Professionals: 75%
Medium Business/Fortune 250: 15%
Education/Government: 0%
Germany: 80%
Other: 20%
* Most figures above are based on 1991, but his been updated with information from 1992.

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00039 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195378 HIGHLY CONFIDENTIAL

[Plaintiff's Exhibit 1513 Page 7 of 10]

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00040 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195379 HIGHLY CONFIDENTIAL

Product Line

Vobis sells a complete line of IBM compatible desktops, laptops, servers and accessory products. This product line is as follows:

Desktops
Name of Product Line: Highscreen Kompakt Serie III
# of Desktops: 6
Processor Types:
AT 286-16 AT386SX-25 AT386DX-33
AT386SX-16 AT386DX-25 AT486SX-20

Laptops:
Name of Product line: OEM Notebook Series (No Name)
# of Desktops: 4
Processor Types:
286 386DX-33
386SX-20 486DX-33

Servers
Name of Product Live: Highscreen Tower Serie
# of Desktops: 5
Processor Types:
AT386DX-25 AT386DX-33 AT486SX-20
AT486DX-33 AT486DX-S0

Software Bundles

Presently, when Vobis sells any of the above PCs to a customer, they receive a cardboard carrying box called "Software Package" in which they receive the following software bundle:

  • DR-DOS 6.0 -OR- MS-DOS 5.0
  • Power Basic Basisversion
  • PC Tools 7.1
  • Microsoft PC-Works 2.0
    -OR-
    Windows 3.1 and WAYS 1.1 with the book tided "Ways" by J. Schieb.
  • Microsoft Works for Windows 2.0 on select 386/486 Windows machines.
  • Microsoft Entertainment Pack Volume 1 on their Video Pro 286 machine.
  • HP New Wave 3.0 (on all 386 tnd 486 processors with more than 4 MB).

Other Products
OEM Brand Name (Olivetti, He-Mett Packard, Epson, and other)
Printers, Monitors, Modems, Scanners and Mouses;

[Plaintiff's Exhibit 1513 Page 8 of 10]

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00041 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195380 HIGHLY CONFIDENTIAL

Unit Shipments and Processor Mix

* The above is based on actuals that we have through Q4FY92.
All other is based on our Forecasts.

[The histogram shows sales of Intel 8086, Intel 286, Intel 386, Intel 486 for each of 11 quarters from Q2FY91 through Q4FY93.  It is difficult to read the histogram due to the poor quality of the image, but total shipments for each quarter range from 60,000 through 120,000, more or less increasing from quarter to quarter.]


[Plaintiff's Exhibit 1513 Page 9 of 10]


[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00042 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195381 HIGHLY CONFIDENTIAL

Royalty Revenue of Microsoft Products

[IMAGE: Histogram of Royalty Revenue of Microsoft Products, over six quarters from Q2FY1991 to Q3FY1992.]
  Revenue for FY91 - $2,506,980.00
Revenue for FY92 - $7.449.632.00

[Plaintiff's Exhibit 1513 Page 10 of 10]

[STAMPED] HIGHLY CONFIDENTIAL
[STAMPED] SR 00043 HIGHLY CONFIDENTIAL
[STAMPED] MS98 0195382 HIGHLY CONFIDENTIAL

License Profile/Contract Summary

  • New License Agreement effective July 1,1992 - December 31. 1993. Total per processor royalty is $29 .70.
  • All products to be pre-installed and they are a Ready-To-Run licensee is well.
  • MS-DOS commitment doubled from 200K to 400K units annually per processor.
  • MS-DOS royalty at $9.00 for all language versions.
  • Windows commitment increased from 100 K to 320 K units annually on all 386 and 484 systems.
  • Windows royalty at $15.00 for all language versions.
  • In addition, they have committed in a new separate one year contract for 100K units of Works for Windows Version 2.0, 80K units of PC-Works Version 2.0 md 40K units of Entertainment Pack Vol.1 Version 1.0.
  • They will also no longer be licensing OS/2.
  • They no longer have update rights for products and will be including Microsoft registration cards in with all product manuals.

Summary of Terms:


Product
Uplifts**
List price License Royalty* Volume m/c Int'l
 
MS-DOS 5.0
with shell
$15.00 P/P
Pre-install
$9.00 400.000 $3.600.000.00 None
 
Windows 3.1 $19.00 P/P
Pre-install
$15.00 320,000 $4,800,000.00 None
 
Works
for Windows
Version 2.0
$26.00 P/System
Pre-install
$24.00
(annually)
100.000 $2,400,0O0.OO None
 
Works
Version 2.0
$12.00 P/System
Pre-install
$12.00
(annually)
80,000 $960,000.00 None
 
Ent. Pack
for Windows,
Version 1.0
$3.00 P/System
Pre-install
$3.00
(annually)
40,000 $120,000.00 None
 
Test
Software
$0.00 P/System
Pre-install
$0.00
(annually)
400,000 $0.00 None
 
Yearly Total in Royalty Revenues: $11,880,000.00






PLAINTIFF'S EXHIBIT 3247 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 3247 Page 1 of 1]

[handwritten] re you _happy?_

Theo Lieven, President and CEO
Vobis Microcomputer AG

MICROSOFT VISIT ITINERARY
In addition to the attached agenda for the OEM Briefing Conference, the following meetings have been scheduled and arranged for your visit to Seattle:

Tuesday, September 22
[handwritten] does he meet BillG? - Debbyjo said for ½ an hour either before or after his speech - I'm waiting for confirmation - apparantely[sic] he is flying in for the speech + then flying out...

7:00 p.m. Arrive Seattle (from Atlanta on Delta Airlines)
Seattle Limousine Service (Steve Morgan [TELEPHONE]) will pick up and deliver you to the Seattle Sheraton Hotel.

Accommodations
Seattle Sheraton Hotel [TELEPHONE / FAX]

[handwritten] Ask Debbyjo to make arrangements + re: Billg mtg.

Wednesday, September 23

Accommodations
Seattle Sheraton Hotel

Thursday, September 24
7:00 a.m.  Breakfast Meeting

Microsoft Attendees:
Steve Ballmer - Sr. VP of Sales & Marketing
Bengt Akerlind - Director of European OEM Sales
Stefanie Reichel - OEM Account Manager

Location:
Banner's Resaurant (Sheraton Hotel)

6:30 p.m.
Seattle Limousine (Steve Morgan [TELEPHONE]) will pick up and deliver the following individuals to the Hunt Club ([TELEPHONE]) at 900
Madison, Seattle:
Theo Lieven
Juergen Huels
Joachim Kempin
Stefanie Reichel
Bengt Akerlind

7:00 p.m.
Dinner Meeting at The Hunt Club (Sorrento Hotel)
Seattle Limousine will transport the entire group back to the Seattle Sheraton after dinner.

[STAMPED] SR 00008 CONFIDENTIAL
[STAMPED] PLAINTIFF'S EXHIBIT [illegible]
[STAMPED] MS-PCA 1164685 CONFIDENTIAL

PLAINTIFF'S EXHIBIT 3475 Gordon v. Microsoft   [ Top ]

[STAMPED PLAINTIFF'S EXHIBIT 3475 BUT OCCURS AMONG RESPONDENT'S FILES ON THE COURT WEB SITE.]

[Plaintiff's Exhibit 3475 Page 1 of 3]

[STAMPED] MS 5055904 CONFIDENTIAL
Second feature: I have a Syquest drive. So far I haven't been able to install DR6. It seems that DRDOS is trying to create a directory in the Syquest Drive. The Syquest drive returns "Not ready error reading drive D" ARF.

I'll try removing the Syquest drive!

######################################################## 474
>From ericst Thu Sep 19 11:36:50 1991
To: dosbug
Cc: bobcook dos5beta
Subejct: Re: didn't work
Date: Thu, 19 Sep [19]91 11:37:31 PDT

You must run from the dos prompt. If you try running it in a Windows dos VM, you will get a message saying you can't run under windows. If you try running directly from windows, screen just flashes.

Exit windows and run.

Bugmaster: Please enter the following bug:

Sev 3 - Fix for 5.0a

Getting Started section on problems running setup needs to tell users to not run from windows.

Eric.


>From bobcook Thu Sep 19 09:32:44 1991
To: dos5beta
Subject: didn't work
Date: Thu Sep 19 09:30:36 PDT 1991

screen flashes and it exits with no messages. tried reading trouble.txt but its too large for vi or notepad, tried setup /i and same thing. running under windows 3.0 on northgate 425i.

######################################################## 475
>From johnen Thu Sep 19 11:46:16 1991
To: bradsi jimla
Subject: FW: Bambi:&DR DOS 6.0, may files turned into crosslinked mess
Date: Thu Sep 19 11:36:08 PDT 1991

>From lawren Thu Sep 19 11:34:15:1991

[Plaintiff's Exhibit 3475 Page 2 of 3]

[STAMPED] MS 5055905 CONFIDENTIAL
To: winbug
Cc: bambi scottq lawren
Subject: Bambi:&DR DOS 6.0, many files turned into crosslinked mess

Windows Bug Report

Title: Bambi&DR DOS 6.0, many files turned into crosslinked mess
Opened By: lawren
Sev: 1
WinVer: 3.10  47
Product: Drivers
Area: Bambi
Origin: Internal
Sugg.Assign: scottq
DOSVer: 6.00
Mode: high
Name:  Phone:

Problem Statement: Had run fine other times. Did a "DIR" and I got garbage characters. Chkdsk found well over 100 files with crosslinked files, invalid clusters, and invalid & incorrect lengths.

Machine Configuration: Cumulus 386 2M VGA
Config.Sys:
Autoexec.Bat:
Form #: SYS003 Version#: 2.00 Revision Date: 08/22/[19]91
####################################################### 476
From karlst Thu Sep 19 11:55:06 1991
To: steveti
Cc: bradsi philba
Subject: jeanp/vacation
Date: Thu Sep 19 11:34:54 1991

I believe we've settled the issue with jeanp regarding his vacation. Don't call him. Kudos to bradsi for making this happen.

####################################################### 477
From philba Thu Sep 19 11:58:17 1991
To: lawren
Cc: bambi bradsi lawren scottq
Subject: Re: Bambi:&DR DOS 6.0, many files turned into crosslinked mess
Date: Thu, 19 Sep 91 11:34:11 PDT

[Plaintiff's Exhibit 3475 Page 3 of 3]

[STAMPED] MS 5055906 CONFIDENTIAL
tee hee hee... thats what you get for running an incompatible OS

sorry I couldn't resist. Scott, lets figure this out asap -- could be a problem thats lurking for real DOS

>From lawren Thu Sep 19 11:34:15:1991
To: winbug
Cc: bambi scottq lawren
Subject: Bambi:&DR DOS 6.0, many files turned into crosslinked mess

Windows Bug Report

Title: Bambi&DR DOS 6.0, many files turned into crosslinked mess
Opened By: lawren
Sev: 1
WinVer: 3.10  47
Product: Drivers
Area: Bambi
Origin: Internal
Sugg.Assign: scottq
DOSVer: 6.00
Mode: high
Name:  Phone:

Problem Statement: Had run fine other times. Did a "DIR" and I got garbage characters. Chkdsk found well over 100 files with crosslinked files, invalid clusters, and invalid & incorrect lengths.

Machine Configuration: Cumulus 386 2M VGA
Config.Sys:
Autoexec.Bat:
Form #: SYS003 Version#: 2.00 Revision Date: 08/22/[19]91

####################################################### 478
From greglo Thu Sep 19 12:10:11 1991
To: bradsi
Subject: Re: bambi
Date: Thu, 19 Sep 91 12:05:40 PDT
The new WzMail version (which uses the WinMail document handling routines rather than the ancient DH library) also gives you a huge perf boost. If you use WzMail you want it.




PLAINTIFF'S EXHIBIT 4178 Gordon v.Microsoft   [ Top ]

[Plaintiff's Exhibit 4178 Page 1 of 1]


Windows and DOS Development
August Status Report
Phil Barrett

Windows 286 & 386 Rel 2.1D
Release 2.1D development work was completed in August. The product was release[sic] to Manufacturing Sept 6 A 2 1D
release of the 386/XT version of Win386 was not released since no bug fixes in 2.1D affected it.

Windows Rel 3.0
Windows 3.0 moved into high gear in August. The Windows 3.0 product plan was completed and presented to SteveB and BillG for review. The basic plan was well recieved[sic] with one major addition(?) -- Protected-Mode Windows. Thanks to some midnight efforts by DavidW and BobGu?s a
demonstration system was available to prove the feasibility(?) of pW. This was very well recieved[sic]. More planning has been done and it appears that an end of July 1989 ship date is feasible.

A fair amount of progress was made in August. Drop shadows on menus and proportions system font are working The keyboard friver reorganization was completed -- this allows us more flexibility for supporting international keyboards. Windows 386 scheduler is working giving us the ability to boost the priority of the VM using the Keyboard providing for a much smoother response.

Interface/Tools
The Windows 2.1 SDK beta release was made in August. This was mostly to gain feedback on the suitibility of CVW for general use. The feedback was quite positive so the product was released to manufacturing and is now shipping to customers.

Planning for the 3.0 SDK began in August with the specification of the work areas that will be focused on. External suppliers of tools are also being examined(?). Whitewater Group and Eikon Systems are two.

DOS
August saw the IBM release of DOS 4.0. August also saw the discovery by the press of DOS 4.0 bugs. Probably the biggest complaint is poor EMS support. This and several other problems have been corrected for the PC-DOS release. In addition the maintenance release will have a reduction of the number of physical pages consumed by expanded memory usage. A Beta OEM release was made in August and the Packaged product release is planned for Sept. Progress on EMM386 was made in August with substantial progress made on both functionality and size. Plans are in place for a second round of improvements (most size) for the maintenance release.

August was a big month for DOS planning. Evaluation of the IBM Lifeboat(?) (4.1) specification occured in August (and is ongoing) DOS plans
were presented to BillG and we recieved[sic] the go-ahead to investigate Protected-Mode DOS. Initial planning started on a Retail Upgrade Package of DOS.

Recruiting & Personel [sic]
Four offers(?) were accepted in August making it(?) a top month for recruiting. We recieved[sic] accepts from Clark Cyr (Tools). Guther(?) Ziaber(?) (GDI/Drivers). Laitha(?) (Printer Drivers) and Sriram(?) Rajagopaian(?) (DOS). Thanks to Beth Davies and Susan(?) Vositer(?) for a top-notch job. The bad news. we are still 7 under our FYSS(FYBS?) plan. We have also obtained a Waterloo Coop student and there is a possibility of a second. This will help out quite a bit. Current recruiting(?) status:
Offers Out: Mike Morgan. Paul Klemond
Offers Rejected: Eric Meyer (salary)
Headcount Plan:
Group Current FY89
Windows 386 5 5
HPC 1 4
DOS 5 8(?)
CD-ROM 2 2
Group 1 1
Total 28 35

[STAMPED] CONFIDENTIAL
[STAMPED] ATTORNEYS ONLY
[STAMPED] 10152375
[STAMPED] EXH 49 DATE _5/17/[20]02 WITNESS Barrett MARY W. MILLER
[STAMPED] MS-PCA 1148405 CONFIDENTIAL

DEFENDENT'S EXHIBIT 417 - Gordon v. Microsoft   [ Top ]

Defendant's Exhibit 417 Page 1 of 5

[STAMPED] EXHIBIT 67 Barrett 5/31/[20]02

[The exhibit is a photocopy of the cover of a book. The cover art shows a man in a business suit (Bill Gates) posed like Rodin's "The Thinker," looking up at a lightbulb with the Internet Explorer logo on it. The lightbulb is flanked with images of the Earth as a globe.]

HOW THE WEB WAS WON
HOW BILL GATES AND HIS INTERNET IDEALISTS
TRANSFORMED THE MICROSOFT EMPIRE

With a new chapter on the impact of the Justice Department trial verdict.

PAUL ANDREWS
COAUTHOR OF THE NATIONAL BESTSELLER GATES

Defendant's Exhibit 417 Page 2 of 5

[Pages 140 and 141 of How the Web Was Won]

nism for Gates to be exposed to the Internet side of the online service argument. Silverberg sensed that because of Gates's close relationship with Siegelman and Myhrvold, the chairman had been hearing a one-note song. Nevertheless, for Silverberg, Shumway was hardly an earthshaking occasion. He listened to the debate of online versus the Internet with interest. It was an intriguing psychological dynamic for Microsoft, but it had little impact on his thinking. His mind was made up. Shumway merely reaffirmed the need for him to keep moving down the trail he had already been blazing. As far as he was concerned, the debate could continue without him. The Windows effort could not affort to sit on its hand, waiting for an elusive consensus to emerge. By then it would be way, way too late.

So Silverberg started the Chicago team down an Internet path that was in many ways parallel to the goals of Siegelman's online effort. Chicago was firmly in the camp of supporting open Internet protocols for things like e-mail, security, and dialing up from home. Marvel was building its service from the ground floor up, on its own e-mail and publishing and dial-up protocols, with the hedge that if users wanted Internet access, they would be able to get there from Microsoft's online service.

The Siamese-twin approach had enormous inefficiencies in development and personnel overlap. It was the kind of budget drain most executives and big companies would never countenance. Choose one or the other, they would direct their managers. But Gates saw benefits to multitasking the online strategy. It gave him the chance again to play two hands at once, as Microsoft had with parallel OS/2 and Windows development. Competition was important, even if it was internal. And Gates was loath to discourage entrepreneurialism within his ranks. Creative tension was needed in an organization for it to thrive and move forward. Gates was not going to stand in the way of a process that would save Microsoft from becoming a Wang or an Apple or a Lotus or an IBM.

Gates also was caught in the bind of the Silverberg-Siegelman personaility conflict. It too was nothing new in Microsoft's competitive, ego-driven culture: "It's just another thing you have to manage," Gates said later. In this case, he saw benefits to a macro, not micro, managed approach. He had given Siegelman the green light well before the Net was a factor. And at the Shumway retreat he had made it obvious that Silverberg was to intergrate the Internet into Windows. Gates was like the basketball coach having two point guards play one-on-one to see who would get the starting assignment.

After the Shumway retreat, Silverberg met with Phil Barnett, a lead systems manager who had just joined the Windows 95 development effort. Silverberg asked him to look at how Internet capabilities could be woven into Chicago. Included on the list were Allard's initiatives regarding TCP/IP, ftp, telnet, WAIS, auto dialer, and other Net acces features. Silverberg added another item to the laundry: browsing capability. Should we include a browser with Chicago? he asked Barrett. What would be the browser's role vis- à-vis Chicago connectivity with the Web? Should we build it ourselves, from the ground floor up? What would that take, in terms of resources and time? Would it be better to license or buy existing technology and improve on it? Silverberg did not want to rush headlong into a drain on Microsoft resources. There were lots of browsers out there, after all, and little discernable demand. There was still plenty of time, it seemed, for Microsoft to make its play in the browser sweepstakes.

Barrett hired two part-time program managers and by midsummer had a college intern on hand to help out with product management. But his primary focus was on Chicago, not the Internet. "Everyone was focused on getting Chicago out," he recalled. "Bill may have said the Internet is very, very important, but organizationally, I don't think that took right away at all." Barrett took on the assignment, but for him the Internet was not a huge action item. As for the browser, Barrett had heard nothing about integrating it into Windows at the Shumway retreat, and he felt little urgency to pursue the issue.

For Silverberg, however, browsing in Windows was a top priority. The Windows three-year plan he presented after Shumway specifically outlined "integrated Net browsing in [Windows] Explorer." [Square brackets in original] He was not sure what form it would take, but browsing needed to be there. On board as well was John Ludwig. "It was clear from Shumway that we needed to Internet-enable our operating system much, much, much more, and that a browser was the most important part of this," Ludwig later recalled, even if all the t's were not crossed or the i's dotted.

Six weeks after the retreat, Silverberg attended Windows World at Spring Comdex in Atlanta, with an eye toward finding out what Windows vendors were doing with the Internet. Sinofsky was there with a similar goal in mind. The two hooked up and strolled the floor together. In a tiny booth tucked away on a side aisle they found gold -- or at least some glitter. BookLink Technologies, Inc., a small software developer based in Wilmington, Massachusetts, was showing an early iteration of Internetworks, browsing technology that integrated tightly with Windows. Silverberg and Sinofsky had learned of BookLink from Allard, who knew one of the company's principles, Bill Hawkins, though various Internet conferences. Hawkins, who

Defendant's Exhibit 417 Page 3 of 5

[How the web was won, pages 162 and 163]
For Silverberg, long the Marvel skeptic, the problem was that the browser, and by extension the Internet, was too low a priority for Marvel. Siegelman had simply blown a huge opportunity, and now Microsoft was back to square one. Silverberg was determined to have browser technology in Chicago. It looked like his team would have to develop or obtain it themselves. John Ludwig, the networking veteran who was "blue-skying" future opportunities for Windows for Silverberg, agreed: We had better reassess where we are in the browser game, he put it to Silverberg. We need to decide whether we are serious about this stuff, in which case we had better start moving faster and more decisively. Or we need to cut bait. It was a short conversation. As far as the Internet goes, Silverberg said, we're as hardcore as Microsoft has ever been about anything. It was more than a year before Bill Gates, on December 7, 1995, would say the same thing to the world at large.

Tall, gentle-mannered, and cerebral, Ludwig brough keen analytical skills and a calm rationality to the browser project. Ludwig monitoring a project was like a submarine tracking a target. He preferred working below the surface, unnoticed, while tirelessly and unflaggingly plotting political strategy, honing in on challenges and charting progress. Silverberg and Ludwig made a great alliance. Both hated ego-boosting or self-aggrandizing schemes. Both practiced a subtle form of leadership where they enabled those around them either to make the right choice or to learn from mistakes -- miscues neither of them might have committed, but which were necessary as lessons learned. Both drew more satisfaction from watching those around them succeed together than from calling attention to their own contributions. From mid-1994 on Ludwig was a critical part of each significant strategic decision Microsoft made on the browser front. Yet, innumerable articles and analyses of the browser competition almost never identified him.

There was an almost audible shifting of gears going on for the Redmond gang. Through early fall of 1994, browser development had more or less meandered along as part of the Chicago effort, but not a huge part. It was not so much that the browser was considered unessential or insignificant. The Shumway retreat, and Gates's mobilization e-mail immediately following, made it clear that integrating browsing capability into the operating system was a vital goal for the company's Internet effort. But it seemed unrealistic to expect that a browser could be cobbled together in time for Chicago's release, at the time still scheduled for the upcoming fall of 1994. Integrating an entirely new dimension would mean lots more coding, debugging, testing, coding, debugging, testing -- the seemingly endless programming cycle. If you altered one line of code in a program as complex as Windows, Silverberg was wont to point out, you usually stood the chance of introducing a bug or glitch, and so on down the line. Software development at its heart was a mind-drubbing, Sisyphan chore of debugs and fixes. Microsoft's ability to persist to the bitter end in ferreting out as many bugs as possible and in addressing user needs helped explain its successes where others had run out of ideas, steam, or initiative.

The previous spring, Barrett had been assigned to look into a browser, but with attention focused on more pressing issues in the Chicago upgrade, he had not put it on the front burner. Over the next few months he talked to a few people, looked over the field -- then consisting of a wild assortment of browsers that did one or two things well but overall were slow, underfeatured, and immature -- and drew up some preliminary specifications. But no team got assembled, no product description or business plan got drawn up, and no code got written. Neither did any alarm get sounded. The BookLink discussions were progressing along a normal path, after all. With most browsers available for free, there was no real commercial pressure on Microsoft. Once the Internetworks code became available, the thinking was the browser effort would be able to ramp up production quickly.

When the BookLink deal fell through, everything changed.

As luck would have it, and Microsoft often did have luck, a coding maniac by the name of Ben Slivka had other ideas. Bearing a striking resemblance to Anthony Edwareds -- Dr. Mark Greene on the TV series ER -- Slivka combined a studious demeanor with alacritous energy, stamina, and will. Starting the previous summer, Slivka had agitated Ludwig's next-generation Windows team to o something like Mosaic for Windows. Although it was not true that, in order to be part of Microsoft's Internet effort, your last name had to begin with "S" and feature some combination of "v," "i," "n," or "l," Slivka was a perfect fit for the company's aborning browser development. A veteran of the OS/2, DOS 5 and DOS 6 projects, Slivka had a ton of code under his belt and was known as a just-ship-it kind of guy. He liked impossible challenges, particularly if he could drag his friends into them as well.

Ludwig, looking for a programmer to start prototyping browser technology for Windows, asked Slivka onto the team. At that point, the Internet was just one aspect of the blueprint for Memphis, as the leapfrog upgrade of

Defendant's Exhibit 417 Page 4 of 5


[Page headings How the Web Was Won]
[Page Number] 164
[Page Number] 165

Windows -- the one following Chicago -- was code-named. The whole idea of projecting an upgrade ahead was a new twist for the Windows team. It harkened back to Silverberg's conviction that software development had to happen in incremental iterations rather than one shot only. Besides the Internet, on the Memphis team's plate were things like wireless comminications, game machines, PCs in the home, the eventual merging of Chicago with Windows NT. Permeating the effort philosophically was the Gates vision of Information At Your Fingertips, approaching its fourth anniversary. How would the Windows of three or four years hence continue the IAYF vision? Ludwig thought about the question every day.

After joining Ludwig's team in July, Slivka initially was interested in the notion of indexing all the content on the Internet. It seemed a natural extension of the IAYF metaphor. In order for the unthinkable amount of data on a vast interconnected network to be useful, it would have to be indexed in a way that gave meaningful access to users. Ironically, by that point, the summer of 1994, Slivka had not even gotten a home connection on the Net. He knew next to nothing about the Web. He hadn't seen the Allard or Sinofsky memos. He hadn't attended the Shumway retreat. Of all the eventual architects of Microsoft's Internet presence, Slivka was undoubtedly the last to the starting line. But in terms of producing actual code, Slivka was first out of the blocks.

Ludwig loved this about Slivka. Ben is not a patient fellow, Ludwig would say. When he identified something that needed to be worked on, Slivka was like a woodpecker, tapping, tapping, tapping till he got to the meat of the matter. "He'll come at you every day with ten things you ought to be doing," Ludwig put it. "Some percent you already are doing, he just didn't know about it. Some percentage are just shooting from the hip, he hasn't really thought through. But some percentage are dead on and you should listen to him. I let him have his say, and he tells me how to do my job, and then I throw away the nine things I don't want to hear about. The one thing he says that's accurate, I say, That's a good idea, I'll try to do better on that one."

First Slivka tracked down a Microsoft technician and browbeat him into providing an Internet tap to Slivka's office. As of the summer of 1994, getting an Internet line at Microsoft still was not a trivial procedure, where security concerns about the Net still kept it from being widely accessable. Once he got on the Net and downloaded Mosaic, Slivka spent twelve hours straight surfing. He would get on a home page, then click to a link, then go to another URL, then find a dozen more links. It was revelation after revelation. This was as close as Slivka had seen to an actual manifestation of IAYF in all its original intent.

After his tour of the Web, Slivka did not just feel the world had changed, he set about making sure it had -- at least, his world. He started sending around e-mail, asking questions, communicating with programming teams. He asked Silverberg and Ludwig where the company was on the browser. Shouldn't we be developing something for Chicago? From the standpoint of programming, the browser did not seem to be a monumental challenge. Even if we can't get it in time for the Chicago release, Slivka told the Windows team, we ought to have it ready within a few months afterward. Directed to consult with Barrett, Slivka became even more convinced that Microsoft needed to move more quickly. WWhat he found was pretty bare bones. Barrett had "already decided this was nuts. This is going nowhere, and I don't particularly want to be in an enormous company." To his mind, Microsoft did not get the Net and was not likely to soon. It was time to move on. By August "I'd already made a decision to leave," Barrett recalled. Knowing he was a short-timer, Barrett ignored the Internet project.

Oblivious to Barrett's disenchantment, Slivka spent little time puzzling over the situation. Microsoft would get a browser, he decided, if he had to write every last line of it himself. Slivka's first step was to take a comprehensive look at Mosaic, break it down feature by feature, figure out how the stuff worked, and where Microsoft had the opportunity to improve. What was the competition in the browser space? Who were the players? What were the feature sets? What problems do users encounter with surfing? One of Slivka's first assumptions was that browsing -- at the time still being referred to as "viewing" -- would supersede gopher and ftp. This despite the fact that at the time, gopher and ftp were by far more popular ways of navigating the Net than any of the browser technologies. Our focus should be on the Web, Slivka told Ludwig and Silverberg. That's where our resources should go.

Like Ludwig himself and Silverberg, Slivka was a systems guy, which meant he thought in terms of platforms. How could Microsoft use a new technology to benefit Windows users? How could the company get thousands of software developers to use Microsoft technology? That was the key question to platform guys. At the time, the Memphis team was well aware of parallel efforts to incorporate browsing into other Microsoft products. Pathe had the Internet Assistant project going for Word. Evslin headed the effort to make browsing a part of Microsoft's Exchange e-mail project. No, no, no,

Defendant's Exhibit 417 Page 5 of 5

[Pages 166 and 167 of How the Web Was Won]

the systems guys were saying. Browsing -- viewing, exploring, whatever -- should be a part of Windows. Not that Pathe or Evslin were misguided in wanting to make browsing a key part of their end users' experience. But writing a browser for Word, and another one for Exchange, and yet another one for Windows would waste resources and create a lot of redundant code.

On August 22, in an e-mail time-stamped 5:10 P.M., Slivka notified the Memphis planning team that he had gotten started on the user interface design for what he termed Microsoft's "WWW Explorer" -- there was that word again. Slivka had cataloged the entire Mosaic user interface -- at least as far as http was concerned; ftp and gopher mechanics were still awaiting assessment. To a crack systems programmer like Slivka, Mosiac was a collection of pieces, as its name implied. There was an html piece, a user interface component, a caching element -- caching referring to the process where things like Web pages, or URLs, were stored on the local machine for ready reference by the browser user or the browser itself. Caching made it much easier and faster for the browser to call up previously displayed URLs. Slivka though it was done pretty poorly on Mosaic, and it became one of the WWW Explorer team's top priorities and early triumphs. From his initial analysis, Slivka concluded that the process of Web browsing was pretty similar to network browsing and hard-disk browsing. It was all exploring, he thought at the time. Slivka started a list of what changes and improvements the Windows team could make to Mosaic, but a key design question also needed to be addressed: "At this point, I'm not sure if I want to be TOTALLY INTEGRATED INTO THE CHICAGO EXPLORER, or if we want a separate window for the html viewer." The reference provided another benchmark in Microsoft's plan to blend Windows with the Web. Eventually, Slivka would have it both ways. The html viewer -- browser -- would start off as its own window but gradually, with the release of Internet Explorer 4.0 in September 1997 and Windows 98 the following June, meld with the Windows Explorer.

Slivka's persistent questioning of the browser effort got back to Silverberg. He looked into the situation, found it wanting, and told Barrett he was not happy with the progress he was making. Silverberg was a patient manager as long as progress was evisent. It looked to him as if Barrett did not understand what the browser did and what Microsoft needed from the technology. Barrett was in no frame of mind for second-guessing. By the first week of October, he told his supervisors, "I'm quitting and I've got four weeks of vacation. See you later." Within days Ludwig was paying Slivka a visit. How would you like to be in charge of the browser effort? Ludwig asked. It was an entirely rhetorical question. Slivka did not even bother to ask what happened to Barrett. As it turned out, Barrett took about a month off and then joined Rob Glaser's Internet startup, Progressive Networks, as vice president of software development.

It was fitting that Slivka found himself on the cusp of Microsoft's biggest paradigm shift since DOS-to-Windows. Everything in his upbringing and career path had pointed toward a day when he would tackle something worthy of his talents. Since childhood, Slivka had been self-driven towards programming achievement. One of twin boys born in 1960 in Seattle to a public librarian mother and Seattle Symphony percussionist father, Slivka grew up playing with a variety of electronics. His first-generation Russian father, Meyer, put together a Theremin, a rare electronic musical horn whose "wooo wooo" sound changed tone when one's hands passed over its surface. Meyer also built an oscilloscope and TV set from Heathkit and, in the mid-1970s, put together his own electronic music synthesizer. Assisting him, young Ben got handy with a soldering iron. It was his mother, Enid, however, who introduced Slivka to programming. In the early 1970s she took a course on programming in BASIC, and Slivka got intrigued by what you could do with computer code. He was still a little on the young side to do much on his own, but a seed had been planted. When Hewlett-Packard came out with its programmable pocket calculators, Slivka would go downtown after school, a half-hour bus trip, and program display models for an hour or two at Seattle's leading department store, Frederick & Nelson. The salespeople, amused at what a kid could do and figuring it might attract buyers, were tolerant.

Slivka learned eraly on the value of hard work and independent thinking. His working mother had the two boys helping out almost from the time their younger sister was born. "The poor things never knew what is was like to sit still and have someone wait on them," Enid Slivka recounted. As a result they learned to speak their minds when they wanted something, a trait heartily encouraged by their mother. Enid Slivka had read a book about the Compton family, which produced two university presidents and Nobel Prize-winning physicist Arthur Holly Compton. "The way their [Compton] children were encouraged to investigate things for themselves made quite an impression on me," she said. Ben eventually drew the admiration, not to say awe, of Microsoft coworkers for being able to repeatedly challenge a boss named Bill Gates, and take the return heat without flinching. [Groklaw transcriber's note: square bracketed [Compton] is original.]

Slivka eventually outgrew the calculators and discovered bigger terrain. Near Green Lake in north Seattle, a treasure trove called the Retail Com-

[END OF MARCH 19, 2004 Gordon v. Microsoft Transcript.]

Gordon v. Microsoft Exhibits Published March 24, 2004
21 Exhibits 82 pp.
(Microsoft, Go Corporation, Compaq Corporation: Pen Computing, PenPoint OS, July 1988 to January 1993.)
Nearing completion.
Last Revised 4/29/2004


PLAINTIFF'S EXHIBIT 27 Gordon v. Microsoft   [ Top ]


NONDISCLOSURE STATEMENT

GO Corporation has disclosed or may disclose business or technical information ("GO Information") to you in connection with a proposed business relationship.

In consideration of any disclosure and any negotiations concerning the proposed business relationship you agree as follows:

1. You will hold in confidence and not use or disclose any GO Information except information you can document (a) is in the public domain. (b) was known to you prior to diclosure by GO or (c) was properly disclosed to you by another person without restriction.

2. If you decide not to proceed with the proposed business relationship or if asked by GO, you will promptly return all GO Information and all copies.

3. You will promptly notify GO of any unauthorized release of GO information.

4. You understand that this statement does not obligate GO to disclose any information or negotiate or enter any agreement or relationship.

ACKNOWLEDGED AND AGREED:

Name: (Print) William H. Gates
By: (Signature) William H. Gates
Date: 7/8/[19]88

GO CORPORATION
By [signature illegible]

[STAMPED] KAPLAN DEPOSITION EXHIBIT 3 4/18/[20]02
[STAMPED] KAP01-16
[FAX TIMESTAMP FOOTER] JUL 05 '88 15:51 GO CORPORATION P. 2/2

PLAINTIFF'S EXHIBIT 58 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 58 p. 1 of 3]

GO Corporation [GO logo]
[ADDRESS / TELEPHONE / FAX]

February 10, 1989

Mr. Jeff Raikes
Microsoft Corporation
[ADDRESS]

Dear Jeff:

This letter documents a joint project between Microsoft Corporation of Redmond, WA and GO Corporation of Foster City, CA.

Assumptions:

* Microsoft is interested in exploring potential business opportunities relating to the GO notebook computer. Potential projects include native applications, adaptations of desktop-based Microsoft products, and insuring smooth exchange of data between notebook and desktop applications.

* GO is seeking strategic partners to develop applications for GO's notebook computer.

Goals:

* Establish a working relationship in which technical, marketing and strategic information may be shared between GO and Microsoft for mutual benefit.

* Support Microsoft's explorations of potential opportunities relating to the GO notebook computer.

* Support GO's corporate and product development efforts to establish its notebook computer.

Project Definition:

Microsoft will assign a minimum of one half-time person to work with GO to identify projects of potential interest to Microsoft.

GO will provide temporary office space, access to proprietary design documentation and related information, and technical cooperation to the assigned Microsoft project representative(s).

KAP0155

[STAMPED] Kaplan DEPOSITION EXHIBIT 4 4/18/[20]02

[Plaintiff's Exhibit 58 p. 2 of 3]

GO Corporation
[ADDRESS]

GO will attempt to incorporate design elements and system services into the notebook computer that complement and support Microsoft's efforts.

The project will result in a jointly authored report describing the results of the investigation, This report will be jointly owned, and will be held in confidence by both organizations.

Term:

The project will extend for a period of 3 months beginning February, l5 1989, and may be extended by mutual agreement.

Confidentiality:

In the course of this project, each party may at their sole discretion provide the other party with access to proprietary, trade secret and confidential information ('Information') includinq, but not limited to, design documents, program code, draft documentation, and market research, as required to further the project.

Each party agrees to treat the other's Information with the same care with which they treat their own.

To protect each party's Information each party agrees:

1. To use Information only for the purpose of furthering this joint project.
2. To hold in confidence and not use or disclose the other's Information except to the extent they can document that it (a) is in the public domain, (b) was known to them prior to disclosure (c) was properly disclosed to them by another person without restriction.
3. To label written Infomation as confidential, and to follov up oral disclosures of Infomation with a written confirmation of the disclosure describing the Information within 30 days followinq oral disclosure.
4. That no copies of Information are to be made without the express written consent of the other party.

[STAMPED] KAP0156

[Plaintiff's Exhibit 58 p. 3 of 3]

GO Corporation
[ADDRESS]

5. To return all tangible forms or Information received from the other upon written request.
6. To promptly notify the other party or any unauthorized release of the others Information.

The participation of Microsoft and GO staff in joint design and implementation efforts will not create an interest or ownership on behalf of either party in the other's proprietary, confidential, or trade secret information.
Agreed:
Agreed:
[signature of R. M. Carr]
Robert Carr
Vice President of Software
[signature of Jeffrey S. Raikes]
Jeff Raikes
Date: 2/13/89 Date:
[STAMPED] KAP0157


PLAINTIFF'S EXHIBIT 101 Gordon v. Microsoft   [ Top ]


To: rc
From: jkaplan
Date: May 10, 1989 6:16 PM

My thinking is that while he is here, we should cooperate fully with him. Giving him doc to read is fine, if we can usefully give him a PC with some tools to try out, that's fine to [sic]. In fact, not doing this may seem a bit odd, given that Gary (for example) has been able to develop a sample aps on his machine.

We should set him up in a booth at the other end of the world from Bob Vallone.

I think we should request that he not remove (or copy) documents from the offices from the time being, and explain that this is not a personal criticism but reflects the early state and relative lack of definition of our relationship with MSFT at this time.

Do you think this will be taken in the right spirit? If not, we should consider a specific agreement about what will be done with our docs (like a log of who sees them that we get a copy of, that they by kept under lock and key, and that they not be duplicated?

J

[STAMPED] Kaplan DEPOSITION EXHIBIT 12 4/18/[20]02

PLAINTIFF'S EXHIBIT 115 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 115 p1 of 3]

[handwritten] File: Notebook Computers
[STAMPED] Kaplan DEPOSITION EXHIBIT 15 4/18/[20]02
[STAMPED] X 584359 CONFIDENTIAL

File : c:wzmailmailbox.fld
Messages: .

####################################################### 50
>From dinash Thu Jun 8 15:19:01 1989
To: cindych
Subject: pls print
Date: Thu Jun 08 13:40:44 1989

>From peteh Thu Jun 8 11:47:53 1989
To: dinash
Subject: pls print
Date: Thu Jun 08 11:44:35 1989

>From ralfha Wed Jun 7 14:20:48 1989
To: peteh
Subject: Lloyd's GO Trip Reports
Date: Wed Jun 07 14:19:20 1989

>From lloydfr Thu May 4 18:39:08 1989
To: gregs jeffh ralfhe
Subject: Trip to GO Corporation
Date: Thu May 4 18:36:37 1989

Trip Report - GO Corporation, 5/4/[19]89

(150 lines long)

Yesterday, Kathleen Schoenfelder (soon to be permanent MS employee, now Stanford MBA) and I met with Robert Carr, Jerry Kaplan, and Suye Toenfskoetter (GO director of marketing). Carr was with us the whole day; Kaplan and Sue dropped in occasionally.

WHAT GO IS DOING

They see themselves primarily as systems software developers for a new class of comptuers. Out of the 35 people that work there, approx 25 actually write code (15-17 on systems kind of stuff, 5-6 on apps), and only a few work on the hardware. They subcontract the hardware out to a bunch of companies, and in the future might have closer ties to the Japanese. They haven't figured out if they are going to license their systems software to other computer manufacturers. Hardware is not their competitive advantage, it's the software. They hope to ship everything by summer of '90.

Hardware

It's a 12 mhz CMOS 286 with 2-8 meg static ram, LCD screen, EGA, 640x400 80 DPI, 2 planes. Special pen that interacts with a magnetic field above screen. They can tell how far the pen is from the screen, it's [sic] angle to the screen, and rotation. A second processor deals with the pen interaction. Either a standard modem or fax modem card can be put in the unit (hopefully they'll be able to oput them both on one card). They have a "home base unit " which includes a battery recharger, 3mb floppy, SCSI, parallel, and RS-232 ports, LocalTalk/TOPS connection, keyboard connection. Optional home base stuff is network card to get to Novell or MS-Net, and a hard drive. Active Matrix screen will be option sometime. Standard with 2 meg and base unit will cost around $4000. The only demo we saw was a year old - this July they should have prototypes. It'll weigh 4 1/2 lbs and be about and [sic] inch thick, still kind of bulky.

Software

They're doing their own GUI, object-oriented, protect mode operating system. They seem to have a lot of bright people who have gotten fairly far along with this. Their design appears pretty similar to our object strategy. They have a kernel, class manager, file system, etc, and on top of this they have an application framework. You build apps by subclassing off of the system objects and adding your own stuff on top of that. Carr talked a lot about embedded objects and other merits of GO OS. I agreed that this is all very nice, in fact we're doing the same thing, but just imagine if this was completely compatible with the x million desktop computers in existence. All of this workgroup, networking, and communication would be so much easier. His response was he needed it soon, and specifically to meet the constaints of notebook type computers. Carr wouldn't let me bring home the API documents or their SDK, and I didn't get a chance to actually see the stuff working. But he wants me to spend a few days down there looking it over and talking wit some of their developers. I think this would be very worthwhile for me to do.

The apps they are doing initially are: notebook manager, notetaker/simple wordprocessor, fax/markup up, e-mail, forms manager, rolodex, calendar/To Do. They are likely to add small apps and change their emphasis as they find key customers who talk about buying several thousand machines. A lot o fthe objects they are using in those apps will be part of the system so that other ISV's will be able to leverage off their work.

They say their handwriting stuff is really good and flexible depending on what each app wants. In a workdprocessor, the apps tells the HW sftwr [sic] to try and map the input to words in its dictionary. A spreadsheet would constrain the input to numbers, formula's [sic] and names. Fields on a date entry form might ask only for date input. These levels make it easier and faster for the HW recognition. They have two modes: one is gesture recognition and the other is interpret what the user is writing. Again, I haven't yet seen the API or if this all actually works.

ANALYSIS

Carr really wants us to write apps for their machine. I pointed out the many difficulties for us to do this - namely a whole new platform. I said I thought Windows with some modifications and enhancements would be ideal for this sort of computer. Essentially he's writing an operating system from scratch of which 90-95% of the functionality can be found in Windows or PW. Why is he doing all of this work when we have already done it? Mostly because he wanted to design something from the ground up which took into consideration the following: a pen being the primary interface, everything is in ram, can't crash the system, have to conserve battery power, and he needs a whole new class of apps. He figures that if people are going to have to write new apps, what's the big deal switching to a new platform? And since it's all object oriented, it's much faster to develop these apps. He certainly didn't convince me that

[PAGE FOOTER] C:TMPDH002595. Thu Jun 08 15:05:14 1989

[Plaintiff's Exhibit 115 p2 of 3]

creating a whole new, non-compatible OS was the right way to address their concerns.

I really doubt they are going to radically change the direction of their company and use Windows or PH as their OS. But they are interested in co-existing nicely with current and future "desktop" operating systems and apps. This is where we can help them out a lot, mostly by exchanging information. Carr sees Microsoft as being potentially their biggest competitor, but he is still very interested in working closely with us on many different fronts.

WHAT CARR WANTS

Systems
(1) Info on our Attributed File System. Should I send him this documentation? Urgent.

(2) he needs a command/control language for his operating system. I mentioned BASIC and he's very interested. Should we consider licensing him some version of our interpreter, maybe OO BASIC, maybe EB? Urgent.

(3) More info on protect mode windows. He still doesn't have a copy.

(4) Info on our object oriented framework. Persistant objects, OOFS. Do we want to show him what we're doing? In what detail? Would it help him be more compatible with our objects?

(5) Info on C++. They use C 5.11 for all of their development, and in their SDK.

Apps

(6) Email - he wants to talk to us, see how we're doing it and get some direction. If we have an engine we'd be willing to license, he's interested.

(7) Microsoft developing apps for GO computer. His idea is that we would give them some money and they would find some developers (or we would send some down) then the developers could start wok on an app of our choice at GO. Carr was pushing for a wordprocessor. (Since lawyers would use this a lot for marking up documents, this would be a good entre into the legal market, he said.) The goal would be to produce a demo, and then decide on the next step. He said one of our main benefits of this would be learning what is involved in writing these types of apps. An interesting idea, but I don't think it makes too much sense for us to do real soon, if ever. We should keep talking to other vendors, and gradually figure out what approach to take.

ACTION ITEMS

  • Resolve Issues (1) and (2) ASAP (Attrib File Sys, BASIC)
  • develop an idea of how much we want to cooperate with GO
  • decide whether I should go down there and spend a few days pouring [sic] over the technical documents, playing with their SDK, and talking to their programmers.
>From lloydfr Wed May 17 19:09:32 1989
To: billg jeffr leno mikemap raleighr sherryr tandyt
Subject: 2nd GO Trip Report
Cc: gregs jeff raitha russw
Date: Wed May 17 19:06:29 1989

GO Trip Report - May 15, 16

(100 lines long)

Purpose of trip:
(1) For me to learn in more detail what GO is developing. In particular (a) their development environment and (b) their notebook metaphor. We can take this information into consideration when putting Windows on similar types of machines.

(2) Discuss GO's interest in licensing our email and BASIC technology.

(3) Continue to tell GO that we believe a modified version of Windows is the best environment for this type of computer.

Results:

(1)
(a) GO Development Environment

This is really bad. Not only do they have a totally non-standard OS, but their way of doing object oriented programming is inefficient and hard to use. The reason for this is they do all of their OO stuff in C, not C++. And since everything in their system is an object, they force to [sic] programmer to adhere to their kludgy model. The way you apply a method to an object is to call the function ObjectCall() with three parameters: (1) the message # which the programmer defined in a .h file (2) a handle to the object (3) a pointer to a block which contains the params for that specific message. The object has to have its own message processing proc which does a switch on the message #, and then calls the relevant function. If you fall through the switch statement then you do a[n] ObjectCallAncestor() with the same params and pass the message up the hierarchy. To create virtual functions, you simply stop propagating the message. To access instance variables somewhere up in the hierarchy, you have to do and [sic] ObjectRead() call, which I imagine is as inefficient as the method dispatch.

A lot of errors that would be caught by C++ at compile time (scoping, typing) will only be found at runtime, and this makes it MUCH more difficult to develop apps. The programmer also has to manually do a ton of things that C++ does automatically which makes the code much longer, more complex, and therefore more prone to have errors. GO claims that in this OO environment, more code will get re-used. This might be true, but a lot of sacrifices are being made.

So why did they do their OO stuff this way? I think mostly because their [sic] isn't a really good C++ compiler in existence. They wanted to build on a very solid base, MS C, and put their objects on top of that. I don't think they realize the dangers of being only part way object oriented.

(b) Notebook metaphor

I think this stuff will work out fine. Essentially it is a

[STAMPED] X 584360 CONFIDENTIAL

C:TMPD[H?M?]002595 Thu Jan 08 15:05:14 1989

[Plaintiff's Exhibit 115 p3 of 3]

[page number] 5

specialized version of a shell. They have tabs on the side of the screen which represent folders, and then a table of contents and index. You can place GoTo buttons and PostIt Notes any place in the notebook. They could do all of this on top of Windows and I think it would be very good.

One thing I might mention is that I never actually saw anything running on a computer, I just buried myself in stacks of documents for a few days.

(2) Email and BASIC

As far as email goes, we don't have anything to offer them. In 6-12 months when slingshot is further along we might talk about possibilities. If GO sticks with their OS, they would only be able to license some very low-level code from us.

They are very interested in licensing EB from us. They have a table object which does indexing and other ISAM-like tasks. They are going to develop a forms app for people who need to do data entry outside of an office environment. And they need a language to snap the two pieces together. Sounds a lot like Visual BASIC to me. They also want to use EB as their "batch" language for the OS. This sounds like a mix between Ruby and VB. So, the question for us to ponder is whether or not we want to license them EB.

(3) GO and Windows

The whole time I was down there, I was politely telling them that it was very unlikely Microsoft would develop apps for their machine because of their non-standard OS and not so slick OO environment. Finally, Carr said he wanted to come up and discuss technical issues with the Win 3.0 people, and explore the possibilities. E.g. what kind of modifications to Windows would be necessary? This kind of caught me by surprise, but I said we'd be happy to do that. If anyone wants to see him while he's up here, let me know.

Other Info:

The one key piece we need in order to put Windows on flat computers made by other manufacturers is the handwriting recognition software. The GO people wouldn't let me get near any of their stuff (with good reason). So, while I was down there, I spent a night in the Stanford library photocopying six IBM research reports on this topic. These reports have fantastic bibliographies and references to groups (at universities and corporations) who have done work in this area. Eventually, we will have to decide if we want to acquire this technology or build it ourselves. I will be looking further into this subject.

[page number] 6

C:TMPD[W?H?M?]2595. Thu Jul 08:15:05:14 1989

[STAMPED] X 584361 CONFIDENTIAL


PLAINTIFF'S EXHIBIT 117 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 117 page 1 of 2]

[STAMPED] Kaplan EXHIBIT 88 5.8.[20]02

GO Corporation Confidential Mar 20, 1991
MEMORANDUM
TO: File
FROM: Robert Carr
DATE: June 16, 1989
SUBJECT: Draft Response to Microsoft Pitch for Cooperation
COPY: Jerry K
SAVED AS: \go\thirdpar\microsft\890616.fw3

Robert will call Jeff Raikes and say: Jeff, here's the range of what we're thinking. I want to give it to you now, within 1 week of our meeting last week.

FIRST OF ALL:

We agree to the principle that it's a shame for both parties to duplicate efforts or fragment the marketplace unnecessarily.

We've really thought through the issues involved, and concluded:

- We believe we've got the right approach for NB computers. There are serious technical tradeoffs in the rival approaches [underline is handwritten] of GOOSE and Win H and we feel there's significant market opportunity for a GOOSE solution.

- it doesn't make sense for us to use Win H in place of GOOSE

- they might want to consider usign the consierable efforts & expertise we've already invested

IF MS DOES NOT PROCEED WITH WIN H:

There are circumstances under which we'd be open to MS being the licensing vehicle to 3rd party HW vendors.

We would split royalties w/ MS. Under this scenario MS would not compete with GOOSE through a Win H product.

MS would make material commitments towards their level of support through evangelism, 3rd party licensing efforts, tie-in development efforts etc.

An equity investment in GO would be an additional possibility.

Page 1

KAP0105

[Plaintiff's Exhibit 117 page 2 of 2]

[STAMPED] KAP0106
[Page Header] GO Corporation Confidential  Mar 20, 1991

IF MS INSISTS ON PROCEDING WITH WIN H:

If you do want to do your own work and system in this area, then we certainly still want to cooperate on:

file format exchange & other tie ins.

we're exploring MS Press publishing our reference books.

you putting apps on our machine (but we're hesitant to have technical exchanqe with you on this until you assign active development teams for GO specific versions or until our technical design is public knowledge anyways). Once GOOSE is public, then we'll actively support you as much as you want.

Handwriting Recognition licensing from GO to MS could be a topic of discussion. We'd probably look for royalty streams and healthy advance on royalties in return.

If Jeff wants a written document: we'll be out of town next week, after that we need some time and discussions with board members to work up a more specific proposal or to capture anything in writing.
[Page footer] Page 2


PLAINTIFF'S EXHIBIT 293 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 293 p. 1 of 9]

[STAMPED] X 188680 CONFIDENTIAL
[STAMPED] Raikes DEPOSITION EXHIBIT 48 10/18/[20]01
[STAMPED] X188680
stylus input) to those features stored in the computer. We [wo]uld add to this feature set to enable recognition of lower [ca]se (e.g. vertical height of the chars), but this would probably decrease the accuracy rate. Jeff said he thinks he knows how to do text recognition using ideas somewhat related to neural nets. This would take several years of development.

Although Grid's handwriting s/w might not be as full fledged as GO's, there would be many fewer conflicts of interest in working closely with Grid instead of GO. Grid is in the business of selling hardware, unlike GO who sees themselves more like Apple selling both hardware and the system software.

Grid wants to come up here around the first week in August to describe their handwriting s/w in gruesome detail and discuss other ways we might work with them on this project. They will also give us a demo of their machine.

- Lloyd

>From jeffr Mon May 28 16:14:19 1990
To: billg mikemap
Subject: FYI-GO competition
Date: Mon May 28 16:14:18 1990
Mail-Flags: 0000

>From luist Mon May 28 15:30:01 1990
To: jeffr rice
Subject: Go and Momenta(?)
Date: Mon May 28 15:39:22 1990

I met a friend that just came from talking to GO and Momenta. Though she was under non-disclosure, she said a couple interesting things:

* If we still had any doubts: GO is trying to licence their operating system. Her words: "they want competition in the hardware side...they want as many people as possible to build the machines"

* About price: "it is a neat idea, but i have a hard time seeing many people paying so much money for it..." It sounds like it will be on the expensive side for a Grid like machine. For some reason she came out feeling they are equivalent (beyond the handwriting recognition side...), which is great news. GO will probably have a hard time positioning the machines (happy, Pradeep?).

* About the 'Momenta' machine: "well, its [sic] so far into the future that it probably won't be very interesting" what is the momenta company about?
:
luis.

>From jeffr Wed Jun 6 17:47:58 1990
To: billg jonll mikemap
Subject: FYI-Momenta on GO
Date: Wed Jun 6 17:47:57 1990
Mail-Flags: 0000

>From pradeeps Wed Jun 6 16:04:07 1990
To: jeffr
Subject: Info on GO from Momenta meeting

[Plaintiff's Exhibit 293 p. 2 of 9]

[STAMPED] X188681 CONFIDENTIAL
[STAMPED] 188681
Cc: lloydfr pradeeps
Date: Wed Jun 6 16:04:05 1990

Jeff - you may want to relay some of this to Billg:

Momenta claimed that IBM is giving $7m to GO; of which $700k is equity; $6.3m is pre-paid royalty. This originated as a single customer deal - initiated because of an $80m deal with an insurance industry IBM customer.

Other perceptions from Momenta about GO: GO is a software company (i.e [sic] looking for OEM licencees). IBM is not restricting GO's ability to licence the GO OS to other OEM's. That there is a third unknown partner involved with IBM and GO. (My speculation: this could be Slate (no news there..) OR it could be Toshiba who is already supplying GO the integrated tablet and other h/w components. Toshiba could be a GO OS licencee..)



>From jeffr Wed Jun 27 07:37:08 1990
To: billg mikemap
Subject; FYI-Intel and GO-ByronB's comments
Date: Wed Jun 27 07:37:08 1990
Mail flags: 0000

>From byronb Tue Jun 26 18:29:49 1990
To: carls jeffr rice
Subject: A few personal impressions from the mtg with Intel
Date: Tue Jun 26 18:28:12 1990

Taken a face value, the meeting resolved around the issue that they are [a] hardware company want to purchase stylus/tablet technology CHEAPLY in order that they may manufacture these machines and we are a software company worried about the confusion caused to out OEMs by an apparent endorsement of GO software by intel.

It may be telling that first major issue brought up was Mike's apparent fondness for the GO folder paradigm. It makes me quite skeptical that their only focus is hardware. I assume their real worry is that they want to make sure that stylus based machines use intel silcon [sic] (they said so) and they don't want to dependent on Microsoft for the software (they weren't willing to say this). They brought up the telling and accurate analogy of Microsoft hypothetically investing in MIPS. So they could be hedging their bets on stylus software.

They seemed willing to do some sort of 'damage control' for the GO investment. They proposed an joint announcement by Intel/Microsoft saying that Intel is investing in hardware technology owned by GO and that Intel supports supports WIN-H as tha OS of choice for stylus based computers. They even proposed bringing in an OEM willing to announce an intel x86 win-h machine. This offer seemed insincere considering they want to sign with GO next week and that we told them we were not willing to announce Win-H until October.

While they would love to hear about what leads Lloyd can give them about other hardware technologies sources. It will not change their mind on GO; they will just view them as additional sources of technologies.

We did not convince them to not invest in GO. We probably convinced [th]em that 1) the investment will give bad signals to OEMS 2) We are

[Plaintiff's Exhibit 293 p. 3 of 9]

[STAMPED] X188682 CONFIDENTIAL
[STAMPED] X188682

a serious contenter [sic] and will be heavy competitor to GO 3) That our [ar]chitecture (open, based on Windows) and positioning are sound.

What should we do? From my humble view point in the trenches ... Reiterate through Bill our concern about confusing the Market and OEMs and work with them on 'damage control' to see if they are serious on this. My read is the deal is as much as done with GO and from their point of view today's meeting was about learning more about what is going on in the notebook computer world and not about reconsidering Intel's investment.

Some impressions from the demo ...

When one of the intel guys tried the tablet (who admittedly had bad writing), I was embarrassed by the results, but he gave the comment he thought it was quite good. I got some genuine Wows! with the cursive demo (4 out of 6 words correct). This tells me one of three things: 1) Go is not as far along as we think. 2) They have not seen GO's handwriting yet (unlikely) or 3) They were being polite.



>From jeffr Wed Jun 27 07:34:19 1990
To: billg byronb carls lloydfr marline pradeeps
Subject: Intel Stylus Technology/GO Corp.
Cc: mikemap nathanm paulma samf steveb
Date: Wed Jun 27 07:34:16 1990
Mail-Flags: 0000

Addition to CarlS's mail:

They didn't say the investment in GO would be for hardware only. In fact, [Pa]ul Otellini specifically said that they might "have the GO operating [en]vironment on the shelf, and if an OEM (AT&T was the hypothetical OEM used) wants to buy it, they would sell it." I pointed out how potentially confusing that was, and Michael Aymar jumped in with a comment about how we did software for the Mac.



>From Carls Tue Jun 26 18:46:38 1990
To: billg byronb jeffr lloydfr marline pradeeps
Cc: mikemap nathanm paulma samf steveb
Subject: Intel Stylus Technology/GO Corp.
Date: Tue Jun 26 18:44:52 1990
Mail-Flags: 0000

We met today with the group of people at Intel that is pushing an investment in Go Corp. We attempted to dissuade them from that investment by seeking to find ways to meet their objectives without this investment and by explaining why we thought the investment would be a bad idea. In particular, we presented the Windows H project to them. We didn't suceed in convincing them not to invest, and they did not appear very open minded. One appropriate followup is probably for us to write a letter to Grove with a followup phone call by Bill

Intel
-----
Paul Otellini, VP, Assistant to the President
Mike Aymar, VP MCG, GM of SMD I
Tom Galvin, Finance Manager, Business development MCG
Mike Bruk, Microsoft Liaison

[Plaintiff's Exhibit 293 p. 4 of 9]

[STAMPED] X188683 CONFIDENTIAL
[STAMPED] X188683

Microsoft: Jeffr, Pradeeps, Lloydfr, Marline, ByronB, Carls

Our objectives for the meeting were to:

  1. present and demonstrate out WindowsH project to show that we have good stylus/handwriting technology; that it integrates with DOS & Windows; and that we are working with leading OEMs;
  2. to attempt to understand Intel's real objectives behind making such an investment; and
  3. attempt to dissuade them from the investment.

Intel believes that stylus computers represent a very important emerging market; they want to be sure that Intel participates in that market both by having Intel chips used in those computers as well as as a manufacturer of such computers. Go is not the only investment Intel has been pursuing in this area--there has already been one that has fallen through and they will consider more. I inferred from listening to them that Aymar, Galvin (who works for Harold Hughs), and Avram Miller (not present) are proponents of the stylus market oportunity, and are trying to get Intel into it.

We had told Intel that we think an investment in Go Corp would be a bad idea. These guys wanted to operate under the assumption that Intel would invest in Go, and how could we minimize or mitigate any damage--they were not very open minded. This is understandable since they have been pushing the Go investment inside Intel and see us as an obstacle toward their objective. They had been told to come and see what we are developing and if it changes the rationale for investing in Go, but these guys clearly want to make the investment.

[L]loyd, Pradeep and Byron gave a good presentation of what we are trying [t]o achieve and how we will go about doing it. We covered [h]ow we are building on top of standard Windows and standard Windows apps, and that we are working with numerous OEMs. Byron's presentation on our character recognition software was particularly good--I think they were impressed that we are pushing the state of the art in this area.

We attempted to understand what their objectives were in making the investment. They gave the following:

  1. They want to learn about the stylus market and technology
  2. They want to license hardware technology for use by their systems division to make stylus computers to sell on an OEM basis
  3. They want to be able to design silicon (chip sets) that OEMs will use in stylus computers; this investment will let them better understand what silicon to design
  4. They want to make sure that the x86 is at the heart of stylus computers in general and the Go system in particular in case Go is successful.

It is not clear what technology it is that they hope to get from Go. They said something about a squishy tablet. When we asked them to list the technology they wanted, the list was:

  • display
  • stylus interface
  • small form factor technology
  • integrated tablets

And a key point they made is that they want access to this technology without having to give up any Intel technology (e.g. chip licenses). [I?] don't know enough about Go--someone tell me if I am wrong--but it

[Plaintiff's Exhibit 293 p. 5 of 9]

[STAMPED] X188864 CONFIDENTIAL
[STAMPED] X188684

seems unlikely to me that Go has such great hardware technology in [t]hese areas--it sounds like stuff you would get from Japan. We [vo]lunteered that we could help Intel to get access to the technology from someone else, and they asked us to follow up on that.

This is what they said. Unclear what the real motives are. The investment in Go would give them a license to Go's design as well as Go's software, including sublicensing rights.

As I mentioned they weren't really interested in finding alternative ways to get stylus/tablet technology--they wanted to take the approach of saying: assume that Intel does invest in Go, how can we mitigate the impact? They say they are willing to position their investment as strictly a hardware technology investment, and issue an endorsment of Windows H. We did not pursue this discussion.

The points we made are:

  • Windows H builds on top of DOS and Windows and will run existing and new apps and be open to everyone.
  • Establishing a new OS is hard and unlikely.
  • Establishing a new hardware and OS standard divides our market, confuses customers and ISVs and is unnecessary
  • An Intel investment in Go will send a strong message to OEMs and ISVs and cause confusion
  • We are already working with key OEMs, and they will not be unhappy with this Intel activity
  • It is an anti-Microsoft move because it competes with our OSs and it weakens the Intel platform standard.
  • Finally--we'll help you to get the technology; we will deliver technology and OEMs making stylus computers using x86; you don't need to make an investment.

[Th]eir response is: they don't care about Go's operating system and whether or not it succeeds; they will work to reduce any impact on OEMs and ISVs by positioning it as a hardware-only investment; it isn't an OS thing so it's not anti-Microsoft; the Go platform can be made to run DOS; and this is a quick way to get technology.

We weren't making any progress, so we ended the discussion. If they are sincere about what their objectives are, then if we show them alternative sources for the technology, we ought to be able to persuade them from the investment.

I fear that either they have further motives, or this is another case where Intel has just decided to do "something", which means put money into something they think helps x86. Without considering how OEMs are going to feel, how we are going to feel, or whether it really makes any sense at all. Just throw money at it. Sounds like the Bell Technology deal. Or some of Intel's Unix work.

Aymar asked a lot of questions about application software, and how users locate documents and files and things; how they navigate through "folders" etc. I think he thinks of Windows as being too complicated. I don't know if he has seen some demos of Go software that look easier to use, but that is my impression. That makes me think they must also be interested in Go's software and just not telling us. I would have to assume that reducing their dependence on Microsoft is also in their minds.

[Th]ere are 2 next steps:

[Plaintiff's Exhibit 293 p. 6 of 9]

[STAMPED] X 188685 CONFIDENTIAL
[STAMPED] X188685
  1. See if we have some stylus/tablet technology sources we can introduce them to. We need to decide if we want to do this, since they could then try to sell Go's technology to them, or otherwise interfere in our work with OEMs. If we want to do this, we need to get this moving by the end of this week. Billg--give your input on email. If yes, Lloyd/Pradeep own the followup to identify tech. sources.
  2. Write a letter from Billg to Grove and follow up with a phone call. I'll draft a letter. Send it out on Email for comment tomorrow.



>From tomos Mon Jul 16 04:52:30 1990
To: lloydfr
Subject: Summary report of Intel Team in Japan
Cc: billg carls jeffr makon mikemap pradeeps samf tomos
Date: Mon Jul 16 20:42:04 1990
Mail-Flags: 0000

Subject: Summary report of Intel Team in Japan.

Followings are summary report for the above mentioned team's activities in Japan. I coordinated meetings with Kyocera, Wacom and Seiko Denshi for their survey of stylus technology . Prior to those meetings. I had a preliminary discussion with them on 7/11 at the hotel to get their objectives of survey in Japan.

1. Team members and visiting schedule.
Mike Moon Consultant, Strategic Technology
OEM Microcomputer Platform Division
Kathryn Harrignton [sic?]
Technology Planning OS & APPs
OEM Microcomputer Platform Division
Tom Gablin Finance Manager, Business Development
Microcomputer Component Group

Mike is the boss of the team and he knows the latest stylus technology and trend. He is the key person to create a report to Grove. He knows the fact that GO is using Wacom's stylus pad for their product under development when I asked the question. Kathryn is software oriented person who want [sic] to know hardware from Hand Writing Recognition software(HWR) point of view. She made several questions related required functions for better recognition rate and she showed much interest in Sony Palmtop when i saw it to her. She might be a key person for Win-H discusion scheduled in Redmond on 7/17. Tom has his interest in possibility to expand market of Intel's microprocessor through stylus computer including development of co-processor for HWR. >From these point of view, Intel's team will create their report from several point [sic] of view not from one point of stylus technology.

2. Objectives of their survey in Japan.

Mike indicated their objectives of survey in the preliminary meeting.

Intel has interest in stylus computer itself as the possible new computer in the future which will expand market opportunity of Intel's microcomputer. They want to establish the platform of microcomputer in this product segment same as they succeeded in PC. >From this point of view, they want to know following points to be reported to Grove prior to the meeting on 7/20.

[Plaintiff's Exhibit 293 p. 7 of 9]

[STAMPED] X 188686 CONFIDENTIAL
[STAMPED] X188686

I will finish this report now. If you have questions, please call me [a]t the office, I will be the office 9:00-9:30AM Japan time then will be out of office.

Regards

Tom



>From johnsa Tue Jul 17 12:07:09 1990
To: billg jonl lloydfr mikehal mikemap steveb tonya
Subject: ibm/go announcement
Date: Tue Jul 17 12:05:13 1990
Mail-Flags: 0000

lee reiswig just called to inform me that there will be an announcement today that ibm and go are engaged in a joint project to pilot certain pen based applications. if the pilot works out it could lead to jointly developed porducts [sic].  ibm is licensing certain go technology.

the Q & As will attempt to address the DOS issues as follows:
Q. Why endorse a new platform to support notebook computers as opposed to enhancing existing platforms?
A. the go platform has been optomized [sic] to make it natural for the user to write with a pen. it will be easy to exchange data betweeen the go platform and those platforms that run on the desktop.
   
Q. Will the go applications run on dos or os/2?
A. No.

He will send us a copy of the press release.



[>F]rom jeffr Tue Jul 17 22:20:45 1990
To: mikemap
Subject: Go ancmnt - GRID's questions..
Date: Tue Jul 17 22:20:44 1990
Mail-Flags: 0000

>From pradeeps Tue Jul 17 14:54:22 1990
To: jeffr lloydfr
Subject: Go ancmnt - GRID's questions..
Cc: billg
Date: Tue Jul 17 14:54:20 1990

I just spoke to Alan Lefkof (Pres, Grid Systems). GO is announcing their machine tomorow. (I don't know if this is a separate, but related event to the IBM press release described in Johnsa's email). Alan wanted to know what was going to be our response and what should he say about Win-H.

He didn't have much data on the announcement, except that IBM will endorse and licence the s/w (not h/w) and that three ISV's will be there: Slate, Pensoft (a 2 man flaky operation) and either Lotus or Borland. We think it's Borland. (Dan Bricklin is coming West this week for the announcement, I guess, not for a board meeting).

Aln Lefkof plans to say that Grid has a large account focus which requires open architectures and adherence to standards. MS-DOS is the standard today. As new standrads [sic] emerge, Grid will look at them. He does not plan to say anything about Win-H unless we decide that we are going to. Even then, he is concerned that we would be forced to talk about unannounced harware [sic] if he said that Grid [wa]s making a machine which will run Win-H.

[Plaintiff's Exhibit 293 p. 8 of 9]

I've told him that I will talk to Jeffr and get back to him with our party [li]ne.



>From jeffr Tue Jul 17 21:59:19 1990
To: billg mikehal mikemap rice sarahch steveb w-connib w-pamed
Subject: FYI-GO says
Date: Tue Jul 2l 21:59:16 1990
Mail-Flags: 0000
>From mcimail?TO:__Gregory_Stikeleather#___EMS:_MCI_MAIL#__MBX:_0004262752 Tue
To: JeffR
Subject: GO Press Briefing
Date: Tue Jul 17, 1990 1:16pm EST

From: Gregory Stikeleather
   EMS: MCI Mail
   MBX: 0004262752
Handling: LETTER
Message-Id: 21900717231512/0004262752NB1EM

Dear Jeff,

Congratulations on your promotion to Vice President of Office Systems. It is certainly clear outside of Microsoft the key role you have played in building the applications Division, and we here at GO are pleased to see you acknowledged for your work.

Everything is proceeding well here; in fact, we are conducting a [pr]ess briefing this week that outlines part of our strategy. We [wa]nted you to hear this directly from us. These are our four [m]ajor points at the briefing:

  1. GO is developing a new operating evironment designed for pen-based computers.
  2. This environment is available for license to computer manufacturers wishing to enter the market.
  3. We have completed an agreement with our first licensee, IBM.
  4. There is already substantial support from independent software companies, VARs, and corporate customers.

I am sending you press releases that you should receive tomorrow.

Jerry sends his regards.

Sincerely,

Greg Stikeleather
Manager of Advocacy
[STAMPED] X 188687 CONFIDENTIAL
[STAMPED] X188687



>From lloydfr Wed Jul 18 23:15:50 1990
To: billg carls jeffr johnsa mikehal mikemap rice sarahch steveb tonya w-connib w-pamed

[Plaintiff's Exhibit 293 p. 9 of 9]

Subject: GO News
Date: Wed Jul 18 23:13:12 1990
[Ma]il-Flags: 8000

UPf 07/18 2304 IBM licenses pen based computer technology

By BRIAN ROONEY
UPI Business Writer

SAN FRANCISCO (UPI) -- IBM Corp. gave an important endorsement Wednesday to an emerging technology that allows computer users to enter data with pen-like writing devices.

The nation's largest computer company announced it was the first licensee of a pen based computer operating system developed by Go Corp. of Foster City, Calif., a 3-year-old startup company.

IBM said it intended to create computers operated by pens on a flat screen. The computers will be able to easily exchange information with other IBM and IBM-compatible personal computers.

Insurance examiners, architects and other mobile professionals could use the portable machines on the job and later transfer the information into their primary computers.

The technology "will make using a computer as easy as writing on a piece of paper," said James Cannavino, IBM vice president and general manager of its Personal Systems Division.

Cannavino said it was too early to tell when IBM would introduce its first pen-operated computers or how much they would cost.

Because of IBM's size in the computer industry, its endorsement could make Go's technology the standard for pen-based computers, just as IBM's licensing of Microsoft Corp.'s DOS operating system made it the standard for personal computers in the 1980s.

Go Corp. said it would make its technology available to any manufacturer that wanted to license it. IBM also invited other manufacturers to adopt the Go system.

Software developers Lotus Development Corp., Borland International, Wordperfect Corp., Slate Corp. and PenSoft Corp. already have announced plans to develop software applications for Go's operating system.

Companies that are developing their own pen-based machines adopted a wait-and-see approach to Go's system.

"I think IBM's endorsement of the whole stylus-based computing market is nothing but good news for people building these types of devices," Rizzo said. "It is way too soon to make any kind of rational statement about whether it will be a standard."

Grid Systems Corp., the first U.S. company to introduce a pen-operated computer last fall, said it would let the marketplace set the standard but would consider licensing Go's technology if necessary.

[La]st page !
[STAMPED] X 188688 CONFIDENTIAL


PLAINTIFF'S EXHIBIT 349 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 349 p. 1 of 3]

112
>From gregw Tue Jul 31 23:56:40 1990
To: billg darrylr jeffr mikemap
Cc: bradsi jabeb lloydfr tonyw
Subject: Re: GO threat
Date: Tue Jul 31 23:42:48 1990

The L&E stuff was not meant to be a real architecture. Our applications could not respond to a real architecture. The AppDT work will form the basis of a real OO architecture - a robust extensible data model which works with the AFX view models.

Our instances can't be viewed as containers of information today. This means that it is very difficult to implement index and content querying. The system would be forced into understanding file formats (we know this is not workable).

(the remainder is long)

The GO machine brings home the following point in a big way.

Once we have implemented enough interesting data types and viewers using [here a "?" appears on a line by itself, apparently a glitch] our OO frameworks and interfaces, there is no need for DOS and Windows as we know it. Instead, the file system can be replaced by a simple memory manager with a backing store to yield persistence. The notion of processes and applications disappears replaced by a single address space with concurrent threads of activity. On the notebook, detached from [the] rest of the world, the security of separate processes is unnecessary. There is still a need for concurrency controlled resource management (memory and screen real-estate).

Why bother with DOS apps or Windows apps as we know them, the DOS apps don't interoperate and the Windows apps are not much better.

These apps are easier to write - no file formats and I/O (only in-memory storage), few format conversions (enough to support content and queries), natural container-containee relationships, garbage collection, objects have well defined behavior (implement a set of protocols). Objects that can be queried support the content protocols. The system can enumerate all objects. If it makes sense to have a container which knows about all instances of a particular type, this is easy to implement and install.

The GO UI is probably the least interesting part of the product from a technical perspective. Like Hypercard and some aspects of Toolbook, it will show how far graphics art can take you. Of all of our products Windows 3.0 is perhaps the best, but it doesn't come close to these other examples.
[here a "?" appears on a line by itself, apparently a glitch]

GO is scary but they are a small player attaching themselves to a limited hardware platform. Their distributed machine plans are interesting and very focused.

In our business there are the following things that are important -

[STAMPED] EXHIBIT 35 [indistinct due to fax or repeated photocopying]
[STAMPED] X 531240 CONFIDENTIAL

[Plaintiff's Exhibit 349 p. 2 of 3]


1. ownership of the information type implementation
2. location of the information
3. diversity of information types
4. end-user ability to integrate information into new components

We need to be taking control of these 4 things with our system and application strategies. GO (or any new platform) is going to have a hard time addressing these 3 issues. They are completely dependent on making the hardware platform compelling soon after the initial introductions. We need to understand what we think are the compelling features and have a response in the form of product and strategy. We won't be able to get a product (an OEM to support us) until we have a compelling strategy to sell.

? [sic]
The OO architected system is the key part of that strategy (we need to [sic?] same pitch to go against New Wave except forget the NT-OS/2 heavy duty features for notebook computers). We need to be able to demonstrate that handwriting does not require new UI concepts and looks by making existing apps work with minimal changes. Most of the apps that people say that notebook computers need are keyboard apps that we are missing today. Other apps like the math equation app are just brain-dead - what ever their implementation, it is unlikely that it can be effectively reused at a low level in other places in the system - high level reuse is easier but the right application contexts need to be found.

What do we have going for us -

1. handwriting is neat but not as reliable as  a keyboard for entry
2. handwriting computers will have keyboard options
3. with a keyboard DOS apps can run
4. handwriting / pen interface techniques integrate smoothly into Windows apps
5. information is naturally exchanged with the primary location (no unreliable format converters - using the same application)
6. diversity of applications for our environments
7. long term strategy that makes sense on the three interesting machine environments - notebook, workstation and server. Keep the picture simple. The GO solution is weak on interoperability with the
?[SIC]
workstation and server.
8. new hardware fits into the big picture of office work - take home and travel (take the information with you)
9. huge ISV support for Windows apps with huge base (can evangelize when software and hardware are ready - need it soon)
10. ... (we have some more going for us - left to the reader as an exercise)

What are we missing -

1. low-end personal organizer apps (windows desktop needs these also)
2. low-end information types (we're dieing [sic] for windows works 2.0)
3. OEM hardware (may be we have it)
4. some of the below

What do we have going against us -

[STAMPED]  X 531241 CONFIDENTIAL

[Plaintiff's Exhibit 349 p. 3 of 3]


1. higher cost due to larger memory requirements
(need to figure out if this is offset by lower software costs - we have  a2 machine / 1 user licensing problem)
2. compelling features of GO UI and software
3. very focused competitors - we need a good economic model for their businesses - the financial bootstrapping process for them is complicated and could be interrupted by successfully using our advantages above
? [SIC]
4. confusing endorsements by OEMs like IBM
5. ... (there must be more)

The bottom-line is that we have a compelling alternative to the GO machine. We are having real difficulties in articulating it.

Enough said by me - use it as you like - I have other things to do.

[BELOW EMAIL IS QUOTED IN THE ABOVE EMAIL]
>From billg Tue Jul 31 21:39:11 1990
Subject: GO threat
Date: Tue Jul 31 21:36:15 1990

In reviewing some sketchy stuff on the GO machine it is clear that the threat posed by GO is as much an integration threat as a handwriting threat. By using an object oriented approach they allow for searching, hypertext linking, and index and table of contets across all data types. They allow new objects to be added easily in their framework. Their shell is quite visual -- with foler [sic, "folder"] tabs and pages. We should try and learn more about it. Meanwhile it is time for our L&E stuff to deal with linking and sorting. I can't believe we don't have this as part of our architecture when a real architecture would al [sic] allow for these things. Our handwriting group should write up what they know. ACtually [sic] the esther dyson article does a good job explaining what should be done
?[sic]
(GO doesn't do all of it).


PLAINTIFF'S EXHIBIT 355 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 355 Page 1 of 14]

[STAMPED] Gates DEPOSITION 56 2/28/[20]02
[STAMPED] X555301 CONFIDENTIAL
[handwritten] GO Corp --> Lloyd Frink [checkmark] 8M 8/13/[19]90

Microsoft Memo

To: billg darrylr, jeffr, mikemap, bradsi [handwritten underline] jabeb, tonyw, gregw, [s?]andyt, pradeeps, marline
Subject: GO Corp Info.
From: lloydfr
Date: 8/1/[19]90

Attached is a packet of information I have collected on GO. There are some of GO's very preliminary specs, a slide show of theirs, two of my trip reports on GO, and some press information.

Essentially, GO has been going out and telling the world that you need a whole new OS and apps for portable, pen based computers. People seem to bite on this, especially when they see that IBM is a strong backer of the GO OS (GOOSE). They use their notebook shell and OO framework combined with gestures and handwriting recognition to show everyone that their platform is indeed very different from the standard Windows/Mac world. They say you can't "hack this into an existing OS."

Well, we know this isn't so, and the best way to prove them wrong is to do our own "notebook" app. The big constraint is that we get it done in a year. So this won't be the most beautiful thing, but it should serve as a stopgap measure until we do our Win 4 shell and have all of our OO stuff in place.

GO Notebook Shell

When you turn on a GO computer, the user sees something that looks like a notebook. You can write directly on the page, probably do pictures as well. On a page you can open many document windows of any type. They use an object oriented approach to their OS, so opening up a document is just creating an instance of that clss type. When I was down at GO, Carr talked a lot about embedded documents, bu thte screen shots we have don't actually show that. These guys are smart, so I'll bet they will make it possible. When you leave a page, and later return to it, all of the documents will be in the same state as when you left (i.e. open/iconic, position). At the top of the screen is a menu bar with system-wide and generic document commands. If you want to create a new, blank document, you use the Create menu and choose the document type. Inside of each document are menus that correspond to that class.

Besides the many benefits of being OO, the thing which sets the GO notebook apart is the ease of navigation. On the side of each page are a set of "tabs" which correspond to folders, just a page in the notebook. Touch the tab and instantly you will go to a predefined workspace (similar to excel, but multiple document/object types). It is not clear if they are going to have only one level of folders, a "show all tabs" command, or some sort of hierarchy. The really nice thing about these tabs is that they are always visible and never get obscured by other windows. There are prev, next, and back buttons as well (the upper right hand corner). I suggested they add the ability push and pop of pages (hypercard). At the beginning of the notebook is a Table of Contents, which is a listing of the folder and page titles. At the end of the notebook is an Index of keywords the user has marked. Clicking on an item in either the TOC or index sends the user to that page. The user can also search for information in all documents and can create "goto" links (not really clear how they have implemented the latter). There are probably several other standard methods which can be applied to all objcets.

This metaphor does have a few drawbacks. If they number every page by sequence, and you keep inserting pages, then your page numbers are always changing and lose some meaning. Since they seem to be avoiding very many levels of hierarchy, what happens when you get tons of info? Does something like a piece of email belong on a page by itself, or with other email? What about rolodex cards? You kind of have next, perv, back navigational type issues within a window on a single page of a notebook. The notion of having a multiple page document, or a multiple record table in a window on a signle page of a notebook might seem confusing.

But this metaphor has a lot of appeal, especially compared to our program/file manager combination. I thnk we can whip something up in a year that embodies some of these ideas. I'll think about this in more detail and write it down in the next few weeks.

[Plaintiff's Exhibit 355 Page 2 of 14]

[STAMPED] X 555302 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 1
[Handwritten caption to diagram: GO Corporation Confidential]

[Plaintiff's Exhibit 355 Page 3 of 14]

[STAMPED] X 555303 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 2

[Plaintiff's Exhibit 355 Page 4 of 14]

[STAMPED] X 555304 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 3

[Plaintiff's Exhibit 355 Page 5 of 14]

[STAMPED] X 555305 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 4

[Plaintiff's Exhibit 355 Page 6 of 14]

[STAMPED] X 555306 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 5

[Plaintiff's Exhibit 355 Page 7 of 14]

[STAMPED] X 555307 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 6

[Plaintiff's Exhibit 355 Page 8 of 14]

[STAMPED] X 555308 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 7

[Plaintiff's Exhibit 355 Page 9 of 14]

[STAMPED] X 555309 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 8

[Plaintiff's Exhibit 355 Page 10 of 14]

[STAMPED] X 555310 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 9

[Plaintiff's Exhibit 355 Page 11 of 14]

[STAMPED] X 555311 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 10

[Plaintiff's Exhibit 355 Page 12 of 14]

[STAMPED] X 555312 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 11

[Plaintiff's Exhibit 355 Page 13 of 14]

[STAMPED] X 555313 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 12

[Plaintiff's Exhibit 355 Page 14 of 14]

[STAMPED] X 555314 CONFIDENTIAL
[This exhibit page shows a pen-computing user interface]

[The diagram is labelled with a document control number] 4.4.89 4 [the last "4" is crossed out and replaced by a handwritten] 13

PLAINTIFF'S EXHIBIT 359 Gordon v. Microsoft   [ Top ]


[resolution too poor to resolve periods and colons: these have been inserted where the text implies them]

>From lloydfr Thu Aug 9 15:32:22 1990
To: mikemap
Subject: GO Info
Cc: cathyw
Date: Thu Aug 9 15:25:21 1990

I sent a memo to you last week which had a bunch of info on GO corp. Turns out some of it was confidential and I was not supposed to make copies. Could you please return all of it to me so that I can destroy it?

Thanks - Lloyd

[handwritten] [indistinct] Material Redacted
[STAMPED] Raikes DEPOSITION EXHIBIT 99 11/2/[20]01
[STAMPED] X 578746 CONFIDENTIAL
[STAMPED] X578746

PLAINTIFF'S EXHIBIT 370 Gordon v. Microsoft   [ Top ]


>From jeffr Wed Aug 29 19:41:42 1990
To: debem
Subject: Please print
Date: Wed Aug 29 19:40:13 1990

For Go Corp file. Thx. Jeff
--------------
>From claraj Wed Aug 29 19:32:28 1990
To: jeffr rice
Subject: GO
Date: Wed Aug 29 19:28:50 1990

Phil Taylor, Dir. of PC Platform at Softview (a forms co.), is quite eager to give us info on GO. One of their software developers has attended two training sessions, April and June this year. Phil has access to technical documentation and will get more detailed information.

Softview has put a halt on pursuing the GO platform due to lack of resources in time and money in writing apps for a new operating environment. He is definitely interested in developing Windows apps. Being a Windows developer himself, he is definitely pushing for Win-H and open to giving us info on GO. I think he's being straightforward in his description. He was flipping through the papers while talking to me.

Due to my novice status, some of this information may be redundant, and the more technical information a bit fuzzy.

So...based on a technical documentation dated April 24, 1990, the GO machine has:

12 MHz. 80c 286. Zero weight [sic, should be "wait"] state.
3 meg RAM std expandable to 8
1 meg FEPROM
2 meg SRAM for user data
640x400 resolution. Black and white. LCD. Super Twister Reflection.
800 DPI 1:1 aspect ratio
4 shades of gray
9.5" x 13" x 1" with screen size of 5.5" x 8"
weight: 4.5 lbs
removable NiCd battery with backup Lithium cell. Battery life is 4 hours with 1.5 hrs to recharge.
optional expansion slot for Group 3 fax board - 9600 baud
Stylus is cordless. (Phil will check on button) supports in & out proximity, pen events (move up, down, window enter and exit)
NO disks. No h/d or floppy disks.

The current GO notepad has three components: memory, screen and pen.

It does offer an optional base station module-recharger that includes serial and parallel port to print to (HP LaserJet II, Dyconax 150

[STAMPED] Kaplan DEPOSITION EXHIBIT 62 4/19/[20]02
[STAMPED] X531215 CONFIDENTIAL

[Plaintiff's Exhibit 370 Page 2 of 3]


[STAMPED] X 531216 CONFIDENTIAL

(portable printer), Epson Fx/LQ), standard IBM PS/2 keyboard and Appletalk RS232. Floppy drives and hardware modules are optional on the base station.

Phil does not believe that GO is manufacturing the hardare, He heard a rumor that IBM is. GO has been having a tough time in receiving the hardware to fit their time schedule.

Handwriting recognition is CURSIVE and trainable. Phil does not think it's neuronetbased, based on it being a 286. The developer who went to the training class was very impressed by the demo. However, it was used only by Go's representative and not tried out by anyone from the audience. Recognition is concurrent with a slight lag but not distracting. He has little information on Go's hwx other than they are using static and dynamic approaches (speed, direction, sequence, timing).

Go does NOT havg a collision problem between gestures and characters. It doesn't sound like they're relying on different modlities. They've spent much effort in defining them and have cycled them twice already in the past few months. As of April, they were using double taps, triple taps, circle, carets, horizontal and vertical slash, question mark, X. Having started out with 20 system gestures, they have cut it down to 10-12. Phil mentioned that they're shooting for the magic seven. In addition to system gestures are application gestures and user-defined gestures (???).

GO uses a constant daytimer metaphor with tabs on the side, even when an application is being used. Thus, it forces the user to define the GO notepad as a daytimer.

Applications bundled are a word processor and a central address book. Tbeir operating system includes file conversion and network communication. Optional software is a drawing program (COM ??), a faxviewer and forms. He mentioned Tenpoint as one of their ISV's.

It does do multitasking, threads -- perhaps pre-emptive multitasking.

With reqards to the operating environment codenamed OSN, it's pseudo object-oriented written in C, not C++. It utilizes special structures and pound defines to look like object oriented code. But it is definitely not an object-oriented language. It contains an app framework, Win class, app class and object-oriented exensions to C.

The developing environment contains standard editors, linkers and source code debuggers. Development is done on DOS. Tools are DOS-based. The debugger and the emulator is done on Codeview. Developers develop on the PC on an emulator which simulates the GO environment. Then, when they receive the hardware, they can cross compile on the GO machine. The developing environment is split. One compiles and links on text-based DOS. But, booting is graphical (Windows). Thus, the developer has to be aware of DOS and Windows. Apparently, Go is pushing their training classes (100

[Plaintiff's Exhibit 370 Page 3 of 3]

[STAMPED] X 531217 CONFIDENTIAL

ISV's in the April session) to encourage ISV development prior to having the hardware prototypes. Phil says converting existing apps to OSN is highly dependent on how carefully the coding has been done and most likely result in re-programming from scratch.

No compatability with DOS or Windows.
No screen rotation suport.

Phil will get hold of more manuals and sort through then, in addition to talking to the person who went through training to get a better feel of UI. I told him I'll contact him Monday--to give him time. Given that further information is probably more technical, I think one of the development team should be here when I talk to him. Hope this information is useful.



PLAINTIFF'S EXHIBIT 443 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 443 p1 of 2]

[STAMPED] X531274 CONFIDENTIAL
>From jeffr Thu Nov 1 07:37:40 1990
To: debem
Subject: Please print
Date: Thu Nov 1 07:35:39 1990

For Go Corp file. Thx. Jeff
----------------
>From lloydfr Wed Oct 31 16:40:34 1990
To: billg gregs jeffr mikemap pradeeps
Subject: Approaching GO
Date: Wed Oct 31 16:39:37 1990

Our primary mission right now is to stop GO. We could do this by either 1) making sure they sign up no OEMs or ISVs, hence have no customers or 2) convert them to Windows. I've been thinking that it might not be such a bad idea to give #2 a shot again. My reasoning follows.

No matter what GO says, they know they don't need a whole new OS to do most of the things they are doing. I believe the reason they've gone into the OS business is mainly to make a lot of money, but also because they did not want to be burdened with the past. The fact that we've sold over a million copies of Win 3 and developers are flocking to Windows might make them see that being "burdened with the past" really isn't such a disadvantage. And, we've heard from someone inside of GO that GO is running into Win-H everywhere, especially the people Billg has spoken to; they call it FUD. If we've spoken to the accounts first, GO has a hard time getting off the ground. And to people GO though they had locked up and we talk to them, GO has to visit them again (Cannon?). The idea is not to get them to convert right away, but to build a relationship with them so that it is not so distasteful for them to go that direction at some point in the future. Right now we are the big bad Goliath, and I don't think we have much to lose by being very open with them about what we are doing and how they could fit into it.

How could GO make money working with Windows? Basically I think they have a lot of smart people and good ideas. They have three things which would work well on top of Windows - their HWX, their notebook shell with applets, and their OO framework. They should see that as well as portables. there is a huge desktop market that could use this same technolgoy. And you just have to have a compatible OS to try and reach the desktop. If they decided to port their stuff to Windows, then they would essentially be competing with both the handwriting group (HWX and notebook applets) and AFX, but not the Windows group. We could license Win 3.1 to OEMs, and then the OEMs could buy GOs stuff on topof that. I'm not sure how much of a business opportunity this is, but at least it is something for GO to fall back on if they are unsucessful in the path they are taking now. Also, do we want to risk losing some of potential revenue?

If we were to approach GO, I assume we would meet with them, give them our API's now and an SDK when it's ready. We might want to show them a demo of the compatibility layer, but definitely not the notebook. We'd just try to be more open and friendly and not ask for anything in return. We'll say that our apps division will consider writing apps for


[Plaintiff's Exhibit 443 p2 of 2]

the GO machine, but realistically we wouldn't make that decision until they have sold a fairly large number of machines.

A last benefit of taking this friendly approach is that we might lessen the chances of them suing us for some unknown reason right as they are about to go under.

Just an idea, any comments?

- Lloyd

[STAMPED] X 531275 CONFIDENTIAL



PLAINTIFF'S EXHIBIT 478 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 478 p. 1 of 4]

GO Corporation Confidential Mar 5, 1994

December 2, 1990

Jerry Kaplan notes for meeting with Jim Cannavino

Executive Summary

* Go has excellent technology, lead time, visibility, and is well received by industry
* GO + IBM licensing has not proved enough to win over other companies to date
* Microsoft strong competitive threat has stalled progress, put financial pressure on GO

GO Status

* Completing "developer release" of 286 hardware and software
excellent reception of development environment by ISVs
200 people trained
Tens of application development projects
good press, visibility, book series
* Penpoint unveiling and developer product announcement Jan [19]91
* Starting on 386 version
software promiseed to IBM end of 3Q91
* Customers excited about product
IBM special accounts signed up and ready to go
Many VARs interested, seeking credible hardware vendor
* $15M raised - will last for the next year.

GO's Plans

* Establish "clean" 32-bit 386 version as market standard
* Limit 286 to developers and pilot testing
* Complete and license 386 hardware version, then
* Spin out hardware/system's busines

January Announcement

* Full day of press and developer activities, San Fran 1/22
* Boston Computer Society meeting 1/23

Page 1

[STAMPED] Kaplan DEPOSITION EXHIBIT 34 4/18/[20]02
[STAMPED] CONFIDENTIAL
[STAMPED] KAP0500

[Plaintiff's Exhibit 478 p. 2 of 4]


[PAGE HEADER] GO Corporation Confidential  Mar 5, 1994

* Third party application products announcements
* With luck, possible additional Penpoint licensees
* Expect VERY heavy press coverage

What will it take for us to succeed?

* Multiple manufacturers with Penpoint machines
* ISV support
* Staying power to fight competition during market development

What are we finding?

* "Wait and see" attitude among manufacturers
    IBM support is not effectively communicated
    Unlike the PC, no IBM machines as example
    Unlike the PC, no proven market need
* Good developer support, but concern about when machines will be available. General availability in 1991 is essential.
* Microsort is stalling our progress, hoping to starve us out

Microsoft - our only serious competition
* Their strengths -
     Perceived as a standard setter.
     Financial staying power.
     Relationships with ISVs, OEMs.
* What they are doing -
     Copying us.
     Creating confusion.
     Promising everything to everyone.
     Tryinq to unhook our ISVs, licensees.
     Nearly giving product away.
* Their basic pitch -
     Microsort is standard setter, not IBM/GO.
     They will do whatever GO does in time.
     Their system will be compatible with Windows.
     IBM will be forced to work with them eventually.

[PAGE FOOTER] Page 2
[STAMPED] CONFIDENTIAL
[STAMPED] KAP0501

[Plaintiff's Exhibit 478 p. 3 of 4]


[PAGE HEADER] GO Corporation Confidential  Mar 5, 1994

GO's strategic advantages
* Great technoloqy - only system really designed for a pen
* Two year lead over Microsoft
* Outstanding development staff - 60 of 90 people
* Attracting the creative talents of the developer community
* IBM "design win" and public support

What we need from IBM
* Increase visisbility of IBM comitment
* Clear messages
    1. IBM sees a major market for pen-based computers.
    2. IBM has evaluated Microsoft and GO, and has chosen to work exclusively with GO.
    3. IBM is supporting GO to make Penpoint an open industry standard.
    4. IBM will bring a product to market running Penpoint in 1991.
    5. IBM is now working with major customers on pilot projects.
* Promotion of GO licensing efforts
    Active support from IBM Japan executives
    Direct high level contacts: NCR, Toshiba, others?
* Ship a Penpoint machine to general distribution as soon as possible
* ISV support
    Assign advocacy/technical support resources
    Seeding/""scholarship" program for universities
    Multi-city road show support
    Internal applications development - ASD?
    Outside applications support - Easel?
    Internal Territory Management Systems effort?
    "Venture fund" for Penpoint ISVs

[PAGE FOOTER] Page 3
[STAMPED] CONFIDENTIAL
[STAMPED] KAP0502

[Plaintiff's Exhibit 478 p. 4 of 4]


[PAGE HEADER] GO Corporation Confidential  Mar 5, 1994

What can you do personally?
* Strong video endorsesment for January announcement
    Personal appearance at Boston Computer Society,
    Authorize a WSJ ad congratulating GO.
* A few interviews - Fortune Magazine (by 12/10), some dailies
* Improve and expand executive focus on helping GO to succeed
    Promote GO/IBM relationship inside and outside
    Clear roadblocks, reduce bureaucracy, control PR
    Centralize and coordinate negotiations
* Allocate some discretionay budget for non-development Penpoint promotional activities

* Do not cooperate with Microsoft in this area at this time!

Summary

* We've got the right technology, the lead time, and the hearts of the ISVs.
* We need to turn up the heat on Microsoft.
* We need to exand IBM standard setting support,
* We need to increase overall investment in Penpoint,

Give us air cover and runway, and we'll create a new (non-Microsoft) systems software standard for pen-based computers!

[PAGE FOOTER] Page 4
[STAMPED] CONFIDENTIAL
[STAMPED] KAP0503


PLAINTIFF'S EXHIBIT 687 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 687 Page 1 of 2]

[Fax Cover Sheet]
To: EXECUTIVE From: MICROSOFT CORP
APR 18 [19]'91 11:00 FROM GO CORPORATION PAGE .001
4-18-91 12:09pm p.1
GO Corporation
[ADDRESS / FAX / TELEPHONE]

DATE: 4-18-[19]91

TO:
NAME: Mr. Bill Gates
COMPANY: Microsoft
FAX NUMBER: [FAX]

FROM:
NAME: Jerry Kaplan

NUMBER OF PAGES (INCLUDING COVER SHEET): 2

IF YOU HAVE ANY PROBLEMS WITH THE RECEIPT OF THIS FAX, PLEASE CONTACT: Holli Maxwell AT [TELEPHONE]

MESSAGE: Thank you!

[STAMPED] Kaplan DEPOSITION EXHIBIT 39 4/18/[20]/02
[STAMPED] X 504392 CONFIDENTIAL

[Plaintiff's Exhibit 687 Page 2 of 2]

[ADDED BY FAX PROCESS] To: EXECUTIVE  From: MICROSOFT CORP 4-16-[19]91 12:09 pm p.2
[ADDED BY FAX PROCESS] APR 18 91 11:00 FROM GO CORPORATION  PAGE 002
[STAMPED] X 504393 CONFIDENTIAL
[ADDED BY FAX PROCESS] ** TOTAL PAGE 002 **

[LETTERHEAD]
GO CORPORATION
[ADDRESS]
[FACSIMILE]
[TELEPHONE]

April 17, 1991
Mr. Bill Gates
Chairman of the Board, Chief Executive Officer
MICROSOFT CORPORATION
1 Microsoft Way
Redmond, Washington 98052-6399


Dear Bill:

This is to follow up on our phone call of yesterday.

I appreciate your offer to sit down with me to discuss any potential issues GO may have with Pen Windows.

As I explained, at this time I am simply trying to learn enough about Pen Windows to make an informed evaluation.

I’d like to take you up on your offer to provide us with relevant materials without restriction. We have already received the Pen Windows developer kit, but it arrived with a license agreement stating that it must only be used for the purpose of developing applications. Given our conversation, I will assume that it is OK with you for us to examine these materials vithout signing the license. If this is not what you had in mind, please let me know by next Monday. I will wait until than before opening the materials.

Should the developers kit not contain the current version of the “notebook applet”, may I request that you forward this to me (object code is fine, as long as we can run it).

After we get a chance to look over these materials, I will let you know if additional materials will be helpful. I will try to complete an evaluation in a timely manner.

I’d like to reiterate that I have no interest in stirring up press interest in a “GO/Microsoft fight”. I appreciate your willingness to cooperate with us, and look forward to resolving the matter.

Sincerely,
[SIGNATURE]
S. Jerrold Kaplan
Chairman

PLAINTIFF'S EXHIBIT 718 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 718 Page 1]

Microsoft Corporation [ADDRESS / TELEPHONE / TELEX / FAX]
[handwritten] Billg
[Microsoft Logo]

VIA AIRBORNE EXPRESS

May 3, 1991

Mr. S. Jerrold Kaplan
Chairman
GO Corporation
[ADDRESS]

Dear Jerry:

After our phone call and your letter to me last week, I decided to write this letter so that my view is clear.

Microsoft has been very interested for a long time in handwriting recognition software using the pen as the primary interface. When Jeff Harbers and I met with you and Bob Carr on July 11, 1988 we discussed the fact that we were working with several companies on system software for handwriting machines.

When Microsoft met with GO in May and June of 1989, GO knew that we were seriously considering putting handwriting recognition into Windows. During these meetings, you and Bob Carr tried to get Microsoft's commitment to develop handwriting applications for the GO operating system. You knew we'd be in the business, whether on GO's OS or on our own, and you were careful about what you showed us and told us about handwriting recognition and gestures. Specifically, Lloyd Frink saw no code at all and received no information about how you did handwriting recognition. He wasn't permitted to take any documentation, any software or the early version of the SDK out of your offices. Lloyd did see a couple of gestures that have been staples of the Associated Press editing guides for years.

Microsoft received no information about GO's pen product, except what we read in the papers, from the time of those meeetings until your recent public demonstrations of PenPoint. What information we learned during our meetings was largely irrelevant since we saw nothing about handwriting recognition, which we cared a lot about, and a lot about GO's OS, which we cared not much about.

[letterhead footer] Microsoft is an equal opportunity employer.

[STAMPED] X 504373 CONFIDENTIAL

[Plaintiff's Exhibit 718 Page2]


Mr. S. Jerrold Kaplan
May 3, 1991
Page 2

We have been clear at all stages of our discussions that we were interested in putting handwriting recognition and a pen interface in Windows. You hoped to convince us to rewrite our applications for your operating system. However, nothing that we learned from Go undermined our belief that Windows was the way to go. As a last attempt to convince GO to write its software for Windows or license it to Microsoft to put on Windows, we asked Robert Carr to come to Redmond and we disclosed a lot of confidential information about our plans for Windows. The upshot of that meeting was your offer to license GO's technology to Microsoft, but on terms we could not accept. Reluctantly we went our own way.

As I said over the phone, if you have any problems after looking at our Pen Windows SDK and other materials, or if you think you need other materials, please call me and l will make the matter top priority, I am ready to sit down and discuss any of this, and any other concerns you might have.

Very truly yours,

MICROSOFT CORPORATION

[signed] Bill Gates

Bill Gates
Chairman

cc: Bill Campbell

[STAMPED] X 504374 CONFIDENTIAL


PLAINTIFF'S EXHIBIT 733 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 733 p1 of 2]

GO Corporation
[ADDERSS / FAX / TELEPHONE]

[STAMPED] RECEIVED MAY 15 1991 CORPORATE
[handwritten] file --> billg. Cc: jeffr mikemap mikenal billn jonl

May 13, 1991

Mr. William Gates
Chairman of the Board & CEO
MICROSOFT CORPORATION
[ADDRESS]

Dear Bill:

Thank you for your letter of May 3 clarifying your views. However, I must take issue with your version of the events downplaying the significance of GO's disclosure of confidential information to Microsoft.

When we initially discussed GO's work in July, 1988, contrary to your statement, there was no discussion that you were working with several companies on systems software for handwriting machines which use the pen as the primary interface. In fact, the concept as a whole appeared to be quite a surprise to you at that time.

After our initial meeting in July 1988, you and executives in your applications division expressed a serious interest in considering writing applications for our system. It was represented to us that your applications division was separate from your systems software division, and that information obtained from us would not find its way into your system's software division and therefore would not be used to deveop competitive products.

Based on these representations, we hosted some Microsoft employees, including Lloyd Frink, for extensive briefings about our designs, plans, and implementation, with the view toward having Microsoft's applications division develop applications programs for our system. We discussed at length the results of our research and investigations, our intended marketing approach, and what we had learned through prototyping and testing of different approaches. Lloyd had free access to our design documents, handwriting interfaces, gesture set, and engineers.

[STAMPED] Gates DEPOSITION EXHIBIT 60 2/28/[20]02
[STAMPED] X 504369 CONFIDENTIAL

[Plaintiff's Exhibit 733 p2 of 2]

Letter to Mr. Gates
May 13, 1991
Page 2.

If we had known that Lloyd might become a principal supervisor of Microsoft's development of a competing product, we would not have disclosed our trade secrets to him, much less provided him with an ofice in our headquarters and unsupervised access to our confidential information.

The demonstrations of Pen Windows we have seen to date, quite frankly, cause us concern about use of the information we disclosed to Microsoft, Frink's involvement in the design of Pen Windows, and the apparent similarities between Pen Windows and PenPoint. I will get back to you with a more definitive analysis after we complete our review of your materials.

Sincerely,

[signed] Jerry
S. Jerrold Kaplan
Chairman

PLAINTIFF'S EXHIBIT 746 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 746 Page 1 of 5]

>From nathanm Sun May 26 10:03:18 1991
To: pradeeps
Cc: billg bradsi cameronm gregs jeffr jonl karenh lloydfr marline makemap raleighr robg steveb
Record-folder: C:NATHANMFOLDERSWSENT.FLD
Subject: RE: PCs that aren't PCs
Date: Mon Mar 16 11:31:28 PDT 1992

I agree with much of what you say. In fact, the part that you said I didn't understand is exactly what I would have expected you to say. The only differences that we have are rather subtle, but I think that they are important. Here are some specific comments:

The assumption that the PenWin mkt is the 'existing PC user' while GO's mkt is the 'new user' is wrong. (What the hell is a 'new user' anyway? Presumably over the past 10 years, we've gathered 50m 'new users'.) GO wants to position us that way -

Yes, that is my point! I think that we [are] in violent agreement. There are several quite different topics that seem that we potentially could discuss:

- What your marketing position *is* - i.e. what people really perceive. This is what you tell them as filtered by what they believe and retain.

- What your posture *should be* - i.e. the message you shouldbe saying. Note that this is usually a strong function of the audience, or market segment.

- What the fundamental "lay of the land" is - i.e. the techincal [sic] and market *realities*. Note that these can be quite different than either what you say or what people perceive. Sometimes you can get away with saying things that are very untrue (like the misuse of the word "Open" in the UNIX community, or the old practice of calling dictatorships "The People's Democratic Republic of..."). The usual case however is that you need to make sure that what you are saying is in line with techincal [sic] realities.

Go's message is very pure in that all three of these line up. They say that they are for people who are likely to only use a Pen based machine, and do so primarily in tasks which are not typical on desktop PCs (that is a good definition of a "new user" in this regard). PC industry people believe Go when they say this, because it is obvious that they can't run any PC industry apps - so it is clear that they want to be something else. Also, Go can make some credible claims to having technically optimized for this case.

My comments about Pen Windows were primarily about the technical and market reality, and I think you interpreted them as statements about what your message should be.

WinMail 1.21 lynnra Mon Mar 16 11:18:50 1992 Page: 109

[STAMPED] EXH 22 DATE 02/14/[20]02 WITNESS [signature] SUSAN ZIELE
[STAMPED] MS 5024470 CONFIDENTIAL

[Plaintiff's Exhibit 746 Page 2 of 5]


The reality is that if you completely subtract the connection to Windows apps, Go is way ahead of us (in my opinion). That is not a message I would suggest you communicate, and I don't think that many people percieve that yet (fortunately) because you have done good work. The good news is that we don't have to subtract Windows out, and you can use the aura of Windows to help in the other markets. This is explained more below:

we're in this business precisely because we have no intention of giving that market up. Better than 50% of pen system sales in the next couple of years are likely to come from fleet sales to people who are so-called 'new users'. We intend to win those sales. Yes, most of these will be sold direct; will have few apps on them, but will still lead to a critical mass of pen machines (and pen OS's). Once the socket gets out there it will attract ISVs. Those sockets must be PenWin sockets.

I agree that it is dangerous to cede these sales to Go, and I had no intention of suggesting that you do this. Your goals, and your message should reflect the fact that you want to win there too.

Even in the case of fleet sales, I believe that your ability to convince people that we have an advantage is directly or indirectly Windows. If you had to give a demo and sales pitch where you never once ran a normal Windows app, or otherwise used the imprima of Windows, you would be at best be on even ground with Go, and I don't think that would be much fun.

So, to sum up on this, I was not misunderstanding your goal of doing well in that market, or your message which supports that goal. I just think that Windows is your ace in the hole, and that within the next 2 years 90% of your sales will be either directly be to Windows users, or to "new pen-centric users" (as defined above) who picked Pen Windows over Go directly or indirectly because of the connection to Windows (part of coroporate strategy, support from machine mfgr, FUD with going "against" Windows, expectation of future apps...).

As an aside, I think that there is a big danger in your message becomming to diffuse if you tell everybody that you'll win against Go in all markets. Go has a very focussed position, and as much as possible you should use this againtst them by letting them position themselves into a corner.

In particular, when it comes time to talk to ISVs, you really want to make the point that Go's intended market is NOT a good place to sell their apps. I think that the point you raise above about fleet sales eventually becoming sockets for apps may have some merit, but this is a terrible thing to say to ISVs. X think the best *message* to ISVs is that Go's market is IRELEVANT to them in turns of selling any maintstream PC app. Kany of the other points I made in my previous email also apply - the machines are priced such that PC industry people are a large part of the early market, PC OEMs are building the machines etc.

A specific point - yes, Pens should be positioned as an integral part of the Windows soln (the PenWin SDK will be bundled with the 3.1 SDK), but it is an oversimplification to say that any good 3.1 app is a good pen app.

I believe that you must MAKE this true as much as possible. People making Windows apps that ship in 92 should feel embarrased as hell if they don't work well with a pen. Note that this is ANY Windows app - I didn't say "Pen Windows ISVs".

We've got to evangelise the pen stuff specifically - no Win developer is likely to build support for 'ink' (scribble objects), interaction with the recognizer, extended gesture support etc unless we raise the awareness of why pens are important.

Yes, you must raise awareness, but ideally it should be in a Windows context. I would for there to be an ISV that says "yes we have a great Windows app, and later on we'll make a Pen Windows app". I'd rather have the guy thinking "we really have to fix our Windows app so that it works well with Pens, outline fonts and other new Windows features". Some apps will target Pen specifically, just as color paint programs target machines with color monitors, but the broad mainstream should cover, it all.

We can get people to buy our OS only because the end-user sees solutions that he can't get elsewhere - thus forcing the OEM to support our OS. Building relationships with consumer OEMs is important as is building a lighter weight Windows versions for the PCs w/o kbds. But NEITHER is as fundamental as getting apps out there which meet those needs - even though they may be on machines sold by traditional-PC OEMs and on a overweight OS. The very existence of those apps will then drive the consumer OEMs to adopt the MS soln. The basic argument is that the proposition laid out in Nathan's email: "PMK = Consumer hardware + ? apps + consumer positioning" is a non-starter.

You HAVE to build from your strength - existing PC's; get the apps built; get people using them, then grow the mktg broader through lowered price points on the hardware. And that's what our pen effort must do.

There are many valid points in what you say, but there are some subtle distinctions which I believe are crucial. I believe that they could make the difference between success and failure in this area.

What you have described is ONE way to approach the consumer market. It could be paraphrased as "PCs move down and take over consumer electronics". We start with PCs, on which we are already successful, get apps built which satisfy consumer needs, let them incubate in the PC industry until the machines are cheap enough, eventually "forcing" (your word) the consumer electronics companies to go with us.

I would love it if this happened, but it would be negligent of us to RELY on this happening. We are presently investing in this direction, because it happens to be a direct extension of our PC business - i.e. this bet is already covered. Pen Windows is one great step in this direction and our Multimedia efforts are another one.

I agree with you that we must build on strength, but this is more than one way to do this! We must hedge our bets.

The consumer electronics companies will regard the scenario you describe as all out war. They are oriented around proprietary standards and have a different mentality than the PC industry. If we remain insular and focus just on "PCs will move down", then consumer companies will create alternatives. CD-I is exactly such a thing, but consider that the tip of the iceberg - they will create many more for PMK machines (and enhance CD-I a lot) . Remember that these guys buy movie studios and record companies just to feed their gadgets today. The amount of money that Philips has invested in CD-I titles and other infrastructure is staggering. They will make similar software investments for PWKs if they feel they have to, and the "PCs move down" scenario is so threatening to them that they will feel they have no choice. I would like to see them do this in partnership with us.

I think that we should continue to press forward on "PCs move down" approach but IN ADDITION we should do a project with the following characteristics:

- Work closely with, at least one major consumer electronics company from the very beginning, and learn to adapt to their culture and approaches (as much as is feasible). I think Sony would be be3t for a variety of political reasons in the consumer industry, but this is a huge topic unto itself.

- It would be nice to have 2 projects with them in the PWK area. One would be a fairly near term thing (such as Sony Bookman?) , and another would be much more ambitious - which might be considered a prototype in the inital stages, but which we want to be a product. The reason is that near term stuff is important to get a foot in the door and establish a working relationship, but it is too limited in terms of the hardware (8088 character mode...) to be a the base of anything longer term. We heed to get in on this, but also leapfrog out to the point where our full gamut of software (Windows etc) is feasible.

- We would try to draw on things which appeal to our partner. As an example, Sony is a MIPS architecture licensee, and really wants to get into chips. Suppose we had a project to make a lightweight portable Wia 32 implementation for a PWK. Sony could focus a project on making a super low cost MIPS system with custom chips, and perhaps even custom CPU. If we found a way to need a couple of special new instructions (say for power management...) and they could feel they were getting an edge on the world this way, so much the better. Another example would be to support some cool hardware that they had in the works - a new kind of storage (next generation mini-disk). We want to work our way into their strategies so that they rely on us in many ways.

- Our "strength" in this case is our credibility in software, our position as a partner, and the connection to the PC industry. I believe that we can get such a project going on this basis.

- Technically we want to tie in to the Windows world. This is what gives us the huge strategic win - we can draw on apps and other things being developed in the PC industry and our other initiatives (Pen, MMsys...). The various options mentioned in the previous mail (drop Dos etc.) still leave considerable overlap with Windows.

- When I wrote "? apps" in the equation you quote, I didn't mean that we would want to run PC industry apps (although there are subsets, like Dos apps, that may not be relevant). The point is that the apps which will be compelling to PWK users are not part of the PC industry's current inventory. We need to get them developed. The consumer companies are more than willing to spend millions doing this (again, look at the CD-I investment) . We want this investment to be channeled into something that will also benefit the desktop - by being based on Win 32.

I would love it if the PWK battle between consumer computers moving up and PCs moving down was fought with Windows on both sides. Our destiny is to be the Krupp Werks of the 1990s. I want us to be in a position such that we DON'T CARE if the PWK of the future is sold by Sony (and leverages connection to entertainment media), or Sharp (and is more of a calculator), or is sold by a communications equipment company (and leverages cellular telephone), or in fact is a PC that moved down and is sold by IBM, Compaq and Tandy. In ALL cases they should be running Microsoft software - in particular, some variant of Win 32.

We have one of these avenues firmly in hand, but we are not far along in getting the consumer people to work with us. I think it is time to start cultivating that community.

Nathan


[Plaintiff's Exhibit 746 Page 3 of 5]

[Plaintiff's Exhibit 746 Page 4 of 5]

[Plaintiff's Exhibit 746 Page 5 of 5]


PLAINTIFF'S EXHIBIT 809 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 809 Page 1 of 4]

[STAMPED] Kaplan DEPOSITION EXHIBIT 43 4/18/[20]02
[STAMPED] KAP0037

Microsoft Corporation
[ADDRESS / TELEPHONE / TELEX / FAX]

July 3, 1991

Mr. S. Jerrold Kaplan
GO Corporation
[ADDRESS]

Dear Jerry:

Thanks for your letter dated May 13. My heavy travel schedule has delayed this response.

As to the July, 1988 meeting at GO, you and I have pretty different recollections. While I don't want to seem at all argumentative, I am confident of the substance of my recollection as set out in the early part of my letter to you dated May 3, and I invite you to consider some corroborating material. I am enclosing a copy of an email message I composed and sent on Thursday July 14, a few days after our meeting. Several points emerge: a) we at Microsoft had considered well before that visit a machine of the sort you were designing; b) I asked our Systems and OEM Vice Presidents to pursue selling Windows to GO; and c) I asked our VP for Applications to consider an applications strategy, albiet pessimistically. Also enclosed is a paragraph from my May 22, 1989 keynote address at the SPA conference, when I touted Windows-based pen systems under design in Japan.

Regarding events after the July meeting, I stand by my description in my May 3 letter of discussions between our two companies, each having a preferred goal: Microsoft to license Windows to GO as the operating system for your system, and GO to persuade Microsoft to become an ISV of applications for the GO machine.

Regarding Lloyd Frink's role, I can understand GO's confusion about his titular responsibilities. The fact is that I asked Mike Maples, our VP for Applications to pursue the handwriting recognition technology and he delegated it to Jeff Raikes who enlisted Lloyd. Lloyd believes he negotiated in good faith the mertis [sic] of GO licensing Windows for its operating system while entertaining GO's arguments why Microsoft should become a GO ISV and that he did not misappropriate any trade secrets in the course of his discussions with GO.

[letterhead footer:]
Microsoft Corporation is an equal opportunity employer.

[Plaintiff's Exhibit 809 Page 2 of 4]

[STAMPED] KAP0038

Mr. S. Jerrold Kaplan
GO Corporation
July 3, 1991
Page Two

Jeff Raikes reports that he had a lengthy and constructive discussion with you and Bob Carr in mid-May about these issues. If there is any further information that we can provide, please let me know.

Sincerely,

[signed]

William H. Gates

Enclosures
cc:
William Campbell Robert Carr
John Dore
Mike Maples
Jeff Raikes
Greg Slyngstad

WHN:pf MiscWHG-Go

[Plaintiff's Exhibit 809 Page 3 of 4]

[STAMPED] KAP0039
Jul 20 08:04 1988 MAIL Page 15

>From billg Thu Jul 14 16:42:49 1988
To: gregm jeffr joachimk jons mikemap steveb
Subject: GO corporation
Cc: charless jeffh tandyt
Date: Thu Jul 14 16:42:45 1988

Jeff Harbers and I met with Jerry Kaplan and Lob Carr of GO corporation Monday afternoon.

Basicly (sic) they are building a machine that Kay and I talked about building along time ago - a machine with no keyboard and no disk using-static memory. Its like an 80286 version of the model 20D with 2meg-8meg using a writing stylus with handwriting recognition forinput. According to Marquardt there are a few other people building things like this - in fact there was one discussed in the WSJ this week. Its [sic] notebook size. The LCD is 640x400 so about 55DPI (which I dont think is enough)

They-are doing ALL their own system software - a protect mode OS for 286 using visual objects (like everyone!). Its [sic] multitasking. The interface metaphor is a set of named folders with tabs on the right hand side each containing any number of numbered pages and each page has on it [sic] just ink (writing) or rectangles that contain application sessions (which can be zoomed/unzoomed) All the old ideas like using gestures for various commands they have "rediscovered" They will announce in 1990 at $3k. Modem is optional. The[y] will bundle some drawing/wp/filing/notetaking/mail software but they want to get third party isv's including microsoft.

We tried out their handwriting stuff and it was terrible. Its very possible to do this stuff correctly and maybe they will but they havent yet.

ANALYSIS: This machine should be built as an open standard by a bunch of Japanese makers. The software layers should be more compatible with desktop-stuff. Kaplan isnt the best CEO. They have some OK ideas but I dont think this thing will be big. We do need to think about note taking and the fact that small machines can be used everywhere especially with this input approach but I dont think we should be an ISV for them.

ACTION ITEMS: Gregm - Carr wants our debug format to do a remote debugger. If it is written up and easy to send then send it to him with a letter saying they will use it for developing on their-machine only. If it is hard then have someone call and say sorry.

JoachimK/Steveb - we should be selling system software to people like this. He is fairly far along at this point. What would we sell him? Either stripped down PM or WINDOWS. WINDOWS is the best choice I think. Tell him to use extended memory. He wont like this but it will sure help him with applications developers. Who can take a pass at this with Carr? Mikemap - Another applications opportunity but unless we want something that fits on this machine for the desktop I doubt it makes sense. They do want to create connections between stuff on their machine and popular desktop stuff so we want to be friends with them even if they are not an ISV. They offered to come up and present their concepts to a larger group up here. I doubt that make[s] sense. I will talk about this class of machine in our saturday morning meeting.


[Plaintiff's Exhibit 809 Page 3 of 4]

great things about this package is the way it uses dynamic data exchange to enable you to use a spreadsheet to model your costs or model certain engineering figures, and in real time feed that data into the drawing so that if you change that drawing the data feeds back to the spreadsheet, if you change the spreadsheet it feeds back to the drawing. In fact it's one of the most graphical illustrations of the benefits of dynamic data exchange that I've ever seen.

One of the ways that Windows will be used in 1990 that I hadn't even expected is that we have a number of manufacturers, all of them based in Japan, building machines where they take Windows and they put it into ROM and instead of putting the keyboard on the machine they simply put a touch-sensitive LCD plastic covering on the machine and so it looks almost exactly like an 8-1/2 by 11 notebook. In fact there's a number of companies working on exactly this design. By using the Windows software built in, and by having a handwriting driver that allows an unmodified application to receive the keystrokes as though they came in from the keyboard, we think that they will be able to tap into a wide base of software and also get people to do special modifications to their software to take advantage of the sorts of things that people want to do in an environment where you can carry the PC around. I think it's fair to say that keyboard machines will never find their way into meetings like this or to meetings inside your company, they also won't be carried around inside the office. And so there is a hugh [sic] area of use that relates to note-taking and communication and sales calls that only this stylus-based type PC can accommodate. There are no technical breakthroughs required for this machine - the idea of following the trace when you do handwriting and recognizing the character is actually not a difficult software feat. So this is not rocket science, this is not voice recognition or something that will take a long time to develop - it's something that you will see in 1990 and I think is one of the opportunities in the Windows environment for people who want to do specialized software.

Well I should talk a tiny bit here about the OS/2 operating system, just to provide you an update about what's going on. It's fair to say that no operating environment ever succeeded solely because of technical strength, but people who create these things like to talk about their technical strength anyway so if you'll indulge me for one slide I'll talk about some of the things we are excited about that are built into the OS/2 system. First of all, this is true multitasking and true multitasking is different than what we have today on the Macintosh Multifinder or on Microsoft Windows. The reason it's different is that at any time an application can come in and get system resources and this is absolutely critical. If you want to receive electronic mail in the background, if you want to have good prints

[page number] 8
[STAMPED] KAP0040

PLAINTIFF'S EXHIBIT 859 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 859 Page 1 of 4]

[STAMPED] X 189615 CONFIDENTIAL
[STAMPED] Raikes DEPOSITION EXHIBIT 56 10/18/[20]01
[handwritten] file
[then a single set of illegible initials, ticked off with a check mark]

Microsoft Corporation [ADDRESS / TELEPHONE / TELEX / FAX]

Microsoft Memo

To: Bill Gates, Jeff Raikes, Steve Ballmer, Mike Maples, Brad Silverberg, Jim Allchin
From: Greg Slyngstad
Date: July 24, 1991
[transcription note: the following "Challenges to PenWin" is circled and annotated with a handwritten note which reads] OBU [or OBO] PCB
Re: Challenges to Pen Win
Cc: Pradeep Singh, Lloyd Frink, Aaron Getz, Marlin Eller

This memo highlights some of the threats of PenPoint and warns that Go is making progress on many fronts and unless we take them seriously, Go may emerge with a majority of the Pen OS business. Now is the time to stop them, not after it is clear we are losing. I list several indications of Go's strength and suggest some items that we need to work on in order to weaken their position.

I am more optimistic than this memo may indicate, but I am concerned that our strengths (wealth of Win apps and Win momentum) are not playing out as well as we would like. The majority of our resources (in the Windows area) are focused on a diferent battle. However, there are things that we should be doing that can help against both PenPoint and OS/2.

Go Marketing Wins

* The press continues to favor Go. We are making some progressbut the prevailing anti-Microsoft attitude and the cool new technology of PenPoint makes Go very appealing to the press. Even those editors that believe we will win, want to see Go succeed.

* Go has a focused sales organization. Almost every major corporation we visit has already had a visit from Go. Many of these corporations complain about the complexity of Windows and are open to exploring an incompatible system. Go is very good at convincing them that PenPoint is amazingly easy to learn and use. State Farm said they would rather their technical MIS people had to put the effort into making the system work, rather than the end user. On our side, the Microsoft field sales and support force is reluctant to spend time on an OEM product (with a few notable exceptions).

* ISV's claim they are doing their new Pen oriented software first on PenPoint (Lotus, Software Publishing, WordPerfect as well as many of the startups) At the recent Pen Conference in San Francisco, David Reed from Lotus said, "We would be stupid not to support Windows for Pens, but our real creative energy is going into Pen Point." They will support Pen Win but they downplay this and most ISVs are planning on doing simple mods to their Windows apps.

* Most OEMs will make sure their system will run either OS. People we know are headed this direction: NCR, Toshiba, Zenith, Dell, Grid, Microslate, Wang and Tusk not to mention IBM being PenPoint only. Recent discussions with Boca on PenWin have cooled. I sense the high level anti-Windows sentiment is stopping the pen group from moving forward.

* NCR is leaning towards PenPoint. They claim OS neutrality but their recent add [sic, "advertisement"?] highlights PenPoint and there are definite anti-Microsoft feelings in their pen marketing group.

[Plaintiff's Exhibit 859 Page 2 of 4]

[STAMPED] X189616
[STAMPED] X 189616 CONFIDENTIAL

Scalability. Go is pushing OEMs to build small form factor machines. We are claiming we will run on these systems as well. In theory Windows should be scalable to these small displays. Lloyd has begun exploring this and it is clear that it has never been tested and full of flaws. He has reported some of these to the Windows team, but I wonder how much effort will go into fixing the problems given the need to ship 3.1 (the program manager and file manager do not even use dialog units based on the system font so they do not scale at all, the system menus and min/max icons do not scale so you end up with an ugly title bar if you increase the system font size.) We have only scratched the surface at looking into this but it is clear we need to test this and be prepared to fix the problems.

Silicon only storage. Go treats their storage space as one contiguous address space and will execute in place when running on a silicon storage (flash) system. Windows 2.1 will be ROMmable but it still requires all apps and much of Windows be duplicated during execution. Since silicon storage is required for highly durable systems and the price for this storage is so high, efficient use is an important benefit. IBM has told us they will never put a hard disk in a portable pen machine.

Notebook User Interface. The Go shell is significantly easier to understand and use than the Windows program manager and file manager.

Instant On. The Go system allows you to turn off the system and instantly return to where you left it when powered back up. Not clear if this works on hard disk based systems.

Windows apps in general work poorly with the pen. Selection and command execution work well but with the exception of graphics entry apps (CAD, Paint, Draw) the standard input expected still comes form the keyboard. This means the user must suffer through character recognition. The best horizontal pen apps will rely heavily on ink in its native form. This applies to PIMs & notetakers, PenMail, handwritten annotations of documents, ink in your database, FAX annotations etc. This means we need to continue to get bandwidth from our ISVs and not let them think that adding handwriting edit controls is sufficient to be a good pen app.

Portable. It is not clear how much work they need to do to bring PenPoint up on a non-Intel system. However, we heard that they offered to do this for Motorola and hear rumors of their working on a RISC based PC.

Screen Rotation. They highlight their ability to do screen rotation so the notebook can be used in either portrait or landscape mode. We have a driver under development, but the assumptions of GDI regarding video memory make the portrait driver run very slowly. They support rotation during any session. We must restart Windows.

PenPoint OO design allows for smaller apps. Unclear how true this is, but the OEMs, press and many ISVs buy it.

Connectivity. PenPoint offers deferred I/O and autoconnect to a network.

True 32 bit preemptive multitasking. Their modern OS design appeals to technical industry people. They push the fact that DOS is 10 years old.

Possible Scenarios

The worst case scenario is that Go ships on time, ahead of us. The system is fast and stable and PenPoint apps appear rapidly. The horizontal market for pen computers is very slow to develop and initial sales are primarily through vertical sales force. The larges group of initial users are the mobile workers and the Go incompatibilities are not enough to overcome the perceived ease of use, more focused sales effort and technical superiority of their solution. It turns out these users really don't need keyboards. We lose.

A more likely scenario is that we ship first or around the same time. They have performance problems and their apps are slow to arrive (with a few exceptions) but the system as a whole is stable and delivers as promised. As with most new PC products, the first users end up being the standard early adopters -- [text continues on next page]

Threats to Windows for Pens Microsoft Confidential 7/26/[19]91 page 2

[Plaintiff's Exhibit 859 Page 3 of 4]

[STAMPED] X 189617 CONFIDENTIAL

technically adept people, with Windows familiarity. They will be split in their preferences. Thy appreciate the coolness of PenPoint but understand the benefits of sharing code across multiple platforms. They have desks and they want to use the same OS and apps at their desktop. However, like the early Mac users, some are willing to pay the price of a new OS. We win this market but Go gets a good piece of it and continues to win the technical comparisons. Go continues to focus on mobile workers and as this market develops (more slowly as a result of the long buying cycle and resistance to change of corporate accounts), they get the larger share (due to lack of focused sales effort on our part and IBM pushing PenPoint). As this plays out, the next generation of smaller/lighter Pen machines begin arriving and with Go's focus on scalability and portability, they exploit these machines more quickly and better than Pen Win. They may have the momentum.

What we should be doing

There are more things we should be doing to ensure that we win. I have broken them down into Win 3.1, WinNext (4.0?), and Management issues.

Win 3.1

The Windows group in general has been cooperative but we are caught in the middle of their priorities and frequently can't get things done. There are a few reasonable things that can be done that relate to working well on small portable systems. Since portable systems sales are growing so rapidly, it seems a logical place to invenst in order to combat both OS/2 and PenPoint.

- Scalability. The Win 3.1 group should buy into testing Windows scalability or at least fixing the bugs we report. Assume a small screen display and large system font (640x480 but 1/2 the size so the system font is around 50% larger).

- Deliver Power Management. This is happening, but the Windows group is not planning on shipping anything so we are concerned that this will lose focus in the rush to ship 3.1

- Deliver ROM Windows. This appears under control, but know [sic, "no"] one seems to be addressing the issues relative to executing in place on a Flash system.

- Reduce code size. Is there code that we carry around to run in all three modes that can be eliminated when running only in enhanced mode? Is anyone exploring other possible ways to reduce the size of Windows?

- Power reduction. Windows is a power hog. Are we seriously looking at ways to reduce disk hits?

WinNext

We would love to see some of these things in 3.1 but it is even more important that 3.1 arrive on time so we can beat Go out the door.

- Deferred I/O. We should be able to print without being connected. Ideally, we would have a general purpose output spooling standard. The Pen group can share ownership in getting this defined, but we ned someone in the Win group to work with.

- New Shell. This is coming. Will it be as simple as Go's? Should we explore doing a simpler version for the Pen?

- Instant On. Make it possible to save the state of Windows and the apps so that they can be restored quickly to their previous state. Some ideas were explored for Win 3.1, but there were enough complexities that the idea was killed.

- Quick net connect. We need to be able to quickly configure the system to attach to a network. I shouldn't have to deal with errors when I boot disconnected and I should be able to connect without rebooting.

[page footer] Threats to Windows for Pens Microsoft Confidential 7/26/[19]91 page 3


[Plaintiff's Exhibit 859 Page 4 of 4 ]

[STAMPED] X 189618 CONFIDENTIAL
[STAMPED] X189618

Size. We can't keep getting bigger. Silicon only storage systems will prevail on pen machines and the forecasts for reduction in price are not rosy. What can we do to optimize for a small Flash based system?

- Portability. I hear that NT requires 8MB. This is not suitable for a pen machine. We need to investigate what solutions there may be that provide pen support on small non-Intel PCs.

- Screen Rotation. We should allow dynamic screen rotation and enhance GDI to increase performance in portrait mode on VGA displays.


Management

- Focus from the field. We will need clear support from our field sales organization (through some incentive?) or if that is not feasible, beef up the Pen marketing group (FY 92 has 5 people) to enable our competing more effectively in the direct sales accounts. Many organizations are currently evaluating the competing technolgies and some are starting their prototype development effors. We will be battling the IBM sales force and the Go sales force. Go's existence depends on the success of PenPoint. The Microsoft sales force has little reason to care about Windows for Pens.

- Support. We have discussed offering support to Pen Win end users but the current plan is to offer support only through the OEM. Can we offer free end user support?

- Does it make sense for us to invest more in marketing? Can we raise the ante so Go is forced to spend more of their resources on marketing/selling. Right now, it appears Go is spending more than we are for pen specific marketing. This is based on their strong presence in all pen events, their wealth of marketing collateral and their ability to call on so many corporate customers. They can't keep this up. Assuming this market develops quickly, it will be at least a couple more years before they can be profitable. Based on $50 per unith, they need 200,000 units to do $10mil (ignoring possible apps revenue). Our proposed marketing budget was cut substantially in FY 92 planning.

Some ideas:

- A pen oriented road show that makes it attractive for the early OEMs to participate.
- Figure out how to sell the OEMs sales force on PenWin. We are already offering a lot of training, but maybe we should consider incentives for their sales force or at least provide collateral that makes it easier for them to sell Pen Win.
- Create attractive marketing programs that help our ISVs. Free catalog of apps, create a forum for ISVs to participate (the road show?).
- I am sure others can come up with even better ideas.

Summary

Go has a technically superior solution (if they deliver as promised), they have been sucessful in convincing a lot of people that the pen requires all new apps and that Windows is too complex and inappropriate for small form factor machines. The battle is no longer for OEMs, although we are trying to find ways to stop them, most will offer both solutions. The battle will be at the ISV and corporate/user level. For ISVs we have succeeded in getting them signed up, but the small ones prefer the lack of established competition in the PenPoint market and the big ISVs do not want us to succeed. We are not well positioned to sell or support an OEM product to end users without good support from our OEMs.

We believe we have a great product. The pen user interface under Windows is competitive with Go. We think our recognition will be better than theirs. We will have more apps. But for broader issues such as simplicity, size and applicability for clipboards, we are weaker. We need to put more emphasis on portability/scalability and address the marketing/sales issues or we risk leting Go win this battle.

[document footer]
Threats to WIndows for Pens Microsoft Confidential 7/26/[19]91 page 4

PLAINTIFF'S EXHIBIT 1002 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 1002 p. 1 of 1]

[STAMPED] HIGHLY CONFIDENTIAL

>From toshim Thu Oct 3 05:19:55 1991
To: richardf
Subject: FJ and Poqet ...Pen
Cc: toshim
Date: Thu Oct 3 21:13:29 1991

Want to contact and ask Yonekawa san straightly. FJ (Japan) seems to not only look into Pen Point but have plan to make the PC in parallel with Win for Pen.

Thank you,

Toshi
------------
>From richardf Thu Oct 3 14:38:10 1991
To: toshim
Subject: We have
Date: Wed Oct 2 22:35:40 1991

>rumore [sic, "rumor"?] that Poquet will license Go fpor [sic, "for"?] new systems.
If this is their plan to offer compettive [sic] offering we will not be interested to give Poqet any of theri [sic, 'their'?] prepaids..1.over to FJ license. Nor does it make me very interested in be
richardf
>From jeffr Wed Oct 2 09:18:52 1991
To: joachimk richardf
Subject: RE: Poqet Computer
Cc: pradeps
Date: Wed Oct 2 09:18:23 1991

My source is a consultant to them, but please don't mention that to anyone at Poqet since it would be a problem for the consultant. Thx.
----------------
>From joachimk Wed Oct 2 09:12:54 1991
To: jeffr richardf
Cc: pradeeps
Subject: RE: Poqet Computer
Date: Wed Oct 02 08:49:48 PDT 1991

They are part of Fujuitsi, and have huge prepaids.
i hope they understand, no mercy is an option for us.
Talk to FURAKAWA in J Richard

>From richardf Tue Oct 1 23:36:35 1991
To: jeffr joachimk
Cc: pradeeps
Subject: Poqet Computer

Date: Tue Oct 1 23:36:32 1991

Thi[s] is the first i have heard if tgus [sic "of this"]

this I asked stevemc about this today, he is with me here in tokyo. he has not heard anything of this either,steve will look into tit [sic], seems like we woudl [sic] have heard about it. who is your source ?
richardf

From jeffr Tue Oct 1 22:54:21 1991
To: joachimk richardf
Subject: Poqet Computer
Cc: pradeeps
Date: Tue Oct 1 22:53:43 1991

Who is the account manager? I'm told that they have decided to go with PenPoint for a new pen computer. I may be meeting with the CEO next monday. Thx. Jeff

[STAMPED] Kaplan DEPOSITION 48 4/19/[20]02
[STAMPED] MS 5003997 CONFIDENTIAL

PLAINTIFF'S EXHIBIT 1522 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 1522 Page 1 of 14]


[This appears to be the title page of a manual, or perhaps a sales document or advertising brochure.]
[late-1980's looking computer logo emblazened "Compaq"]

Compaq
PDA Operating System Selection
Portable and Software Marketing
PC Division

Compaq Confidential
Need to Know

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008073

[Plaintiff's Exhibit 1522 Page 2 of 14]


[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008074

Overview

  • Process for launching PDA business
  • Chronology of events and future milestnoes
  • Recommended PDA O/S vendor
  • Executive Summary
    • Potential for market leadership, risks are acceptable, critera for evaluation, evaluation of risk and return, outstanding issues
  • Compaq commitment
  • Summary
  • Backup
    • Microsoft and GO Proposal summary
    • Summary of PDA market opportunity from 11/10 McKinsey presentation
    • PDA relative expected results
    • Market forecasts for PDA hardware and software
    • Technical assessment
    • Product functionality, early market entry, WinPad and Amstel architecture, resource assumptions
    • Microsoft and GO risks and rewards
    • Alternative Approaches

Compaq Confidential [Page Number] 2
1/13/[19]93 (Pda.ppt)


[Plaintiff's Exhibit 1522 Page 3 of 14]


[The exhibit page is a powerpoint slide, showing a Gant chart/timeline arrow from left to right across the width of the page. These arrows are indicated in the transcript using the string - - - >".]

PROCESS FOR LAUNCHING PDA BUSINESS.

[A legend in the upper corner of the chart makes a reference to the right hand half of this table being the subject of] Today's focus


McKinsey presentation November 10
v
Team to complete vision for final review March 1
v
Product announcement 1Q 1994
v

Assess market opportunities

- - - >


Define business and product vision

- - - >

Launch product and build business strategy

- - - >

DELIVERABLES
Detailed assessment of opportunities
Decision on whether to proceed with oppotunity
Build strategic roadmap for launching new business and product

Determine key software and communications skills required, current skill, gaps, and plan for skill acquisition

Develop proposal process for selection of O/S vendor

Choose O/S vendor


Chose microprocessor.
Deliver product to market meeting cost, resource, and time commitments

Implement key business strategies to ensure product success
  • Market acceptance plans
  • Skill acquisition
  • Partnerships
  • Distribution options
  • Technology requirements

cqm002jan156.rjm

[Plaintiff's Exhibit 1522 Page 4 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008076

CHRONOLOGY OF EVENTS AND FUTURE MILESTONES

Assess market opportunities

- - - >

Define business and product vision

- - - >

Launch product and build business strategy

- - - >


[The list of dates below occurs in a giant line-drawing of a "milestone," like the mile markers in the Parker Brothers card game "Mille Bournes" or like the rounded-off-top shape of the tombstones outside English churches such as that in Stratford-upon-Avon.]

November 10 -- McKinsey presentation to Eckhard and Division Management
November 16 Comdex-- Meetings with GO and Microsoft to share general plan to select O/S vendor
November 13- FAX to Microsoft and GO describing details of O/S selection process and proposal topics to be covered
December 3, 9- Microsoft and GO meetings in Houston to review topics and timeline for closure
December 10-Chase Definition Tiger Team Kickoff
December 28- Microsoft proposal received
December 31- GO final proposal received
January 4- Compaq proposal review meeting
January 5, 6- GO and Microsoft presentations at Compaq
January 7- Compaq proposal review meeting, recommendation
January 8- Recommendation review
March- Completion date for Chase Definition Tiger Team

cqm002jan157.rjm

[Plaintiff's Exhibit 1522 Page 5 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008077

Recommended OS Vendor

  • Industry News Flash
    • Compaq selects GO's PenPoint operating system for its Chase PDA
  • Some months later...
  • Industry Headline
    • Compaq announces its industry leading Personal Digital Assistant, the Chase PDA. Chase is the perfect companion for today's business professional. It is loaded with unique features that make Chase stand out from all of today's competitive products. Chase is based on GO's PenPoint, a state of the art, object oriented operating system.

1/12/[19]93 (Pda.ppt)

[Plaintiff's Exhibit 1522 Page 6 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008064

[Star Trek "Starfleet" logo clip art]

Potential Reactions to GO PDA Decision

  • Sabotage our efforts to integrate PDA with the desktop (desktop integration)
  • How?
    • Chicago would break our desktop integration design and implementation
    • Microsoft would withhold Chicago information and DSK
    • Industry compatibility demands should restrict extent of breakage
  • Impact
    • We would have to fix our desktop integration functions
    • Existing example: Novell's DR DOS
    • Risk is manageable.
1/13/[19]93 Pda.ppt
Compaq Confidential
Slide 6

[Plaintiff's Exhibit 1522 Page 7 of 14]


[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008079

["computer terminal" clip art]

Potential for Market Leadership

The potential for Compaq to take a position of market leadership with GO far exceeds oppotunities with Microsoft

A. GO would probide Compaq with significantly greater sustained differentiation potential that would translate into greater market share and higher margines

B. With sufficient Compaq commitment, a GO/Compaq relationship would provide greater product functionality to the customer resulting in faster market adoption and greater market share

C. Compaq could generate significantly potentially greater non-hardware business with GO than with Microsoft

D. Finally, Compaq's ability to work better with GO on a daily basis should help Compaq achieve its goals more quickly and effectively

Compaq Confidential
[page number] 7

[Plaintiff's Exhibit 1522 Page 8 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008080

["computer terminal" clip art]

Risks are Acceptable

Risks associated with a non-Microsoft decision while significant can be mitigated and should be more than compensated by the potential benefits

A. Market acceptance of a GO/Compaq combination although less certain than Microsoft acceptance, could be ensured as a result of a number of factors
  • Friction between ISVs and Microsoft
  • Zero applications for Microsoft migrate to the PDA
  • With significant investment, Compaq could help a critical mass of GO applications
[The last two bulleted items above are underlined by hand, and the handwritten annotation attached:] evangelize ISVs.

B. Although a GO decision might negatively impact Compaq's relationship with Microsoft in other areas, long term cost to Compaq should not be prohibitive
  • Microsoft less focused on a PDA O/S
  • Alternate O/S provider could actually help our position with Microsoft
C. Attempts by Microsoft to sabotage Compaq/GO Integration to the desktop is likely

D. Additional resources (investment and unit cost) are likely to be higher with GO but potential benefits from higher revenues and gross margins should outweigh these costs

E. Although Microsoft could compete against Compaq/GO with other vendors (e.g. HP, AST, DEC, Dell) Compaq's greater product functionality and sustained differentiation should win in the long run

F. GO has enough financial backing to survive for the next 2-3 years and with Compaq support should become self-sufficient by 1994

G. Finally, a GO decision provides a "fall-back" option as long as X86 is the chosen microprocessor, while a Microsoft decision eliminates any near term options with GO


1/13/[19]93 (Pda.ppt)
Compaq Confidential
[page number] 8

[Plaintiff's Exhibit 1522 Page 9 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008081

["computer terminal" clip art]

Criteria Evaluation for PDA O/S Providers

CRITERIA
Weighting
Microsoft
GO
RETURN




 - Sustainable differentiation potential
10
2
4
 - Early market entry
8
3
3
 - Product functionality
7
3
4
 - Non-hardware business potential
5
2
5
 - Organizational fit
3
2
4
 - Total


77

129
RISK




 - Market acceptance
10
2
4
 - Impact on relationship in core business
7
1
3
 - Competitive threat
5
2
4
 - Desktop Integration threat
4
1
3
 - Investment / unit cost risk
3
2
3
 - Hardware OEM
2
1
2
 - Company viability
2
1
3
 - Total


50

101


1/13/[19]93 (Pda.ppt)
Compaq Confidential
[page number] 9

[Plaintiff's Exhibit 1522 Page 10 of 14]


[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008082

Compaq Confidential

GO Expected Business Scenarios

Preliminary



Aggressive / Optimistic
GO Heroic
1997 Assumptions
GO Realistic / Hardnosed
GO
Microsoft
PDA Hardware Market - $Millions
$2700
$2700 $2700 $2700
PDA Software Market - $Millions $2025
$2025 $2025 $2025
Compaq HW Market Share
15%
25%
15%
30%
Compaq SW Market Share 1%
5%
0%
7%
Hardware Gross Margin
20%
50%
75%
75%
Software Gross Margin 50%
75%
75% 75%
Total '93-'97 Investment O/S Decision- $Millions
$43
$23
$10
$23
1997 Results

Compaq Total Revenue - $Millions
$385
$709
$365
$871
Compaq SW Revenue - $Millions
$20
$101
$0
$142
Compaq Total Gross Margin - $Millions
$94
$216
$73
$289
1993 NPV - $Millions
$56
$263
$86
$377


[Plaintiff's Exhibit 1522 Page 11 of 14]


[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008083

EVALUATION OF RISK/RETURN TRADEOFF

[The exhibit page shows the upper right hand quadrant of a two-dimensional graph. The horizontal axis is labelled RISK, and runs from Low to High. The vertical axis is not labelled RETURN, but it is a safe assumption given the context. The vertical axis also runs from Low to High.]

[The area of the graph in the upper left hand corner (high return, low risk) is labelled "Attractive." ]

[The area of the graph in the lower right hand corner (high risk, low return) is labelled "Unattractive."]

[A dotted line runs vertically across the full height of the graph about 40% of the way from low to high risk. Along the horizontal, or "Risk" axis, this rectangular area from zero risk to 40% risk is bracketed with the legend "Compaq typical operating zone." (i.e., Compaq operates well below the average risk, whatever the rate of return.)]

[A solid line runs diagonally from the origin upward and to the right at 45 degrees  along the locus where RISK = RETURN. The solid line is labelled "Normal return." ]

[There is a big black dot about 25% of the way up from the origin on this solid "Normal return" diagonal, as described above, the set of points on the graph where risk and return are equal. The dot is labelled "Microsoft today" which presumably means early 1993. There is an arrow leading straight vertically downward from this dot, pointing to another dot, labelled 1996. Presumably this means that in three years, Compaq predicts Microsoft will have a lowered return, at the same risk.]

cqm002jan107.rjm

[Plaintiff's Exhibit 1522 Page 12 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008084
["computer terminal" clip art]

Outstanding Issues

  • Do the benefits outweigh the risks?
Sustained Differentiation
Market adoption
Initial product functionality
Microsoft retaliation
Competition to Microsoft (customers and ISVs)
Microsoft sabotage
Opportunity for software revenue

Technical Leadership ISV response
  • Can we execute?
Generate application support
Deliver on product vision
  • How much additional non-hardware revenue is possible/realistic?
  • Is this the best use of resources
1/13/[19]93 (Pda.ppt)

[Plaintiff's Exhibit 1522 Page 13 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008085
["computer terminal" clip art]

Compaq Commitment

  • Human resources
  • Marketing funds to compete with MS and their WinPad OEMs
  • GO/ISV investment
  • Geographic divisions mind share
1/13/[19]93 (Pda.ppt)
Compaq Confidential
[page number] 11

[Plaintiff's Exhibit 1522 Page 14 of 14]

[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 008086
["computer terminal" clip art]

Summary

  • Choice: GO's PenPoint
  • High rewards merit risk
    • Key risks
      • Market acceptance and competition with MS and their OEMs
      • Investment level
    • Key rewards
      • Sustainable differentiation
      • Market leadership in new product segment
      • Software business entry
  • We are at a unqie point in time where the future rewards are worth the apparent risks
1/13/[19]93 (Pda.ppt)
Compaq Confidential
12


PLAINTIFF'S EXHIBIT 3174 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 3174 Page 1 of 2]

[STAMPED] KAP0011
[The page is an illustration of a screenshot of a PDA-style user interface.]

[The user interface appears as if it were a tabbed user interface window.]

["Tabs" along the right hand side of the user interface are labelled, from top to bottom:]
Contents Current Products Proposed New Products Sales Customers New Bottle Design

[There is an icon bar along the bottom of the user interface, with icons labelled, from left to right]
Help Preferences Tools Stationery Disks Keyboard Installer In Out

[The user interface has a title bar, which reads:]
Notebook Contents (1)

[The user interface has a menu bar, which reads, from left to right]
Document Edit Create [illegible] Show Sort

[On the menu bar of the user interface, the Create option has been selected, and a small popup or context window appears, with these options from top to bottom.]
Business Memo MiniText [a horizontal separator bar] Section

[The main content pane of the user interface is a tree-structured indented list interface, with a Name and Page for each element. The contents of the tree-structured list interface are, from top to bottom, giving Name and Page number (these are separated by a long ellipsis.......) :]

Current Prod[hidden by context menu] 2
Inventory 3
Standard Order Form 4
Proposed New Products 5
New Product Ideas 6
Capers 7
Chili Mixes 8
Sales 9
Sales by Region 10
New Hires 11
Customers 12
Chain Stores 13
Compaints to Action 14
New Bottle Design 15
Salsa/Condiments 16
Package Design Letter 17
Proposed New Design (3/2/91) 18
Bottle Design Sketch 19
Revised Bottle FAX 20
CONTRACT 21
New Product Draft 22

[The caption for the entire illustration is:]
PenPoint - Table of Contents


[Plaintiff's Exhibit 3174 Page 2 of 2]


 [The page is an illustration of a screenshot of a Microsoft Windows 3.1 style user interface. The caption for the entire illustration reads:]

Microsoft - Notebook Applet

[The title bar of the application reads: ]
NoteBook Page 1

[Below the title bar, the user interface has a menu bar. The menu selections are: ]
Page Edit Go Format Help

[Below the menu bar, the user interface has a toolbar. The toolbar buttons include a selection tool arrow; a row of four buttons whose function is hard to make out (although the first one seems to be depressed, radio-button style, and another one has a squiggle which may suggest handwriting recognition); a button with an icon of a blank document; a group of two buttons, the first of which looks like an SLR camera, and the second of which looks like a file folder; a set of three navigation arrows (back, previous, next?); a pair of text style buttons for Bold and Italic type, and three horizontal text alignment buttons (left, centered, and right.)

Table Of Contents

[A two column table follows, with titles and page numbers: in this transcript, colons have been used to indicate column separation.]

XYZ Co. Mtg: 2
Ideas for Project MoonShot: 3
Scribe Co. Proposal: 4
Scribe Mtg Notes: 5
TechTalk Comments: 6
Sketches: 7
OEM Proposal: 8
Marketing Plan: 9
Page xxx: 10
Page yyy: 11
Page zzz: 12
February 1991: 19
Index: 48

[at the bottom of the user interface is a horizontal row of buttons similar in style to a tabbed dialog. The selections are:]
Table of Contents Index Rolodex Calendar To Do Expenses Customers Ideas Personal





Exhibits published to Court Web Site on Mar 25, 2004 (Incomplete)
Last Revised 4/29/2004


Plaintiff's Exhibit 1520 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 1520 Page 1 of 1]


[STAMPED] Kaplan DEPOSITION EXHIBIT 54 4/19/[20]02
[STAMPED] HIGHLY CONFIDENTIAL Subject to Protective Orders in Coordinated Proceedings Microsoft I-V Cases, J.C.C.P. No. 4106 (S.F. Super. Ct.) and In re Microsoft Corp. Antitrust Litigation
[STAMPED] COMPAQ 007881

Microsoft Corporation [ADDRESS] [TEL] [TELEX] [FAX] [LOGO]

January 13, 1993

Mr. Eckhard Pfeiffer
President and Chief Executive Officer
Compaq Computer Corporation
[ADDRESS]

Dear Eckhard;

Bill and I appreciate the opportunity to meet with you and the Compaq senior management team this Friday. In advance of this meeting, I want to express our apologies if you got the wrong impression regarding our sincere desire to partner with Compaq due to the tone of our meeting in Houston last week.

During our meeting on Friday we hope to accomplish the following:

1. Discuss our desire to improve our relationship in ways that benefit both companies' needs and positions both of us strongly for long-term success. This discussion should include how to provide the industry's best solutions to our mutual customers in the areas of Plug and Play, Chicago and client networking.

2. Put the history of our relationship in perspective. This should include a discussion of how we have helped each other in the past when it came to product development, exchanging information and promoting our relationship with key customers. We would like to fundamentally analyze where and why the relationship has broken down and how we can avoid this in the future. We want to work closer than ever before so that together, we can look even better and promote each others [sic] products in front of  our mutual customers. We hope that these improvements in the relationship will be supported at the highest levels of management in both our companies and that the spirit of cooperation will be a top-down commitment.

3. Further discuss a "Strategic Partnership" Agreement.

Eckhard, I appreciate you scheduling time for us on Friday and personally believe we can reach agreement on the key issues to allow us to further develop our mutually rewarding relationship.

Sincerely
[signature of Joachim Kempin]
Joachim Kempin
Vice President
OEM Sales

cc: Bill Gates
Doug Johns
Gary Stimac

Microsoft Corporation is an equal opportunity employer


[END OF Mar 25, 2004 TRANSCRIPT]


Gordon et. al. v. Microsoft Exhibits Published to Web April 01, 2004
Not Complete
Last Revised 4/25/2004


PLAINTIFF'S EXHIBIT 2348 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 2348 p. 1 of 2 Gordon v. Microsoft

[STAMPED] PLAINTIFF’S EXHIBIT 2348
[STAMPED] GOVERNMENT EXHIBIT 278

From:         Bill Gates [ITG/OXYGEN/billg]
Sent:           Friday, July 07, 1995 9:43 AM
To:             Brad Silverberg; Carl Stock (carts); Paul Maritz (paulma)
Cc:             Richard Tong; Craig Mundie (craigmu); Jim Allchin (jimall); John Ludwidg (johnlu); Jonathan Lazaurus (jonl); Laura Jennings (laural);
                  Roger Heinen (rogerh)
Subject:      FW: Our dinner

I spent 3 hours in San Jose Wednesday night having dinner with Andy Grove.

Fundamentally Intel and Microsoft have not been doing a very good job of working with each other in most areas. There are a few exceptions like some NT related marketing or Plug and Play or TAPI.

I started out by explaining to Andy that for us Windows 95 is job #1. I talked about how we spend over $500M taking calls from end users and use that data to drive our development. I explained how successful we expect Windows 95 to be. I went through how serious we are about testing software.

The main problem between us right now is NSP. We are trying to convince them to basically not ship NSP.

Andy said he did feel bad about a few things. First he felt bad that the software was focused on Windows 3.1 and not Windows 95 or NT. He says that was a mistake. Second he feels like the bad feelings between all the software groups at Intel and Microsoft makes them reluctant to work with us and even makes them hide what they are doing since they think we will crush it. We both agreed we have to start on attitudes and try and get communication going. Andy thinks his IAL people are smart and hard working and he won’t believe there isn’t some way for us to work to get a benefit from what they do.

On the other hand intel feels we are virtually impossible to work with. They feel like if they offered something to use free we would say it is too expensive - they still have bad feelings over being called up and told we were doing Picturetel and paying for it when they feel like they offered to do anything to work together.

I went thru the DCI episode and how that was a step backwards. I went through the problems we have with Intellectual property issues. I told Andy that I think he should cut down the number of software people that Intel has. I got the feeling he doesn’t plan on doing that.

Intel has software in 2 groups - the Frank Gill platform group and the Craig Kinnie IAL group. The Kinnie group is a little over 600 people. Andy says that 200 of the IAL people are the compiler type people who work for Wirt. Andy feels like we asked them to have these people to do hard core compiler tuning to compete with MIPS and he thinks they are complimentary to our work. I didn’t disagree since I think there is some kind of relationship with Roger’s group that works. He says 200 of the people are doing communications related work like NSP. I wasn’t clear what the other 200 are doing. Under Gill there is a communications group run by Gelsinger, a networking group run by Mertz and an Internet group [sic] was just started with 100 people run by Mcgeady. Since Intel is just getting started on Internet, [sic] Andy and I agreed that that is an area we should try and have a very complimentary relationship. We agreed that there should be a high level meeting with Gill to get this kicked off - Paul or Jim Allchin will have to get involved in this. I tried explaining to Andy that our Internet strategy needed to involved getting specific code (in our case) and x86 instructions (in his case) on the Internet.

I told Andy that he shouldn’t be concerned about software not eating up his cycles. I said that MPEG2, object orientation, social interface, 3-d realistic rendering, natural language and speech were all things we are making great progress on that will be popular and use lots of cycles. I said they should stop resisting MPEG2 and they should stop fighting with Rockwell over NSP. I said we are doing amazing graphics work because we have hired the best people in that area. Andy asked if NSP doesn’t do some interesting things that we don’t handle. I said we hadn’t even seen all of what is sometimes called NSP but that the audio related stuff did very little of value. Andy said Intel will not “pay” people to ship NSP by using marketing dollars and they won’t force it onto intel motherboards. He said he is the one who asked them to get a game show off NSP and he said retailers and others he has show the game to have “gasped”. Although Andy is super smart in software related areas and in some aspects of the PC market it is amazing what he doesn’t know. Andy is going to keep his 600 people and he wants us to give them something to do that is constructive. In a spirit of openness he asked who besides Paul was very calm and open minded about technical work with Intel. I wasn’t sure who to say but I mentioned John Ludwig as someone who I totally trusted to let me know if Intel was offering to do things that were constructive or destructive. Andy said he had heard good things about John.

One point I [sic] kept pushing to Andy is that we are the software company here and we will not have any kind of equal relationship with Intel on software. It is in his interest to have us making the hard decisions and doing world class industrial strength work. I kept asking how he would feel if we showed [sic] up with a bunch of microcode to throw into the P6 at the last minute. I said equality cannot be the basis of how we work together on software. Andy understood this but he said that Intel has to push the “platform” forward and they see things we weren’t doing. I said we have to discuss those in advance because their “rabbit” approaches are just destructive. I said I thought it was humorous that it took his 600 people to alienate us so much and cause such poor communication that no one seemed to be talking about how poorly the P6 runs Windows 95 applications. Andy said a second step P6 would run 16 bit operations better – however when he checked on

[STAMPED] MS98 0169009 CONFIDENTIAL
[STAMPED] MS-PCA 1092297
[STAMPED & HAND WRITTEN] [UNINTELLIGIBLE] A5 03/12/02 [ININTELLIGABLE] Stork

Plaintiff's Exhibit 2348 p. 2 of 2 Gordon v. Microsoft

page 2

that Thursday he found out that the P6 second step will still not look good on Windows 95.

On chip relationships Andy asked why we [sic] were being slow to close the P7 framework. I said we were being slow because: a) they didn’t want to use the corporate NDA which simply said they would never sue us over software b) they didn’t want to use the P6 framework and c) they had been super slow getting a proposal to us and d) we were distracted by the [sic] NSP crisis - making sure no one ships that pile of problems. I promised we would get back to them within 10 days on their proposal but that if it would expose us to 338 type garbage we weren’t going to be able to do anything with it. Andy said he recently decided to hire 1000 extra engineers so they can do parallel design work on pure x86 chips and P7 type chips with 2 teams each [sic] for leapfrog development. They like the P7 but they want to be hardcore about both approaches.

On marketing I said it was awful they weren’t helping us in education or with Windows 95. Andy said he is giving a speech at Sun Valley and to a cable and TV event (CTAM?) and at Siggraph and he needs support from us to do his softimage demo and he feels he isn’t getting what he wants. Laura - please call Andy’s office and find out how we can help there. This is urgent. Andy said he wants to focus on solving the software development relationship before we spend our time trying to solve the marketing relationship. He agreed Dennis Carter is stubborn. The one area they want to work together and I agree is to pair NT and P6 as a major phenomena to change computing. Dave House is involved in this. Dave feels we haven’t taken full advantage of his willingness to work with us, Rich Tong should form a strong relationship with Dave and really push it to get as much as we can out of it.

Andy and I talked about their relationship with Compaq and everyone’s fear that Intel is going to take over everything by making all the boards and not letting anyone add value. Andy thinks these fears are overblown.

I personally think we need to try and construct a new approach with Intel where they take on some of the hard problems like speech with us. I think we should figure out some way for Intel to help make Windows 97 more P6 friendly. I think we should get them involved in NT performance.

I believe we will see somewhat less pressure form Intel to ship NSP broadly this year but they want to ship in h1 96. It will take a major effort for us to convince them to back off from this. Andy wants me to talk to him more regularly so I am going to get involved more deeply to find ways for us to work together.

I see Andy again next week in Sun Valley and so I can talk about specific problems or opportunities with him then. I am asking people to have a positive open minded view of how we can leverage Intel’s work.


From:         Andy Grove [SMTP: Andy_Grove@[INTEL EMAIL ADDRESS]]
Sent:          Thursday, July 06, 1995 12:10 PM
To:             billg
Subject:     Our dinner

Bill,

Thank you very much for taking the initiative and the trouble to come down and have dinner with me. I am very happy to have had the chance of a relaxed and in depth discussion with you - in fact, I slept fitfully after that, sort of continuing the dialogue in my sleep. Of course, you had it worse, having to fly home…

This morning I talked with Ron Whittier and Carl Everett about some of the things we discussed. Ron will follow up modifying the Internet sessions to include Frank Gill, and of course will pursue the NSP saga with Paul.

Carl is meeting with Carl Stork this Friday. I hope they can energize our work on the P6. On that subject, Carl (Everett) was unhappy with me telling you that it’s OK to leave the 16 bit code in Win 95. It seems I was overly optimistic on the improvements the next version of the P6 will bring. So, let me retract this comment - and pls. Pass on to Carl Stork that we need all the 32 bit support in Win 95 that we can get.

Again, many thanks for your visit. I hope the cassette tape you were presented with during dinner will turn out to be a worthy investment for you … :-)

Regards,

Andy

[STAMPED] MS98 0169010 CONFIDENTIAL
[STAMPED] MS-PCA 1092298

PLAINTIFF'S EXHIBIT 2374 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 2374 page 1 of 4

[STAMPED] PLAINTIFF’S EXHIBIT 2374
[STAMPED] GOVERNMENT EXHIBIT [UNINTELLIGIBLE]

[INTEL LOGO, DROPPED E STYLE]

INTERNAL CORRESPONDENCE
Intel Architecture Labs

                                                                                                                                                                                
TO:              A. GROVE           D. HOUSE            R. SULLIVAN                                 Date: 8/2/95
                    M. AYMAR          F. GILL                 L. VADASZ
                    C. EVERETT       M. MAERZ           A. MILLER
                    A. YU                   C. KINNIE            S. NACHTSHEIM
                    F. POLLACK       S. McGEADY       S. CHOU

FROM:           Ron Whittier

SUBJECT:     Microsoft Meeting August 2, 1995
                                                                                                                                                                               

1. Processors/OSs

  • Gates/Maritz very interested in understanding the “silicon side” forecast (Is Moore’s law running out of gas?) (CPI?, MHz?). AR: Sunlin Chou/via Rob Sullivan

  • Gates issue: “Fundamental problem with “free” software from IAL, cross-subsidized by processor revenues.” Gates would not agree to let processors/Oss programs to progress unencumbered by platform, communication program issues.

  • Win’95/P6 16 bit problems: agreed to put program on highest priority. Sense of direction was that optimizations of GDI, User (current code base) is the solution space. Gates/Maritz position was that a full 32 bit GDI/User capability was a 48 month deal. So tweaking the OS (and the P6) is the direction POR.
    This will be followed up at the DPG/MPG Summit with Microsoft on Aug 15. (AR: Everett/Yu: Microsoft side is Maritz).

    Make the Aug. 15 session two days to really get results (versus just another metting)… Grove directive.

  • MMx IP issue… basically, will we take a proprietary, closed approach to the MMx instruction set (copyright mnemonics, patents on ISA), or will we go “open” on the ISA and protect implementations.

    A second question is, if “open” when will we disclose [Gates: Q1’96: Aymar Q4’96 with product]? One suggestion was to disclose at the time the tools ship from MS (~Q2’96).

    1

    [STAMPED] MS-PCA1545107
    [STAMPED] MS CID 00077 – Intel Confidential

    Plaintiff's Exhibit 2374 page 2 of 4


    The tool effort itself (at MS) is pretty straight forward (per Maritz).

    AR: We need a policy decision on this quickly (Aymar/Yu); and then a “simple e-mail” to Maritz (Whittier)…by August 7th, please.

  • P7 Contract…agreed that we need to elevate (from Legal) to executive staff level so that the business issues can be understood, resolved. The AR is with A. Yu/P. Martiz assisted by Sullivan/Stork. Close ASAP (in August).

  • Processor/OS Marketing issues

    • P6 – position as a server/WS product with NT – even Gates said 32 bit is OK. Close on detailed positioning (Everett/Stork). Needs to be done in advance of Aug. 15 session.
    • UNIX: big flap – MS wants lots of UNIXes.
    • Enterprise marketing…the server summit should be between House team and [UNINTELLIGIBLE]llchin/Bob McDowell…Marshall Brumer will set up. (AR: House)
  • Long-term Processor/OS Consideration…a Microsoft issue – Martiz to review with A. Yu as part of the Aug. 15 meeting


2. Internet
  • Martz’ presentation covered ground of the July 26 session…agreed MS (Intel and behind, that Internet is a platform [standards engineer end user values, which engenders more standards with more enduser value-“feedback mechanism”], and the key is to win the client (patch up the server later).

  • MS Strategy is to focus on MS Client 96-technologically not tough stuff, but needs to be out “today”, Maritz

  • MS Server is the basic WEB stuff (on NT) plus value add (Back office, SQL, secure transactions, authentication)…integrate Back office into the WEB, giving one-stop shopping for business/SOHO.

  • Issues:

        MS: Very sensitive to what Intel might do on the client side. Example: JAVA, a show stopper. (Later Stork played back “you are doing JAVA”.)

        BG: Supporting certain third party deals will be a problem …we need to consider in the context of their (pervasive) internet program to assure we are not unknowingly stepping on one of their key strategies!

        BG: On the 30/70 use of 3rd party technologies, Intel using Netscape in a Windows environment is not a problem (provided we do not set up the “positive feedback loop” for Netscape that allows it to grow to defecto std.).
    2

    [STAMPED] MS-PCA1545108
    [STAMPED] MS CID 00078 – Intel Confidential

    Plaintiff's Exhibit 2374 page 3 of 4

    page 3


  • What Should Intel Do With It’s Internet Resources?

        BG: “Go do a high end WEB server (super duper server). This could be a tied to their ‘Tiger’ program”. [Or we could go climb a mountain].

        BG: Internet will be deeply integrated into the OS over time, just like messaging, conferencing, etc.

  • Internet Resources, by Company

                    Intel (150)                                         MS (400-500) + MSN (600)
                    NICS (ok)                                         No single focal point
                    Infrastructure (ok)                             Tools (Blackbird, etc) ~ 175
                    Web Server (ok)                                Soft Image (275)
                    High Performance Clients (~ok)       Browser/Shell ~ 45
                    Security/RT Media (not ok)             + Best iNet SQL, Best iNet Content, etc

  • “Intel needs to understand all the stuff MS is doing on the iNET.” Agreed that Microsoft will make a “sales call” disclosing the total scope of their internet program.

    AR: Maritz to set up sales call visit to Gill/McGeady/Maerz

3. Closure of Afternoon Meeting
  • Agreed upon process for defining projects:

    Disclose plans --> Find Holes --> Fill holes

  • On disclosure (the Maritz rules, with McGeady adder):
    - Close to the Platform: MS to come discuss with Intel
    - Close to the OS: Intel to come discuss with MS.
    - Communications (including internet): Maritz to come talk to Intel.
4.  Dinner Discussions – range of topical issues discussed; then serious discussion on various aspect of the relationship and the way we work with
each other. The most important conclusions:
  • On Processes

    1. Regular Disclosure Sessions (Disclose – Holes – Fill)
    2. No “going around” execs (if there’s a problem with Whittier, Gill, etc, address directly with the individual, versus lobbying with other execs to put pressure on the problem owner).
    3. Parallel processing of projects by directly connecting execs in the two companies (see below).
      3

      [STAMPED] MS-PCA1545109
      [STAMPED] MS CID 00079 – Intel Confidential

Plaintiff's Exhibit 2374 page 4 of 4



4. Start treating each other with more respect, i.e. like a valued customer. Also, cut out the non-value added aspect of competing to lead the industry.

Specific AR
Kinnie to review the Win ’97 plan for specific holes (and more generally), and review with Gates directly. It was suggested this be a written review with Face-to-Face delivery. (AR: Kinnie)

  • Parallel Tracks (ESM members to establish their connection)

        Aymar --> Silverberg
        Yu/Everett --> Maritz
        Kinnie --> Ludwig
        Gill/McGeady --> Spix
        House --> Allchin/MacDowell

        We need to exercise these tracks on a direct basis (not via Stork)

  • Misc Comments

    On IP: Gate only trusts Maritz and Neukom, not Stork! (This explains some of our problem.)

    Regarding Product Development …only 16% of resources on actual doing the origional concept development; the remaining is test, QA, productization…we are way out of balance. (AR: Whittier)

    Lots of talk on their program in the video space, including video used in conjunction with internet channel. And they are playing with an investment with Turner. So we need to review our disclosure plans in this area (AR: McGeady)

    Ms8295.doc

    4


    [STAMPED] MS-PCA1545110
    [STAMPED] MS CID 00080 – Intel Confidential


PLAINTIFF'S EXHIBIT 2418 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 2418 Page 1 of 1

[STAMPED] PLAINTIFF’S EXHIBIT 2418
[STAMPED] GOVERNMENT EXHIBIT 281

From:                    Bill Gates
Sent:                     Wednesday, October 18, 1995 11:19 AM
To:                        Brad Silverberg; Carl Stork (carst); Jim Allchin (jimall); Paul Maritz (paulma)
Cc:                        Bernard Vergnes (bernardv); Craig Mundie (craigmu); David Cole (davidcol); Jeff Raikes
                             (jeffr); Joachim Kampin (joachimk); John Ludwig (johnlu); Jonathan Lazarus (jonl); Nathan
                             Myhrvold (nathanm); Paul Osborne (Paulo); Pete Higgins (peteh); Rich Rashid (rashid);
                             Richard [UNINTELLIGIBLE] (richta); Steve Balmer (steveb)
Subject:                 Intel Grove - some opportunities to work together!

1. MMX! (bradsi, craigmu, carts)
    Intel has instruction set extensions to allow for higher speed in multimedia operations called MMX. All Pentiums coming out after summer 95 (P55C) and all Pentium Pros coming out after Fall 96 will include these instructions. Intel thinks they will have 200mhz Pentium Pros in 96, Intel has a plan to fix the 16bit performance of Pentium Pro but not until sometime in 1997. Intel has decided to make the MMX instructions available to the people who do x86 clones. Intel is expecting support from Microsoft as it goes public with these instructions in February 96. I am assigning Brad to prioritize more discussions with Intel about MMX – where can we use these improvements? For example: we should study Intels MPEG2 software using MMX and understand if software only video conferencing becomes realistic. Even though MMX isn’t perfect in its first release we need to support Intel on this one and help them evolve it. If we really don’t like it we need to be very articulate NOW. Our multimedia group in PSD and Craig’s video group under Jay Torborg need to take a hard look. It is past the point of major change for Intel. With Intel’s share I believe the best thing for us is to push everyone to adopt MMX even though the alternative may be technically superior. Intel would like elements of Nashville to visibly support MMX even if its mainly a marketing thing. If MMX enables video we should consider how our content and applications should use it.

2. NT/Pentium Pro (steveb, jimall, paulma, craigmu, jon l)
    Intel wants to work with us on marketing the NT/Pentium Pro combination. Intel has an architectural sales force that calls on large accounts. They are willing as part of a cooperative arrangement to have this group focus on promoting NTS on Pentium Pro servers and NTW on Pentium Pro desktops. Grove thinks we could really help each other on this. We would both still have modest activities where we are processor independent and they are OS independent but they are willing to have almost all of their message focus on the combination of our products. I think this is the first time where we can really leverage Intel’s marketing investment to help a Microsoft product. We might even be able to get Intel to help us focus on tough accounts like National Security Administration or other SUN accounts. We might be able to get Intel to help us focus on Publishing and Higher Education.

   Under the new organization Intel has 2 relevant groups – Dave House for servers and Carl Everrett for Desktop. Grove wants to increase the bandwidth between BSD and these groups. Grove wants to know personally things that hold us back for working closely with these groups to deliver a strong message. If we want help on tuning, benchmarking, working with particular ISV’s or customers Grove thinks Intel can be supportive. Intel is actually doing 2 launches for the Pentium Pro – one by each of these groups.

   Intel as a customer has decided to skip Windows 95! They have a roll out plan to go straight to Windows NT. Grove says they put a lot of time into thinking about this. They don’t want to go public on it yet. I am actually surprised personally at the number of corporations looking at going straight to NT.

3. Real time Communications (paulma, craigmu, bradsl, jonlu)
    Paul – Intel feels we have all the OEMs on hold with our NSP chill. For example they feel HP is unwilling to do anything relative to MMX exploitation or the new audio software Intel is doing using Windows 95 unless we say its ok. This is good news because it means OEMs are listening to us. Andy believes Intel is living up to its part of the NSP bargain and that we should let OEMs know that some of the new software work Intel is doing is Ok. If Intel is not sticking totally to its part of the deal let me know. Andy knows you are down seeing Albert Yu on October 2[UNINTELLIGABLE]. He will grab a short meeting with you. He think Intel still has a lot of strength in “real time communication” and that combining this with the great strength he fears we have developed in graphics/video is the key to getting interactive applications on the Internet to require more than a simple terminal. He wants us to recognize that despite problems in the past combining our strengths here makes sense. He says the key groups involved here now work for Gill. He is willing to lay out everything they are doing. He understands telling us at the last minute does not work.

[STAMPED] MS98 0169058 CONFIDENTIAL
[STAMPED] MS-PCA 101686
[STAMPED] EXH. K[OR X]8 DATE 03/12/02 WITNESS STORK SUSAN ZIELS


Plaintiff's Exhibit 3667 - Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 3667 Page 1 of 8

[STAMPED] Government Exhibit 920
[STAMPED] DT EVOL, SLIDE 1                                        DC 75
[STAMPED] MS-PCA 1020245
[STAMPED] MS CID 00208 Intel Confidential
[The exhibit is a Powerpoint-style presentation with a background logo that says] ADL ARCHITECTURE DEVELOPMENT LAB  INTEL CONFIDENTIAL
Destop Platform
Goals/ Strategies
  • Grow the Market for Intel Architechture PCs
  • Requires new Media and Comm based Apps
  • Requires a new Common Baseline of media and comm functions in mainstream PCs
  • New Common Baseline must be easy to use and manage
  • New Common Baseline must fit within cost constraints of today's $2K MM PC

NSPRP = NEW COMMON BASELINE
NSP = KEY TECHNOLOGY FOR COST AND EOU


Plaintiff's Exhibit 3667 Page 2 of 8

[Still present is the background logo that says] ADL ARCHITECTURE DEVELOPMENT LAB  INTEL CONFIDENTIAL
[STAMPED] MS CID 00209 Inter Confidential
[STAMPED] MS-PCA 1020246
NSP Reference Platform
[Beneath and to the left appears a down-pointing arrow-box with the logo of a CD  and an illegible word written in a clip art starburst shape over the upper right hand corner of the arrow itself.  Within the arrow-box  are written the following:
Open Design Guide
Industry Work Groups
SDKs
DDKs
[To the right of the arrow are the following:]
ISV's
Media & Comm Based Applications
Base Capabilities
OEM's                    NSP Reference Platform
[ with check marks next to each of the following, excepting "Scalable Performance"]
Scalable Video                  Transparent Connectivity
3D Graphics                      Plug and Play
Sharable Audio                  Manageability
                 Scalable Performance
Balanced Partitioning
[Here follows a box containing a picture of a plain seesaw-looking balance that weighs on the left two hard disks and a pentium processor logo equally with (on the right) a square with the words, ' Signal Processing H/W" in a square with multiple squares indicated behind it of equal size.  The letters "IHVs" are in the upper left hand corner of this box.]
[STAMPED] DT EVOL, Slide 2                        DC '95


Plaintiff's Exhibit 3667 Page 3 of 7

[Still present is the background logo that says] ADL ARCHITECTURE DEVELOPMENT LAB  INTEL CONFIDENTIAL
[STAMPED] MS CID 00210  [and barely decipherable the words:] Inter Confidential
[STAMPED] MS-PCA 1020247
Intel- MS Platform Engagement Summary
  • Beginning in late 1991, Intel shared with MS a long range application and platform vision as well as detailed requirements for system software.
  • We were not able to engage on any meaningful joint development on any of the goals identified.  We did get active resistance/competition from MS business units.
  • Intel began to engage directly the relevant groups at MS to develop the eight key areas that now constitute the  NSPRP.  (PCI, DCI/Indeo, TAPI, P&P, DMI, 3DR, Native Audio, NSP)
  • Intel developed the key standards and driver level S/W required to implement all eight areas in Windows 3.1.  MS has publicly supported 6 of these 8 areas at one point or another.
  • Only three of these areas are scheduled to be supported by MS in the first release of Win '95 in a compatible way.  (PCI, P&P, TAPI)
  • There are many cultural, strategic and legal issues that cloud our relationship, but the fundamental issue is that MS firmly believes that the largest developer of Pentium Processor based platforms has no business developing platform level software!
  • Intel has proven its ability to lead the PC platform evolution and develop platform level software.  It is fundamental to our business that we continue to do so.   
[STAMPED] DT EVOL, Slide 3                          DC    '95

Plaintiff's Exhibit 3667 Page 4 of 7

[Still present is the background logo that says] ADL ARCHITECTURE DEVELOPMENT LAB  INTEL CONFIDENTIAL
[STAMPED] MS CID 00211   Inter Confidential
[STAMPED] MS-PCA 1020248

Intel - MS Platform Engagement
What joint development would look like

  • MS acknowledges/accepts Intel's leadership role in defining PC platform standards and developing software.
  • Cooperation on establishing a new common baseline Win '95 Desktop PC.
  •        NSP/Native Audio:
             -MS public endorsement of NSP.
              -Define joint development plan for MS support of Native interfaces in future Win '9x and Win NT.
             - Define Intel support of Direct Audio interfaces on Native Audio.
  •       DCI/Indeo:
             - Joint development/ evolution of DCI/Direct Draw interface.
              - Support of Indeo V4 in future Win '9x and Win NT releases.
  •       DMI:
             - Joint development and support of 16 and 32-bit DMI MI and CI interfaces in Win '95 and Win NT
  •       3DR:
             - Acknowledgment of coexistence on RM and 3DR interfaces.
             - MS to port RM onto 3DR libraries.
             - MS to freeze 3D-DDI spec.
             - Intel to evangelize 3D-DDI interface for graphic drivers.
  •       Instant On:
             - Acknowledgement by MS of Instant On as valid Win "95 applet.
             - Joint development and evolution of APM spec. 
  •       Continued cooperative development of TAPI, WinSock 2, USB and PCI.
 [STAMPED] DT EVOL, Slide 4                                        DC '95

Plaintiff's Exhibit 3667 Page 5 of 7

[EXACT REPLICA OF PAGE 2]

Plaintiff's Exhibit 3667 Page 6 of 7

[EXACT REPLICA OF PAGE 3]

 Plaintiff's Exhibit 3667 Page 7 of 7

[EXACT REPLICA OF PAGE 4]

PLAINTIFF'S EXHIBIT 4410 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4410 p. 1 of 2

[STAMPED] PLAINTIFF’S EXHIBIT 4410
[STAMPED] GOVERNMENT EXHIBIT 283

[INTEL LOGO, DROPPED E STYLE]                 Internal Correspondence                         -Intel Conf

To:             List                                                                                                 Date: 8 November 1995

From:         Russell Barck, 408-xxx-xxxx [PHONE NUMBER REDACTED]
                   Frank Ehrig, 206-xxx-xxxx [PHONE NUMBER REDACTED]

Subject:     Summary: 11/7 Kinnie/Gates Feedback Session                                             cc:

Attendees:         Intel                                MS
                         Craig Kinnie                   Bill Gates
                         Gerald Holzhammer        Paul Maritz
                         Frank Ehrig                     Carl Stork
                         Russell Barack                Marshall Brumer
                                                                David Cole

Summary
Overall a very good session. Craig did an excellent job of constructively identifying and addressing underlying differences in the manner in which each company translates its vision into requirements and diffuses its technology to the industry. Lack of detailed MS product schedules was highlighted (and acknowledged by Bill) as a major issue that needs to be addressed to enable successful engagement. Also highlighted was difference in fundamental philosophy: Intel Focus=Demand Creation, MS Focus = Demand Fulfillment. Concerns over acceptable quality levels surfaced as a result of divergent focus - Martiz reinforced desire to make “quality” primary goal of any joint activities. MS places the quality imperative first and foremost in their efforts to improve the end-user experience.

General discussion was open, honest, and free-flowing. Atmosphere was one of “genuine desire” to fix problems that lead to rift in the relationship. Craig’s feedback was received well by MS with a few exceptions - most notably MS’ lack of active new app development focused on driving new platform capabilities and MS’ generally reactive approach to incorporating OEM and IHV input in their products. Several key areas for follow-up and further engagement were identified: Realtime Services, Native Audio, 3D API/DDI (which Maritz finally agreed to release to Intel and 3DR-team), VBI/Intercast, S/W Midi, DMI/RDMI, USB, Surfboard/Schroeder, POTS-VC, and Dino.

Bill concluded the session by saying that he felt it was “super valuable” and that he would like to sit down with same group in 4 months and review the progress. He also commented that it would be extremely beneficial if Intel would disclose our software activities to MS as soon as possible.

Insightful quotes from Bill:

  •  “Graphics is the single greatest area that MS is investing in” [referring to Talisman project]
  •  "Intel wants to push [the market] - this is very good thing”
  •  “IAL having 700 software engineers running around in the industry is an OK thing, as long as MS know what they’re doing first!"
  •  “These are areas we should definitely spend more time with each other in” [referring to Craig’s overview on Surfboard/Schroeder, Teladdin,  POTS-VC, and VBI/Intercast
  •  “Today’s API is tomorrow's DDI” [MS doesn’t want to relinquish control over APIs]
  •  “Sounds like Intel’s motherboard strategy!” [tongue-in-cheek - referring to creation of Universal Driver Model]


[STAMPED] MS CID 00128 - Intel Confidential
[HAND WRITTEN] BARCK #1

Plaintiff's Exhibit 4410 p. 2 of 2

[PAGE 2]

Focus Areas to Advance the Platform
  •     RTS & RTMS in Win9X and NT - Define clear roadmap to the industry on how we converge
  •     3D API and DDI - Provide clear, scaleable interface direction to the industry
  •     DMI & RDMI - Implement common roadmap launched at DMTF
  •     USB - Need commitment for aggressive driver development and device support
  •     Video Capture and DDI - Need clear direction in Quartz 2.0 for Video Capture


Bill commented that timeline is the real issue on RTS - it could potentially be 2H96. He stressed (again) need to share detailed schedules so we (Intel) can do analysis of whether or not it’s worth waiting for, MS asked to add VBI, Native Audio, Dino, and Soft MIDI to list of short-term engagement items.

AR Summary/Next Steps
AR:     Stork/Kinnie - Explore renaming NSP to something with less negative industry equity
AR:     Barck - Coordinate disclosure of POTS VC, Surfboard/Schroeder, HHD, and Teladdin to MS similar to SIPC presentation to Intel)
AR:     Cole/Kinnie - Identify appropriate contacts within respective companies to initiate early disclosure/engagement
AR:     Brumer/Barck - Coordinate MS SW Midi disclosure/evaluation
AR:     Brumer/Ehrig - Coordinate UDM “Sales Call” to Intel - use USB as model
AR:     Brumer/Landsman - Coordinate Talisman disclosure to Intel/Resolve open MS IP issues
AR:     Brumer/Ehrig - Coordinate discovery meeting with MS Research to discuss MIPS-intensive apps
AR:     Brumer/Barck - Provide detailed 6 month roadmap and schedules on MS deliverables
AR:     Brumer/Barck - Coordinate follow-on SIPC presentation to Intel w/next-level of detail
AR:     Maritz/Ehrig - Release 3D DDI to Intel (no caveats on 3DR exposure)
AR:     Ehrig - Coordinate MS visit to Hillsboro for USB Driver Debug/Development
AR:     Brumer/Barck - Provide Intel with plug-in point for Quartz work
AR:     Maritz/Kinnie - Explore less-restrictive driver model implementation.
AR:     Barck - Initiate and manage engagement on Top 7 issues

[STAMPED] MS CID 00129 - Intel Confidential


PLAINTIFF'S EXHIBIT 4411 Gordon v. Microsoft   [ Top ]

[Plaintiff's Exhibit 4411 Page 1 of 2]

[STAMPED] PLAINTIFF’S EXHIBIT 4411
[STAMPED] GOVERNMENT EXHIBIT 563


From:         Gerald S Holzhammer [Intel email address]
Sent:           Thursday, April 13, 1995 2:40 AM
To:              Gerald_S_Molzhammer [Intel email address]; Claude_M_Leglise [Intel email address]; Bill_R_Miller [Intel email address];
                   Robert_Sullivan [Intel email address] Ronald_J_Whittier [Intel email address]; Craig_Kinnie [Intel email address];
                   mcg [Intel email address]
Cc:              Ken_Rhodes [Intel email address]; Murali_Veeramoney [Intel email address]; Joe_Casey [Intel email address];
                   Frank_T_Ehrig [Intel email address]
Subject:       Microsoft Face to Face Summary

Text item: Text_1

Bottom-Line:

- Based on W95 experience MS finds that they need to own all drivers
- Expect no real cooperation from MS on NSP or media; input/education is welcome, of course
- At Games developers conference focus on “Design scalable games for Pentium” message; DON’T make a half-hearted attempt to push Native
   audio without a compelling ISV story (current POR). MS messages are tuned for this audience; we’d come across defensive at best
- Bill, let’s do the Native Audio press release before GDC; it won’t harm our MS relationship :-)

Gerald

Details:

We met with Carl Stork, Marshall Brumer, and Eric Engstrom for 3 hours. WinHEC clearly opened their eyes regarding the scope of NSP. They are upset with us being in “their” OS space - no surprise there. Interestingly, they have evolved their thinking on what their “space” needs to be.

1-MS moving towards vertical integration of driver SW.

In short, they feel they need to own ALL driver software “to the metal” i.e. silicon - a significant shift from owning the core OS and allowing innovation at the driver/HW level. A recurring theme was that nobody but MS is qualified to do good driver SW. In their mind, W95 was delayed 9 months by 3rd party driver dependence. That this means for the games SDK is that they are focusing only on the ISV (API) aspect and all but ignore the IHV (HAL/DDI) aspect. They intend to provide ALL drivers for all interesting HW - a fallacy given the horizontal nature of the PC industry. Nonetheless, they are serious. (If they follow through with this, innovation at the HW level would grind to a halt since silicon vendors would need to rely on MS to get driver support for their new stuff. The good news is that IHVs would have to flock to Intel supported open NSP DDIs as the only way to innovate without depending on MS driver support for their widget.)

2- Cooperation - NOT!
Stork openly admitted that MS has completely missed the boat on developing a compelling state of the art media subsystem for Windows95. But this is high priority for W96/97. He eluded to an upcoming reorg that may combine Osborne’s and Ludwig’s team to develop the next generation media architecture (nothing concrete yet but it will build on the Games SDK). Intel is welcome to provide input - but they own it down to the silicon :-)

3- Games Developers Conference

I don’t expect MS to openly attack NSP at the GDC, BUT their thrust is clearly counter NSP. Eric went as far to insist that “any DSP

[STAMPED] 477MSCID00235 – Intel Business Information – Highly Confidential

[Plaintiff's Exhibit 4411 Page 2 of 2]

page 2

regardless how poor in performance is better than doing processing on the host - since the game is left with all of the CPU” (something games ISV’s are very vocal about). Amusingly, Direct Audio has an NSP-like SW emulation layer if the HW does not support a certain feature. This is not an inconsistency in their mind, of source, since MS will provide this SW.

We should be watching out for a games-enabled ISA card/reference design from MS that provides advanced audio, 3D, and 2D support. Performance is achieved the old-fashioned way: lots of acceleration and memory behind a slow ISA bus.

4- DCI primary surface support makes it into W95; offscreen gets rearchitected later.

According to Engstrom the offscreen aspect of DCI 1.0 cannot be reliably supported in W95 - they were not paying attention when DCI 1.0 was implemented. (I’m sure he is totally objective here given that DCI 1.0 was done by his predecessor jointly with Intel) He claims the DCI drivers provided by IHVs are buggy and violate the Windows programming model. He is paying lip-service to MS providing a Direct Video incarnation of the DCI 1.0 draw handler to support off-screen surfaces in the future but since video is not part of the games SDK this remains at risk. And, NO, Intel cannot help since this code has to be integral to W95.


Games SDK Schedule (I expect this to change)
Beta 1 Games Dev Conference
Beta 2 July
FCS September

[STAMPED] 477MSCID00236 – Intel Business Information – Highly Confidential



PLAINTIFF'S EXHIBIT 4559 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4559 p. 1 of 2

[STAMPED] PLAINTIFF’S EXHIBIT 4559
[STAMPED] GOVERNMENT EXHIBIT 284

S. McGeady
From:         Steven McGeady [mcg]
Sent:          Thursday, November 09, 1995 5:43 PM
To:             Frank Gill
Cc:             Andy_Grove@[INTEL EMAIL ADDRESS]
Subject:      Re: marritz meeting summary

I interpreted [sic] our meeting today consistent with Frank’s last paragraph. We now have completed our AR and will begin the Java program full force, consistent with the direction to do so without public disclosure.

I take minor exception to the “insurance policy” positioning. I’m OK with that for MS, but Java isn’t an insurance policy for us at this time – it’s an important component part of a major application strategy. However, I don’t think that this needs to be resolved.

I’m assigning about 5 existing ITL people to this, opening reqs for another 3-4, and engaging with IAL’s media suppliers to budget resources to include support in their media components, all per the GSR outcome.

I take an additional AR to continue to educate Frank on the merits of Java and the position and motivation of Sun.

On other topics, Richmond will get AR to “win” MS for Intercast architecture convergence at (virtually) any cost,

mcg


----------- Original Message Follows

>From: Frank Gill Frank_Gill@[INTEL EMAIL ADDRESS]
>Subject: marritz meeting summary
>---------
>
>    overall very good but java remains a major controversy. Paul
>    really appreciated the spirit of openness and strong desire to
>    cooperate. Major points summarized below.
>
>    -on intercast and their direct tv program, we both felt we are
>    early enough that we can cooperate to make these programs
>    complementary and even divide up the work. i want to really make
>    this happen and demonstrate [sic] we can do a successful program
>    together.
>
>    -we told them again we are using netscape in first release of our
>    internet-in-a-box and reasons why. while disappointed, they
>    understood and paul told me after “this is how we should handle
>    things with no surprises”
>
>    -re java, we told them that we felt java is on the way to becoming
>    and internet std and felt a need to optimize around IA and
>    particularly optimize our media components for the java
>    environment. they see this as supporting their mortal enemy….and
>    argue SUN is our enemy as well. Paul describe to me as “this is
>    equivalent to us optimizing for powerpc”. i left it with him that

[STAMPED] 475MSCID00015 – Intel Business Information – Highly Confidential

Plaintiff's Exhibit 4559 p. 2 of 2

[PAGE 2]

>    have not starting anything until we talked today, we will not publicly endorse java, and i needed internal review before authorizing this work to >    commence. So, bottom [sic] line, this is BIG BIG DEAL to them.

>    i am still pretty dumb in this space, but think we should do the
>    Work quietly as an insurance policy and see a few more cards
>    before any public disclosure. Lets talk. frank
----------- End of Original Message

[STAMPED] 475MSCID00016 - Intel Business Information – Highly Confidential

PLAINTIFF'S EXHIBIT 4560 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4560 p. 1 of 2


[STAMPED] PLAINTIFF’S EXHIBIT 4560
[STAMPED] GOVERNMENT EXHIBIT 285

S.McGeady                                                                                                                                                            

From:        Dave Landsman [Intel email address]
Sent:          Monday, November 13, 1995 7:01 AM
To:            [Intel email addresses]
Cc:            [Intel email addresses]
Subject:     Draft Notes from Gill/Maritz Summit on 11/9


Attorney/Client Confidential

Here are my notes from the Gill/Maritz summit on 11/9. Give me any comments and let me know if you’d like me to publish a full set of minutes, with more detail.

Attendees from MS:
Paul Maritz – Exec VP Platform Division
Brian Moran – Program Mgr – Broadcast PC
Craig Fiebig – Director Biz Strategy – WW Strategy Group
Marshall Brumer – Acct Manager
On Phone for Realtime Discussion:
- Rich Rashid – VP Research
- Paul Leach - ??
- Yoram Bernet – Senior Architect – Multimedia Server


Key Issues:

Java. We told MS that we were seeing this become enough of a factor in the market that we felt the need to fund efforts into performance tuning. Frank stated that we would not evangelize publically. Maritz reacted strongly to this, saying “it is premature for Intel to do this, there are very few apps being written for Java today, and nobody is going to make an architecture decision based upon Java”. MS’s stated concern in this area is not the Java language, per se, which they will support in their products as “just another language”, but Java component object model. MS wants Intel to support OCX’s and their object model, which is integral to their OS platform. Maritz said he would try to convince us to spend our energies elsewhere.

Server. We told MS that we are using non-MS products (except for NT) in the initial [sic] version of our server. Reasons: 1) Our OEM’s are asking for the products we are using; 2) MS’s does not have a complete product set, specifically with respect to integrated HTML authoring and document management; and 3) MS has not treated us as a customer. For #3, we cited a number of issues. For example, we still don’t have ITL and ISD people registered as Gibraltar and Blackbird users, and that we only got Gibraltar through Alan Holzman, who happened to get it at an Internet briefing he was invited to, to which ITL and ISD were not involved. Bottom line: they are not treating us as a customer. Maritz said he would “fix it”.

Where we can work together. Maritz identified two key areas for working closely together: 1) Integration of Internet and voice/video/telephony; and 2) 3D, specifically Talisman, within a month. On #1, we did not identify a specific AR, so AR is for ICG to identify how we proceed on this and have MS team close back with Brumer for who/how to proceed at MS. Many discussions are already underway around this topic. Perhaps we also need to roll up a vision here and present back to Maritz as a proposal, or ask him for one. One #2, Landsman has already been in touch with Jay Torborg. Torborg told me, as of 11/8, that he is waiting for Maritz’s go ahead to engage. So AR is: Landsman to get relevant Intel folks (PCD/OPSD/IAL) engaged with Torborg’s group.

                                                                                                         1

[STAMPED] 457MSCID00040 – Intel Business Information – Highly Confidential

Plaintiff's Exhibit 4560 p. 2 of 2

page 2

Other Topics

MS Overview of Internet Plans. Maritz gave an overview of their Internet plans that essentially stated the same info we heard in Redmond in meeting on 8/30. They plan to integrate the overall browsing environment into windows by having the right tools, runtime support for various content languages (including their own native document types, like Word and Excel, as well as 3rd parties, like Java, etc). Maritz said he would do an update of these plans to us in Dec/Jan, once they have gotten to the next level of detail.

Real-Time. BSD is working on RSVP on Winsock2 sometime after the shell release of NT. Rashid is working on real-time as well, in context of Tiger. They agreed to have Abel to meet with R. Rashid (Research) and Dave Thompson (NT) on real-time. AR: Weinrib/Landsman

Intercast. Maritz is very positive about Intercast. They want to continue the engineering discussions. He broke the topic into Windows platform issue and content. On platform, they want to continue the engineering discussions: “we do the DBS stack, you do the VBI stack”, and make sure it is supported well in Windows. As for standards group: “let’s synch on the engineering and that will take care of itself”.
Content was kind of [sic] left off the plate in this discussion.
AR: Haight/Landsman – Followup on Intercast engineering discussions.
AR: Gill/Maritz – followup discussion in approximately one month toreview progress.

Other Stuff

I noted a comment that “DirectX will not be in the Shell release of NT”.
                                                                                                       2

[STAMPED] 457MSCID00041 – Intel Business Information – Highly Confidential


PLAINTIFF'S EXHIBIT 4562 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4562 p. 1 of 1


[STAMPED] PLAINTIFF’S EXHIBIT 4562
[STAMPED] GOVERNMENT EXHIBIT 565

[TRANSCRIBERS NOTES: AUTHOR USES MANY ARROWS AS BULLETS, WHICH ARE NOT REPRODUCED. ARROWS THAT CONNECT SENTENCES ARE NOTED]

[HAND WRITTEN NOTES]

[STAR] Microsoft Java Meeting - 4/9/96

Rewritten the VM [ARROW] Goes back to Sun
        Plan to ship broadly w/Inet Explorer
        Natively knows how to talk to COM

Should work together to define APIs for Java Media
        “low-level manifestation” should be through COM object ifaces embedded

VM
        COM object interfaces / will publish these soon [ARROW] Slightly before Beta 
                ifaces:                                                                         mid may/end of May
                            JIT
                            Byte code Invocation
                            native methods
                            [ARROW CONNECTS ifaces GROUP TO] Rob Wellend /John Thomason
                                                                                                  Mike Tutongie[NAME?] [STAR]

“Broad distribution rights”

Ludwig [ARROW] Media Interface
        Dean McRory – in Muglia’s group
        Rob Wellend – Ludwig
        John Thomason – “
        “Joel Spegal’s [NAME?] organization”

IE 4.0
[UNINTELLIGABLE] – Fall ’98 // IE3.0 – This Summer

[STAMPED] 477MSCID00271 – Intel Business Information – Highly Confidential

PLAINTIFF'S EXHIBIT 4563 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4563 p. 1 of 2

[STAMPED] PLAINTIFF’S EXHIBIT 4563
[STAMPED] GOVERNMENT EXHIBIT 566

From:         Alan J Hotzman [email address]
Sent:           Friday, April 19, 1996 2:56 PM
To:             [email addresses]
Subject:      MS/Intel Java Meeting Minutes – 4/18/96

Text item:

Trip Report from Alan Hotzman & Burt Perry:

MS/Intel Java Meeting 4/18/96

Attendees
Bob Muglia, VP Tools                                        Microsoft
John Ludwig, VP Interfaces                                Microsoft
John Thomason, Program Mgr                            Microsoft
Mike Tatongy, Prject Lead – Java                      Microsoft
Alan Hotzman, Strategic Partners                        Intel
Burt Perry, Eng Program Mgr                             Intel
Greg Eastman, VM Project Lead                        Intel
Frank Ehrig, SSTM Mgr                                     Intel

MS will take part in the Sun 4/30 announcement of Java OS platforms
MS VM built from spec, no Sun code
MS has changed native interfaces. Not compliant w/Sun’s current native
interface at the binary level:

    As part of rewriting the Java VM, MS has completely changed the
    internal object model to accommodate COM.  We think they have not
    told Sun and this may be an issue for Sun as well as us if Intel and
    MS give a single optimized IA Java RT back to Sun. The current
    method for calling native methods in the MS Java VM is different and
    not compatible with Sun’s. MS will provide include files that allow
    you to bridge between the two al the source code level but not at the binary level.

MS believe their VM is 2X faster than Netscape
Discuss w/Clary. how can we work w/Microsoft ?
        Get Sun permission
        Tell Clary to call Muglia on us working together if there is a problem with Clary agreeing to it
        Get a letter from Clary giving us permission to work together

MS has not done interpreter optimizations and is very interested in working together to produce a single reference implementation for IA Java on Windows.
The IE 3.0 timeframe of June is not possible for integration of Intel &MS VMs, but IE 4.0 in fall is IE 3.0 will have new MS VM in it through.
        IE 4.0 (as part of Nashville) is right target to combine MS & Intel VM work
Muglia/Ludwig want exclusive access to IA’s VM work:
        Ludwig does NOT want us to give Netscape our VM work
        Netscape access to VM is VERY touchy w/MS!!!
        Ludwig wants us to work through Microsoft, not got to Netscape unless we were unhappy with MS cooperation (or lack thereof). Let Ludwig know we’re unhappy BEFORE we got to Netscape

Issue: If we work with MS, they want us to start with their new code. Technically this is not a problem but when we jointly give Sun the code, they will be astonished because it will be totally different

[STAMPED] 477MSCID00026 - Intel Business Information - Highly Confidential

Plaintiff's Exhibit 4563 p. 2 of 2

page 2

than their source base and will have no shared code with any other implementation. Rather, it will be Microsoft’s architecture. How will Sun feel about this and our participation in this (Wintel all over again?).

A Java JIT will be standard in the OS from MS. Will always be there, make decisions on what to JIT and why
        Borland doing the same with Netscape!

Class Libraries
MS has integrated COM as inherent part of Java VM so they want all classes to simply be COM interfaces. Plan to develop cross platform (Windows, Mac, UNIX) class libraries (built in native code, not Java) for: Media and graphics, Data building / data access, forms.

MS is considering a UNIX client for IE
Software AG is/WILL port COM & DCOM to all UNIX platforms
MS assuming that class libraries will be anti-MS coming from Sun no matter what!

        thus, get ISVs & developers to write to MS class libraries
        Get anti-MS developers by porting COM to the MAC etc.

Forms class libraries were THE most important to MS
Can we work together on defining Java Class libraries “outside” of the Sun Class Library definitions ???
Muglia wants to work with us on defining media classes that work directly with COM


BOTTOM LINE:
They want developers writing to their APIs not Sun’s Java APIs and strongly want us to rethink what we are doing. They want to work with us to define media class libraries that we are happy with. Ludwig will give us a date for their Media Class Interfaces Specs on Monday 4/22 after he meets with the engineers.

Key Ars:

1. AHotzman/BDawson – get Clary letter on MS & Intel working together on VM (combine work).

2. AHotzman/BDawson – Investigate – How do Intel/MS work together on defining Class Libs. Look at our Sun Media NDA and see if there are legal issues that prevent us from working with MS. Do we want to work with MS on these?

Attached document placed in: /tmp_mnt/ffs/f/mcg/attach/java/micros~1.doc on Sat Apr 20 13:21:40 PDT 1996

[STAMPED] 477MSCID00027 – Intel Business Information – Highly Confidential

PLAINTIFF'S EXHIBIT 4564 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4564 p. 1 of 2

[STAMPED] PLAINTIFF’S EXHIBIT 4564
[STAMPED] GOVERNMENT EXHIBIT 1204 Jackson 10/7/98

[1 line unintelligible]
Date: Mon, 13 Oct 1997 00:53:36 –0700
To: [email addresses @prognet.com]
From: Rob Glaser <robg>
Subject: MS-RN meeting 10-9
Cc: [email addresses]

RN Confidential

On Thursday (10-9) Jim B and I met with Anthony Bay to discuss how we might move the MS-RN relationship forward in a positive way. The meeting was a follow-up to a meeting that took place last week.

Joining Anthony (unexpectedly) at the meeting was David Rinn. The tone of the meeting was frostier than my recent interactions with Anthony, indeed colder [sic] than expected.

Anthony began by saying that the relationship between our 2 companies is in his mind “95% competitive.” At first he said he couldn’t think of anything positive that had happened between our companies. I countered that we thought they would think the ASF area was positive. He backed off a bit and said he agreed that our joint ASF work had been very positive.

We (Jim and I) expressed a strong desire to have a positive relationship between our two companies where possible, recognizing that while there would be substantial areas where we would compete, we thought that there were substantial areas where we could cooperate.

Anthony was not exactly receptive to this, although he didn’t want to close the door. He and Rinn indicated that in their opinion, RN had not always behaved honorably since the MS-RN deal was done. He suggested (and Rinn said point blank) that we had not been straight with MS about our intentions on a number of fronts.

For instance Rinn said we must have had a secret plan to go public before we did our deal with them, suggesting that we did this intentionally to devalue their warrants. Jim told them the exact date of the board meeting where the topic first got serious attention, and the date of the board meeting (when I was in New Zealand) where we made the decision. Since it was outrageous statement season, Rinn again reiterated the MS finance view that at the time we did the deal, RN had “screwed MS over the warrant.” I disputed this and asked him to explain why he thought this was true. He declined to do so.

Moreover, they said that they had hoped/expected that RN would not be staying in the horizontal streaming media business, and, perhaps even more outrageously, they claimed that I had privately suggested this to senior people at MS. I completely disputed this, pointing out for instance that the structure of our deal with MS ($25 and $35 million future options) runs directly counter to the notion of RN not doing future horizontal work. They backed off a little, and suggest I talk privately with the person who they say claims that I said this. I certainly plan to do so.

Throughout this Jim and I stayed calm (not as hard as I thought it would be, because their comments were so ridiculously off-base). After they finished venting, we tried to get the discussion back on track to focus on constructive things we could do together.

Anthony again suggested “hey why don’t you guys just become like SAP, building everything on top of MS’s full platforms”? We said that it was unlikely to make sense for the foreseeable future for us to just do this for a variety of reasons.

[STAMPED] MS-PCA1548647

Plaintiff's Exhibit 4564 p. 2 of 2

page 2

[1 line unintelligible]
but adobe has a big apps business that supports MS platforms. Anthony was not super willing to embrace the “Adobe” model, though it was not absolutely clear why.

Going from abstractions to practicalities, we said that were truly were open to supporting MS platforms in value-added ways, including where appropriate even their streaming media platforms if they provided a business model that made sense. We said we thought the right set of initiatives could move the needle from 95% adversarial to say 80% adversarial and 20% cooperative, and that there was real value in this especially if one focused on the trendline. We touched on ideas such as having Timecast list programs in ASF and/or Netshow formats, or having our player play their formats as well as our own.

We took the action item to write up a list of proposed ways to work together. I will draft up a list this week then pass it around internally for comment. The plan is to meet with Anthony in about 10-15 days to go through the list.

We also touched on a few tactical issues, including:

n MS’s warrants and decision regarding stock conversion (a subject of great emotion for Maffei and Rinn). I proposed a compromise -- that if MS agreed to take nonvoting shares, including potentially the warrant if exercised, we would agree that these shares could convert to voting upon sale. Rinn said he would check with Maffei on this and get back to us. Mark please follow-up with Rinn.

n MS pre-paying us the full $30 mil license fee. They “want the $10 mil back” until we fulfill the obligations. I agreed to look into this and get back to them this week. Mark please drive this and keep me posted when we’re prepared to get back to them.

n Them saying that they had not gotten the full deliveries from us of all the bits. On friday Anthony sent me a list of missing pieces, and Phil is in the process of formulating a response and delivery by EOD monday.

n On-going issues regarding branding and uninstall. The[y] said they would get back to us whether they wanted us to just do the uninstall work as part of our 6 person-months of consulting. (Phil have you gotten an estimate on this yet? If not please do so). Re branding, Kelly ball is in your court to write a letter re: our position on what they currently have up on their site. Kelly pls let me know when you can get this letter done, hopefully by EOD tuesday.

n Anthony it’s clear that Anthony is choosing to more closely identify with the hostile troops than with a more balanced view of the relationship. This is unfortunate, but we will do the best we can given this reality.

[STAMPED] MS-PCA1548648

PLAINTIFF'S EXHIBIT 4574 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4574 p. 1 of 2

[STAMPED] PLAINTIFF’S EXHIBIT 4574
[STAMPED] GOVERNMENT EXHIBIT 289

From:       Bill Gates
Sent:         Sunday, June 09, 1996, 10 26 PM
To:           Paul Mantz
Cc:           Carol Stork, Marshall Brumer, Bob Kruger, Joachim Kempin, Brad Silverberg, Jim Allchin
                (Exchange), Rich Tong, Hank Virgil, Craig Mundie, John Ludwig
Subject:     Intel – Andy Grove meeting June 7

I spent 2 5 hours with Andy on Friday. We met 1-1 and discussed a variety of important topics.

Background. Our relationship with Intel is advancing on many fronts. I was very impressed by the Intel roadmap presentation although I wasn’t there in person. They are going to deliver 50% performance increases every year in massive volume. If you compare their plans to the other processors vendors Intel seems to be pulling ahead. I am pleased that more group at Microsoft are taking the risk of really involving Intel.

Issues

1. DMI. Intel is super serious about DMI. Although we have promised to ship DMREG they feel we continue to see DMI as a want on our real approach. Andy thinks cost of ownership is super important and he thinks DMI is here now and we should embrace it rather than trying to go around it and do something better. I told Andy that I thought we still didn’t see DMI as central to our strategy but I would dig into whether there is an approach or a change that might allow us to be more synergistic here. If we are going to leverage Intel and the industry work done on DMI we may need to think of this as more than a mapping layer. I pass this one along to Bob Kruger to consider.

2. Cost of Ownership. As discussed in #1 Intel is heard core about trying to show PCs can be low cost of ownership. Andy showed me a diagram with a 2x2 and where centralized admin is one axis and power needed is the other axis. He has airline res terminals and the like as the centrally admined and lower power. He has decided that someone in Frank Gill’s group needs to get involved with this. He was excited when I talked about including Gartner. Like us they will move this from their NC group. Steve Nochshern will continue to do Webphone working with CPD. We didn’t talk about Webphone specifically but he thinks their effort is important. Someone under Frank will work with us on this cost thing. Andy was very excited about the idea of a “stateless PC”. Unclear to me how we leverage Intel’s resources to help out on this initiative. A minor note is that Andy thinks their cost of ownership is $8k per pc per year. I said that sounded 4x high to me but it must be a difference in what is included.

3. AGP. Intel is hard core about AGP. They [sic] know it requires some system support. They feel it is super important for them as they induce Klamath. They [sic] would like to really push MMX, Klamath and AGP as part of a “big bang” as early as next April even though these programs don’t come out on exactly the same day. They will only ship a few hundred thousand MMX parts this year so they won’t be pushing that in the mainstream. They are counting on us for AGP support.

4. Merced LOI. I told Andy that we were quite optimistic things will get resolved with HP in the next few weeks and then we will move quickly to sign the Merced LOI. In the next meeting I have with Andy in September we will review how much Intel should count on our work. Andy will determine how much they focus on other operating systems based on whether he feels he needs them as an insurance policy. I said based on the way we are working with Intel and getting them to do most of the Merced work, I thought he would be pleased 3 months from now how much has gotten done. Intel is disappointed we seem to think that their 98 date has a lot of risk. They don’t feel that it does. Andy wanted to understand our DEC 64bit announcement.

5. NT/Pentium Pro marketing. Intel is not shy about spending money IF they are really our only partner on this and if they get equal treatment and its not just them supporting us. They want Pentium Pro to get prominent positioning from this. I said I thought we would make them glad they are part of this but I would double check to make sure their expectations would be fulfilled.

6. Intel and Intranet day. I told Andy how great this event is going to be. He said they had heard that but they were told they couldn’t attend. I said I would make sure we freed up 3-4 seats for them and have someone contact them on Monday. It is URGENT that someone follow up on this. Marshall should check how to do this and make sure the invites get offered.

7. Security. They are not happy about our work with them. I said we have to work with W3C as a top priority. I didn’t have enough data to explain why we aren’t able to get along on this. We should try again.

8. Conferencing collaboration. I got the mail about the problems we are having with them too late to bring this up. Andy heard we are working well with them. The recent mail I got says we are not. I suggest Paul call Pat Gelsinger (who Andy thinks walks on water) and/or Frank before I get involved. I am happy to get involved. They may never have gotten over the issue of this being a P&L versus purely strategic. It sounds like we have concrete evidence of them not being a good partner.

9. JAVA. I told Andy that it’s [sic] inappropriate for their group to take anything resembling a Windows API and wrap it as a JAVA API. He agreed this was out of line but he didn’t think that was what his 6 people were doing. He thinks they are just optimizing for Intel. If they are taking DirectX aps [sic] and wrapping those then I need to register a much louder complaint.

10.  Browser. I thanks Andy for pushing his web people in our direction. I said it was important to us that they NOT ever publicly say they are standardizing on Netscape browsers.

[STAMPED] MS98 0169187 CONFIDENTIAL

Plaintiff's Exhibit 4574 p. 2 of 2

page 2

11.  Hood River.  Talisman/Wizard.  I told Andy these are excellent very important efforts that from our point of view they should make sure to put a lot of resources into because they fit perfectly with out strategy.

We will meet again in September and continue to exchange lots of email between now and then.  The Fortune magazine coming out in a week features a joint Interview with Andy and I where we are very friendly to each other.

[STAMPED] MS98 0169188 CONFIDENTIAL


Gordon et. al. v. Microsoft Exhibits

Published to Web April 06, 2004
Not Complete
Last Revised 4/25/2004


PLAINTIFF'S EXHIBIT 4423 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4423 Page 1 of 2

[STAMPED] SIGNED ORIGINAL
?51625?

AMENDMENT

AMENDMENT NO. 1 TO THE LICENSE AGREEMENT
BETWEEN
ZEOS INTERNATIONAL, LTD. AND MICROSOFT CORPORATION
DATED JANUARY 1, 1990, CONTRACT NO. 4934-0130

This amendment to the License Agreement between MICROSOFT CORPORATION (''MS'') and ZEOS INTERNATIONAL, LTD. ("COMPANY") dated June 1, 1990 ("Agreement"), is made and entered into this 1st day of November 1990.

1. The Effective Date of this Agreement will change from June 1, 1990 to April 1, 1990.

2. Section 9, second paragraph located on page 8 is hereby replaced with the following:

Upon execution by both parties. this Agreement supersedes and replaces MS Contract #4934-0022 ("PRIOR AGREEMENT") between COMPANY and MS. COMPANY shall receive a credit equal to: (i) the minimum commitment payments made by COMPANY pursuant to Exhibit B of the PRIOR AGREEMENT, minus (ii) the amount of said minimum commitment payments determined to represent earned royalties in accordance with the PRIOR AGREEMENT. Any such credit shall be applied to reduce COMPANY's minimum commitment payment due upon signing of this Agreement, as reflected in Exhibit B.

3. Telephone numbers found in Section 15 (page 11) will change from:

(COMPANY) 612-633-1175 to 612-633-4591 and (Legal) 612-633-4591 to 912-633-1175

4. Any Customer Systems shipped between April 1, 1990 and May 31, 1990 without Windows 3.0 shall be excluded from any royalties due for Windows 3.0 as stated in Exhibit C1. Any and all royalties due for MS-DOS on such systems shall still be in effect. The Customer Systems excluded from royalties must be stated as such on the royalty reports.

5. Any Customer Systems shipped between January 1, 1990 and July 31, 1990 under PRIOR AGREEMENT and this Agreement that included Packaged Product MS-DOS purchased from MS under license #4934-9125, and/or MS-DOS purchased from Phoenix Technologies, Ltd. shall be excluded from any royalties due under PRIOR AGREEMENT or this Agreement. The Customer Systems excluded from royalties must be stated as such on the royalty reports.

[STAMPED] PLAINTIFF'S EXHIBIT 1302 No: 2:96CV645E
[STAMPED] ENTERED MAR 06 1991 Document Database
[STAMPED] MS-PCA 1193954 CONFIDENTIAL
[STAMPED] MSC 5007949 Highly Confidential

[STAMPED] MICROSOFT LICENSE NO. 4934-0130

Plaintiff's Exhibit 4423 Page 2 of 2

6. In the event of inconsistencies between the Agreement and this Amendment, the terms and conditions of the Amendment shall be controlling.

IN WITNESS WHEREOF, the parties have executed this Amendment to the License Agreement as of the date set forth above. All signed copies of this shall be effective upon execution on behalf of COMPANY and MS by their duly authorized representatives.

MICROSOFT CORPORATION

   [signed]   
by

  Michael R. Hallman, President  
Name & Title (Print)

   2/9/91   
Date

ZEOS INTERNATIONAL, LTD.

   [signed]   
by

  Greg Herrick, President  
Name & Title (Print)

   11/29/90   
Date

Date Of Issue: November 1, 1990

This Agreement shall be deemed to be invalid, unless executed by COMPANY and returned to MS within thirty (30) days of the above "Date of Issue". FAX copies are unacceptable.

11/02/90 3026M

[STAMPED] MS-PCA 1193955 CONFIDENTIAL
[STAMPED] MSC 5007950 Highly Confidential



Gordon, et. al v. Microsoft

Exhibits Published to Web April 07, 2004
Not Complete
Last Revised 4/25/2004


PLAINTIFF'S EXHIBIT 4576 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4576 p. 1 of 5

[STAMPED] PLAINTIFF'S EXHIBIT 4576 Gordon v. Microsoft
[HANDWRITTEN] Follow up 5/10[/1993?]
[HANDWRITTEN] F Microsoft
[STAMPED] Reiswig [?] Exb. 1
[HANDWRITTEN] 931270287
[STAMPED] CONFIDENTIAL IBM 0410327245

Administrative Assistant,
Office of the AGM, Market Development and Operations
[ADDRESS] [TELEPHONE]

May 6, 1993

MEMORANDUM TO:
Mr. W. W. Casey
Mr. J. V. Kalb, Jr.
Mr. E. M. Lineen
Mr. J. R. Patrick
Mr. L. R. Reiswig, Jr.

SUBJECT: Briefing Sheet for Gerstner Call to Bill Gates May 26

The attached briefing sheet has been prepared for Lou Gerstner for his May 26 call with Bill Gates. Please review and return your comments to Mike Nolan by May 10 so that he can incorporate them into the final version for Jim.

Mike Nolan's PROFS ID is [PROFS ID]; Fax [FAX] T/L [TELEPHONE]. Thank you.

R. L. Blair

RB/ag
Attachment

Plaintiff's Exhibit 4576 p. 2 of 5

[STAMPED] CONFIDENTIAL IBM 0410327246
[FAX TIMESTAMP] May 6 '[19]93 8:33 FROM IBM BOCA [FAX NUMBER] PAGE 001
Fax Transmittal Memo 7672
To [HANDWRITTEN] Ray Blair
Number of Pages 4
From [HANDWRITTEN] Mike Nolan [?]
Today's Date [HANDWRITTEN] 5/6[/1993]
[FAX NUMBER] [TELEPHONE NUMBER]

Microsoft
Company Profile:
Founded: 1975
Incorporated: 1981
IPO: 1986

CEO: William H. Gates
Exec VP Products: Michael J. Maples
Exec VP Marketing: Steven A. Ballmer

Employees: 11,500 (WW 12/92)
[COLUMN HEADER FOR PERCENTAGES BELOW] % Rev
Total Revenues (1/92-12/92) $1.2B [?]
Cost of Revenue $0.5B 17%
Gross Profit $[?] 83%

Operating Expenses
 - R&D $0.4 B 13%
 - Sales and Marketing $1.0B 32%
 - G&A $0.1B 3%
SG&A Expense/Revenue 35%

Statement of Operation: Engaged in Design, Manufacturer, Marketing, And Support Of Microcomputer Systems and Application Software Products And Related Books, Hardware, And CD ROM Products

Vision: "Information At Your Fingertips"
 - Empower The User By Making Inforamtion Easy To Find & Use
 - Make Technology More Approachable Yet Offer[indistinct] People Access To More Information Than Ever Before
 - "Windows Everywhere" is the Foundation Of This Vision

Strategy Summary:

Microsoft develops and markets the broadest line of PC software products for Intel, Apple, RISC and UNIX platforms. Their goal is to establish "Windows Everywhere" by extending their operating systems from the desktop up to workstations, servers and mainframes and down to consumer devices such as wallet computers, fax machines and television cable converters. The operating systems then leverage Microsoft's programming tools, applications and services. Products are acquired or developed internally using proprietary tools and customers are used exclusively for testing. Being early to market is more important than function or quality.

Their business strategy is focused on defining industry standards through market share. Specific tactics to drive market share and long-term customer commitment include:
 - Exclusive contracts for preloading Microsoft software on OEM equipment
 - Deep discounts for running Microsoft software on all PCs in large accounts [indistinct]
 -  Aggressive upgrade pricing and bundling of competitive functions in the operating system

Microsoft also publicizes their strategies and product directions frequently and aggressively, highlighting future offerings years before availability. Microsoft executives are highly visible with customers, press and the consultant communities. Generating negative information about the competition seems [indistinct] to be part of their PR effort.

Bill Gates:

Bill Gates is Microsoft. The company's identity and products are all perceivedto emanate [indistinct] from him. He is a Harvard University dropout... very driven... an innovator and visionary. Bill Gates has a special allure that attracts curiosity, as well as an abundance of press about him. He has ben labeled the "Julius Caesar of the software industry" .. ruthless, incisive, and tactically brilliant. "Unlike most founder-chairman types, Gates is a strategic thinker...at every turn of the road, he's made long-term decisions that have paid off. He licensed MS-DOS to IBM instead of selling it. Most 25-year-olds would have sold the farm." At the brink of success of Visicalc, Gates decided that application software (spreadsheets, word processors, data bases, etc) would be a category he would have to dominate in order to achieve his goals for Microsoft.

Bill maintains a very casual demeanor (e.g. he uses words like "cool" & "rad"). He works 15-hour days, 7 days a week. He is frequently considered "quirky"...likes expensive sports cards and did not own a television until recently. He has publicly stated that his goal is to "bury IBM" and that "OS/2 is dead."

IBM Confidential

Plaintiff's Exhibit 4576 p. 3 of 5

[NOT YET TRANSCRIBED]

Plaintiff's Exhibit 4576 p. 4 of 5

[NOT YET TRANSCRIBED]

Plaintiff's Exhibit 4576 p. 5 of 5

[NOT YET TRANSCRIBED]


Gordon et. al. v. Microsoft Exhibits

Published to Web April 15, 2004
Not Complete
Last Revised 4/25/2004


PLAINTIFF'S EXHIBIT 62 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 62 Page 1 of 3

[STAMPED] PLAINTIFF'S EXHIBIT 62 Gordon v.. Microsoft
[STAMPED] X180836 CONFIDENTIAL
[text begins with the end of a sentence]
we need.

From billg Tue Feb 21 19:55:09 1989
To: mikemap
Subject: Microgrphx
Cc: steveb
Date: Tue Feb 21 19:55:09 1989
Mail-Flags: 0000

Steve talked to grayson at the conference. The[y] said they are wiling to make it 10k for big companies and 1k for companies with less than either 3M or 2m in sales. Based on this us getting full use for you including source code and rights to incorporate wherever at noe [sic] extra charge after that seems like a shoe-in. Talk to steveb about this. I am excited about us making this happen.

From steveb Wed Feb 22 10:18:49 1989
To: billg mikemap
Subject: Micrographx
Date: Wed Feb 22 10:18:47 1989
Mail-Flags: 0000

as billg said mike only question is corp accts[.] They dop [sic "do"] not resell[.] should be 1000 not 10000 but confirmation with MG is needed[.]

From billg Tue Feb 21 17:09:01 1989
To: mikemap
Subject: Re: Mirrors
Date: Tue Feb 21 17:09:01 1989
Mail-Flags: 0000

These guys told me explicitly they were not trying to make money on this product and they did it for their own use at first and were pleased others might want to use it. Their price is $10k. Agreeing to pay them $100k to get total source rights and all updates is a fantastic deal for them. I want to call and tell them they are liars and we would have to create this ourselves if they don't get reasonable. They explicitly said we could have source. They will get a great deal if they take updates from us. I dont [sic] want there to be any restrictions on us after a year because we may incorporate some comatibility stuff into windows or PM and althought [sic] we dont [sic] plan to use their code I dont [sic] want any garbage from them. windows or pm after a year we pay a small sum. They should want us to make windows and PM closer. This is all crazy. We need to get moving and that means closing a deal fast. I thought these guys bragged about quick decision making. GO ahead and work on them so[me] more and keep me posted.

From cameronm Thu Apr 13 22:28:45 1989
To: martyta paulma petern waggenerltwg warrenm
Cc: alistair billg bobt jonl lioneljo mark[?]sc mikemap patbe peteh steveb viktorg
Subject: Comdex OS/2 ISV Luncheon
Date: Mon Apr 10 21:23:53 1989
Mail-Flags: 0000

On Tuesday 4/11[/1989] IBM and MS hosted a luncheon at Comdex for the top ISVs in the industry to evangelize and push OS/2 PM application development.

Plaintiff's Exhibit 62 Page 2 of 3

[STAMPED] X180837 CONFIDENTIAL
Talks were given by Lee Reiswig and Jim Cannavino of IBM and by Billg and Steveb. After the brief talks it was opened up to discussion and comments/suggestions by the ISVs.

Lee Reiswig began by going over the marketing promotions we will be offering ISVs (advetorials, co-op advertising, road show, trade show space, etc.). Bill and Steve went over the technical support we are offering and the device driver schedules and distribution plan. Jim Cannavino disclosed IBM's plan to remove the additional cost from the customer[']s decision in choosing OS/2 as an environment (he hinted and later said more clearly that IBM would ship machines configured with 4Meg at no additional cost). IBM will also offer low cost memory upgrade kits for OS/2 buyers.

The luncheon was to reiterate MS and IBM commitment to making OS/2 successful and helping ISVs successfully develop and market OS/2 applications.

Problems that were brought up:

  • The end-user cost of OS/2 (memory requirement) needs to be reduced (Samna)
  • Co-residence of OS/2 and DOS on same machine is difficult (Micrografx)
  • Cost of the OS for future versions of OS/2 is a factor that can affect when people go to OS/2. If corporate buyers don't know the upgrade policies for future version of OS/2, then they might think they can save money by waiting for the next release of OS/2 rather than buying now. (WordPerfect)
  • Merely removing the barriers to OS/2 in terms of cost, memory, config, etc. is not enough. This makes OS/2 less painful but does not provide a compelling reason to go to OS/2. (Ventura)
  • MS/IBM have done well in promoting OS/2 within the industry but there's not enough mass market and general business awareness of OS/2. (Microrim, Symantec)
  • People don't understand the benefits of OS/2. We must create end-user demand for OS/2 products. (Ventura, Microrim)
  • Device driver availability and performance (in line drawing) is a problem. We are weak in fonts and need additional typographic characters. (Ventura)
  • Fonts are very hard and awkward to install, in general usability and installation are not as good as the Mac. (Aldus)
  • Our PostScript driver is woefully inadequate. (Aldus)
  • We need to get OS/2 on corporate approved purchase lists. We need to identify corporations who are predisposed to adopting OS/2 early. (Symantec)
  • MS OnLine support does not give as fast a turnaround as it should. The 16Mb virtual memory limit is too restrictive for some apps. (ISoft)
  • MS OnLine interface is very bad and doesn't run under OS/2 itself. (Asymetrix)
  • There aren't enough tools for moving Windows apps to PM. (Aldus)
  • App developers need cascading and pop-up menu support. (WordPerfect)
  • Positioning of Win 3 versus OS/2 is unclear and confusing especially to corporate customers (ISoft)
  • Windows 3/386 may be misinterpreted as being better or more advanced than OS/2. (Asymetrix)
  • OS/2 needs a 32-bit coordinate system. (Micrografx)
In general the biggest point made was that while everyone understands the


Plaintiff's Exhibit 62 Page 3 of 3

[STAMPED] X180838 CONFIDENTIAL

costs and problems of OS/2, very few people understand its benefits. IBM and MS need to create demand for OS/2 products by explaining OS/2's benefits. While technical problems were raised they were enver belaboured because the ISVs seemed comfortable that we knew about them already and that we would fix them (Billg took them through the latest device driver schedules and distribution plan).

Suggestions that were made:

  • Provide a better dual boot and simpler installation procedure for people who want both OS/2 and DOS to reside on the same machine. (Micrografx)
  • IBM should offer OS/2 and DOS machines at the same cost. IBM should offer machines with 4Meg installed for same cost bundled with OS/2. (which IBM basically said they were going to do). (Samna)
  • Provide a low cost and liberal upgrade policy for future versions of OS/2 to 1.1 buyers so they won't be tempted to wait. (WordPerfect)
  • Do more marketing to corporate users and business people (as opposed to industry-related promotions). Do a lot more classic mass market advertising including TV advertising. Run an ad during the SuperBowl. (Microrim)
  • Do not run an ad during the SuperBowl, but do market aggressively to user groups and PC managers/coordinators - emphasize corporate buyers. (Ventura and Symantec)
  • Identify corporate early adopters and share this information with OS/2 app developers. Help get corporate beta sites for OS/2 apps. (Symantec)
  • Identify specific OS/2 solutions that corporations can being buying, we need to show them where to implement OS/2 in their company. (Symantec)
  • Provide more Win to PM tools, especially a metafile conversion utility. (Aldus)
  • Promote OS/2 technical futures at the time Windows 3/386 comes out so people don't think that only Windows will have advanced features like demand paging. (Asymetrix)
The ISVs want to see more end-user demand for OS/2 and OS/2 applications and generally agree we should concentrate and focus on corporate buyers. We need to make OS/2 cheaper, easier to implement, and promote specific OS/2 application solutions to those who actually buy and use the apps.

ISV Attendees:

Ashton-Tate, Joe Brillando, VP Corporate marketing & Strategy
WordPerfect, Alan Ashton, President & CEO
Borland, Philippe Kahn, CEO
Microsoft, Mike Maples
Aldus, Mike Solomon, VP Marketing
Symantec, Gordon Eubanks, CEO
Ventura, John Meyer, President
Microrim, David Hull, President
Samna, Said Mohammedioun, CEO
Asymetrix, Steve Wood, VP Marketing
DRI, Dana Hooper
MDBS, Gary Rush, President
Informix, John Millovich (server business)
Informix, Ron Ferguson, VP Marketing (Wingz & Smart)

PLAINTIFF'S EXHIBIT 4588 Gordon v. Microsoft   [ Top ]

Plaintiff's Exhibit 4588 p. 1 of 2

[STAMPED] PLAINTIFF'S EXHIBIT 4588 Gordon v. Microsoft
[STAMPED] Grovernment Exhibit 277
[STAMPED] MS98 0169352 CONFIDENTIAL
[MEMO LETTERHEAD] Bill Gates
From: Bill Gates
Sent: Thursday, May 25, 1995 9:42 AM
To: Carl Stork (carls); Paul Maritz (paulma)
Subject: FW: NSP

Here is mail I sent to Andy. He and I will talk on the phone in the next few days. Any furhter [sic, "further] guidance is valuable although I think Carl's mail laid out the issues quite clearly.

We need so[me] decent compromise here and its not clear to me what it is. Should we agree to work with this group on specific things to get them redirected?

It's a shame we have no input from them on Windows97.

From: Bill Gates
Sent: Thursday, May 25, 1995 9:34 AM
To: 'Andy Grove [EMAIL ADDRESS at Intel]'
Subject: NSP

I've learned some more about this. I am not emotionally involved in this particular debate but there is a significant problem here that is on a path to get a lot worse.

NSP means a lot of different things. What it means in terms of hardware and the processor being able to take over more and more functions in [sic, "is"] non-controversial. What is a problem is the incompatibility and overlap of systems software done under the label NSP. Actaully [sic, "Actually"] some of the software elements are Ok - we still seem to have a common view of TAPI, IrDA and APM. Each of these has to be improved so with our current relationship we will probably diverge on each of these. The ones that are a major problem for Windows95 and WindowsNT today are SPOX, Native aduio [sic, "audio"], DCI, 3DR, DMI, PCMCIA card services and Instant on. These are incompatible with where we are taking Windows and have many problems with Windows95.

When we talked I didn't realize that Kinnie was involved in NSP. In fact not only is he involved but he feels that he has personally championed all of the elements against resistance inside Intel and Microsoft and that he will overcome non-believers. He seems to feel that although Intel doesn't share what it is doing in system software with us that he feels like we are supposed to explain exactly what we are doing in advance.

Just recently we called to get a SPOX development kit and nnow Ron and Craig are thinking about whether to send us one or not. They first demanded we guarantee compatibility with SPOX before they would send us the kit.

ISVs are more confused about what to do by Intel's attacks on our Windows plans than they are by Apples.

Structurally it[']s very hard to have our people working as best they can to advance PC software standards with a group of 200 people fully funded to basic[al]ly try to do the same thing in parallel with no guidance to coordinate with us at all. They don't share with us because of Intel's IP attitude. We tried shipping some Intel code relative to DCI and ended up in this Apple lawsuit that has been very damaging.

I don't understand why Intel funds a group that is against Windows95. Craig's group uses VxDs. Even if he fixes his stuff to run with Windows95 we don't have VxDs on Windows NT. Craig[']s attitude is he doesn't care about NT and our family strategy.

It turns out that the P6 slows down certain 16bit operations so it only runs Windows95 about 15% faster than an equivalent pentium. Windows NT gets the full benefit of the P6. It seems like Craig's not believing in Windows NT will be a major boon for AMD chips since they will run todays Windows 3.1 faster than any Intel chip.

The problem we have is that we have to sort of choose in software related issues which company will lead and which will follow. In chips its very clear. In software you hav a group that won't allow us to lead and has all the prestide [sic, "prestige"] and profits of Intel to drive them forward.

If we picked someone neutral and asked them to compare what we are doing in graphics and sound to Native Audio, 3dr DCI and your other work I have no doubt they would say our work is equal or better. However Craig and Ron don't believe they should give up on anything. Meanwhile we have told our group that they need to evolve Windows software including Windows NT.

[PAGE NUMBER] Page 305

Plaintiff's Exhibit 4588 p. 2 of 2

[STAMPED] MS 0169353 CONFIDENTIAL
We should talk on the phone more about this.
[PAGE NUMBER] Page 306




Last Updated Monday, December 07 2009 @ 08:21 PM EST


Groklaw © Copyright 2003-2013 Pamela Jones.
All trademarks and copyrights on this page are owned by their respective owners.
Comments are owned by the individual posters.

PJ's articles are licensed under a Creative Commons License. ( Details )