Now that we know what both Novell and Microsoft lawyers said in their opening statements Tuesday at the antitrust trial just starting in Utah over WordPerfect and QuattroPro, folks here are Groklaw are starting to find exhibits from the Comes v. Microsoft antitrust trial collection here on Groklaw that do seem to raise questions about the facts asserted in the Microsoft opening statement.
So I decided to put the report about the Microsoft opening statement from the courtroom up again, with some of the exhibits we've uncovered, in the hope that it will be useful by comparing them. If you find more, please add them in your comments. We're still transcribing the exhibits, so feel free to help with that as well, by all means. We're concentrating on this page now, but you can work on whatever interests you.
First, here's Chris Brown's report on what Microsoft's attorney told the jury in his opening statement:
Microsoft's Opening Statement:
Now let's take a look at what the gang has unearthed from the Comes exhibits. As you'll recall, that case settled early in the trial, but the judge made the exhibits public. So here on Groklaw, we've been plugging away at making them available to the world as text, to make them more easily searchable. The PDFs have numbers, not names, so it was almost impossible to search them before.
David Tulchin then provided the opening statement for Microsoft.
Mr. Tulchin told the jury that this case is only regarding the period from July 1994 until March of 1996, the period that Novell owned WordPerfect.
He said Novell made bad decisions with WordPerfect.
He said the Windows 95 beta was not a final product, that the namespace extension was a Microsoft invention, owned by Microsoft.
He said that Microsoft had an operating system monopoly, but that Novell's expert has previously testified that Microsoft legally obtained its monopoly.
He went on that conduct is not anti-competitive if it was taken for a legitimate business purpose.
He described Microsoft history and growth... Bill Gates, the writing of Basic, PC/MS DOS, Windows.
He showed the timeline of Microsoft Word and Excel GUI release for Apple Macintosh in 1985. Then he showed the WordPerfect (GUI) timeline with WordPerfect 1.0 for Macintosh in 1988, then nothing for Microsoft Windows until 1991.
He says the evidence will show the reason they were slow was that WordPerfect was always late.
Novell bought WordPerfect, but they didn't own it at this time (referring to the GUI releases shown on the timeline).
He said that at the time Novell was well-known for NetWare, that the emails Novell will be showing were from 1991, 1992, 1993 about competition that was about Microsoft competition with NetWare, not the software at issue in this case.
He said in 1994 Novell bought Quattro Pro from Borland.
He showed a timeline of WordPerfect market share of DOS and said that WordPerfect dominated the market. He showed the Microsoft Word timeline for DOS share (very low).
He said Microsoft Word was very successful on the Macintosh, a GUI platform.
He said that when Windows 3.0 was released in May of 1990, users moved in droves from DOS to Windows. But WordPerfect didn't have a product for Windows -- they were miles behind.
He shows on a chart the first WordPerfect release for Microsoft Windows was 1.5 years later.
He showed a chart of WordPerfect market share for Windows and Microsoft Word, generally showing Wordperfect's share increasing until 1995 then dropping, while Microsoft Word's share was increasing.
(Chris: In my notes I estimate the data points as: WordPerfect, 1991 release 15%, 1992 32%, 1993 29%, 1994 25%, 1995 18%, 1996 8%. And for Word: 1990 92%, 1991 70%, 1992 58%, then line increases to 90+% tracking decrease in WP).
Mr. Tulchin showed a Novell (WordPerfect) document dated Dec 21, 1993 describing WordPerfect 6 as "too slow, too many bugs to be considered sufficiently stable. Requires too much memory and too much disk space."
He shows Novell's 1996-1998 business plan dated June 29, 1995 which reports that they are still recovering from WordPerfect 6.0 for Windows which was too slow and buggy.
He shows a document dated July 15, 1994 from a PerfectOffice Suite Business Review Exercise "Microsoft Office is currently the industry-leading suite... It's one-two punch, Word and Excel, is the strongest combination in the industry."
Mr. Tulchin told the jury that if you have a product, and its quality is not too high, and you have a competition that's the "one-two punch", well, you can tell what happens.
He showed a Novell August 1994 business plan reporting that their partnering with Borland had not been received well.
He showed a chart indicating the market shift to suites, and it showed revenue from sales of suites versus stand-alone applications. It depicted a decline of stand-alone spreadsheets and word processors and a corresponding increase in revenue from suites. He pointed out that by 1995 suites were three times the combined revenue of standalone spreadsheets and word processors.
He said that the time Novell purchased these products, there had been two major changes in the market -- first, the shift from DOS platform to GUIs and second, the shift to suites.
He claimed that in effect Novell was purchasing products in decline, that Novell and the companies before it were not adapting to technology.
He showed a graphic of a 24-frame film roll versus an 8GB SD card holding up to 5,000 photos. He said that if you don't adapt, you'll be left by the wayside.
He showed a chart of Novell's stock price at the time Novell announced the purchase of WordPerfect and Quattro Pro, saying it indicates the market reaction to Novell's decision. He said the announcement came after the market closed on March 21, 1994, that it closed the day at $23.75, but the next day it was down to $20, then the day after that to $19. He says that Novell's market capitalization fell 20% or a drop of $1,467,849,750 on an announced WordPerfect and Quattro Pro purchase price of $1,555,000,000.
He claims that investors recognized that it was the purchase of declining products, that these two products were not worth what Novell was willing to pay.
He said that Chicago (Windows 95) represented a big advance in the user interface, that namespace extensions were only a small part, that the namespace extension APIs accounted for just 4 out of approximately 2,500 APIs in Windows 95, that their expert estimates Window 95 has as many as 2,800+.
He said there were three reasons for withdrawing support for the namespace extension APIs.
First, third-party programs that called the namespace extension APIs could crash the Windows 95 shell.
Second, namespace extension APIs were not compatible with future versions of Windows then under development in 1994 (Windows 95, the unreleased Cairo, and Windows NT). He said the Windows NT team, developing Windows for servers, were very concerned with stability. They lobbied Bill Gates hard to eliminate support for the namespace extension APIs.
Third, the namespace extension APIs did not achieve the functionality Bill Gates hoped could be achieved. He had wanted users to be able to launch from the hierarchical view in the left pane and run in the right hand content pane.
He said that Windows 95 allowed users to launch WordPerfect and Quattro Pro multiple ways -- from the Start menu and from WordPerfect and Quattro Pro icons on the desktop. He said these were new features in Windows 95. He claimed Microsoft invented these methods.
He said what Novell wanted was a third way to get to WordPerfect and Quattro Pro via the Explorer. He showed a graphic of a prototype of the unreleased "InfoCenter". The graphic showed Explorer with a namespace mailbox inbox with email messages in the content pane.
He showed how ISVs were able to add folders, visible in Explorer, to get to WordPerfect and Quatro Pro.
He further said that namespace extensions were never removed from Windows 95, and if Novell really needed them, they were there. Support was removed, and it was not guaranteed to stay there.
He referred to Novell's exhibit 1 and highlighted the portion before that section pointed out by Novell, "... Having the Office team really think through the information intensive scenarios and be a demanding client of systems is absolutely critical to our future success. We can't compete with Lotus and WordPerfect/Novell without this".
Mr. Tulchin said Novell's highlighted section is not about the namespace extension APIs; it's about the application developers making their product the best possible.
He said that Novell could have written their own software to get that functionality, or they could have used the remaining APIs.
He then referred to a Microsoft non-disclosure agreement/contract for pre-release products signed May 24, 1994 by WordPerfect which reads: "The Product may not operate correctly and may be substantially modified prior to first commercial shipment. Company assumes entire risk with respect to the use of the product."
He said that Novell claims the reason they were late is because they had to do the work themselves rather than rely on Microsoft's work.
(break for lunch)
Resuming, Mr. Tulchin then showed reasons Novell was late.
He displayed an email dated August 3, 1994 from Ad Rietveld to Jim Tolonen "After Windows 3.1 (16-bit) products are released this year, move large percentage of Windows resources over to Chicago (32-bit)... There are very few resources on Chicago at this time."
He displayed an email dated April 3, 1995. Novell Business Applications Business Plan "Tremendous growth in the suite category has resulted in a corresponding drop in the growth rate of standalone applications... Weak vis-a-vis Microsoft in perception for corporate strategy, vision, and ability to develop software... Third to suite market. Microsoft has enormous momentum."
A May 3, 1995 Novell Employee Survey: "48% of employees originally hired at WordPerfect are thinking about quitting."
From December 23, 1995 email from Bruce Brerton regarding Quattro Pro developers quitting: "15 additional people submitted their resignations. All (except 1) will be going to Oracle. From a development standpoint, this leaves us with just 2 people."
Mr. Tulchin said that the Novell suite could not get released until Quattro Pro was complete.
He showed an August 3, 1994 email from Ad Rietveld regarding Novell/WordPerfect/Quattro Pro Unification Plan: "Increase resources on internationalizing Quattro Pro for faster time to market support for localized versions."
Another email dated February 2, 1995 which Mr. Tulchin pointed out is 6 months later: "Quattro Pro folks still working on international version of Quattro Pro 6.0. Expect to finish that by end of March and then will begin on next version of Quattro Pro."
Mr. Tulchin said this is 2 months after the namespace extension decision. He says this is still the DOS version they are working on. He said Novell claimed the only reason for the delay was the namespace extension, but here Quattro Pro says it won't be until the end of March before they start on the next version of Quattro Pro, only a few months before Windows 95 comes out.
He showed a chart of market share of office suites for the Windows platform from 1993 to 1996. (Chris: I estimate the data points to be: Microsoft 1993 80%, 1994 81%, 1995 88%, 1996 89%, Borland Office/PerfectOffice 1993 3%, 1994 4%, 1995 5%, 1996 7%).
He said that Microsoft was the first company to come out with a suite and that Novell's office suite was never successful.
Novell contends it's Microsoft's fault they were late, that it was solely Bill Gates' decision on the namespace extension. He claimed that, instead, Novell made bad decisions -- they paid too much (the market indicated that), Novell's products were slow, they were buggy, they were late to market even before the namespace extension.
He said that Microsoft made its decision: not to make the best product for Novell, but to make the best Windows 95 for Microsoft.
According to Novell's theory, if a version of WordPerfect for Windows 95 had been released much earlier than it was, that event would have resulted in a sharp decline in Microsoft Windows popularity.
He says he has a witness, a Microsoft employee, that will say that a WordPerfect employee told him that WordPerfect will be "OK" with Microsoft's decision on the namespace extension.
Mr. Tulchin said there was no complaint from Novell way back then, that they only filed this lawsuit 10 years later. If it was such a big deal, do you think they would have remained silent for these 10 years?
Here are some exhibits from Comes v. Microsoft that do seem to contradict what was said in Microsoft's opening statement, or at least raise questions that need an answer, and we'll keep adding to the list if new ones are discovered:
It seems to me this would demonstrate at least two things: 1) Novell did complain at the time, contrary to what Microsoft's lawyer reportedly argued; and 2) Microsoft's Jim Allchin wanted to "kill" Novell while pretending to be friendly.
From: Jim Allchin
To: billg; bobmu; bradsi; davidcol; jonl; jonro;
mikemap; paulma; richt; steveb; w-clairl
Subject: Customers and Novell
Date: Saturday, September 18, 1993 3:24PM
Reading the PCWeek article about the mud-slinging between
Microsoft on the "black screen of death" again points out
the fact that
we don't have everyone trained with the same approach to
lot of people talk to the press and (and customers) and we
giving the "we just want to help customers" message
Instead, we're hitting back at Novell - in a losing way.
nothing other than set us up as the bad guys. Novell is
and we continue to fall for it. And every time some
or exec falls for it, we look like we're fighting with
Novell in front
of customers. This is very, very bad.
I want to ask you to get everyone to begin and end every
discussion/quote with the press dealing with Novell with
the same basic
message: (take your pick — they're all basically
examples of the same
+ "we don't care where or what the problem is, we will
work to solve
the customer issue"
+ "it doesn't help customers throwing stones between
just going to fix the customer problem"
+ "our mission is to work with every company especially
their large marketshare to solve customer problems. We'd
be crazy not
to work super hard with them."
+ "Pointing fingers is stupid. We will work to Novell and
internally to solve the problems"
+ "Yes, we have a group that competes with Novell on the
isn't competition good for customers? Novell support is
us because it's critical to customers. We'd be pretty
working as hard as we can with Novell on compatibility and
interoperability. That's what customers want. Again, we're
thinking of customers here. We welcome Novell's
they work with us to help customers."
Novell uses every opportunity that we "lose it" to say
against us. "See they really hate us and won't work with
work with everyone. Microsoft works with no one." (Then we
defensive and hit back – exactly the wrong thing to
do in my view.)
We MUST get control of this – especially as we move
to ship our own
NetWare support. We must be able to explain clearly "we
just want to
solve customer problems". "Of course, we're going to
working with Novell. We absolutely want their software to
"We also must think of customers and customers are begging
this support to come directly from us." [we need to put
strong list of names that we can use here.] "Why else
would we invest
so much effort in this client (and server) work? We sure
any more money) We had to do this for our customers." [we
need to hammer home this last point about it costing us
money to do
this and we're not getting anything back from it. It is a
supporting argument about us just trying to help
Sentiment is against us. We can and MUST turn this around.
become more aggressive against Novell product and
must get our mouth in order. The press, etc. is very
sketical of us
so one slip up and we get set back quite a ways.
This really isn't that hard. If you're going to kill
isn't much reason to get all worked up about it and angry
– you just
pull the trigger. Angry discussions beforehand are a waste
We need to smile at Novell while we pull the trigger.
From: Nathan Myhrvold Well. Mr. Myhrvold, now at Intellectual Ventures, in a discussion with Bill Gates, who wanted to make sure APIs were patented, so using them would require infringement: "I want to make sure people understand. I want something where anyone doing our APIs or using our formats will infringe on our ideas. I have received nothing on this. I think there are misunderstandings about what can and cant be done. This should be treated seperately but I want to know who the champion is, what the status is. How many patents would a chicago clone on top of Unix violate? If not a lot then I will help you be more creative."
To: paulo; craigmu
Subject: FW: Chicago and 3 key areas
Date: Saturday, September 18, 1993 12:11PM
This is interesting with respect to our Chicago MM plan. I
presume that much of what we want to do will
be in the "power piece".
From: Bill Gates
To: bradsi; davidcol; dennisad; paulma
Cc: jimall; mikemap; nathanm; steveb; stevesi
Subject: Chicago and 3 key areas
Date: Saturday, September 18, 1993 7:52AM
Chicago seems to be marching on like a series of army
rifles or a series of british highways:
M1, M2, M3, M4...
This teutonic aspect of the project is valuable. I am
impressed that the machine is rolling - its professional
its critical its great. The group should feel
very good.However Chicago has to have a soul too. I want to
make sure the following 3 areas are handled properly.
1. Ease of use. I want a hard core audit of
whether we are eliminating everything we should
eliminate. I want a hardcore comparison to the mac on
these issues. For example is it easier to add a
cdrom? Is it easier to connect to a network printer? Is it
easier to use aliases? We have been humiliated on
ease of use for so long. Get some mac lovers - not our
internal attenuated mac lovers but some real true
unblemished mac lovers to really try this stuff and tell
us what they think. Get a champion of this area who
will ignore the current march of rifles and development
saying something is too hard or cant be backward
compatible to really audit this thing hardcore and tell us
what we are not doing that might make a
difference. Where are we better than Macintosh? What
happens when a floppy disk gets inserted? This
apect of the project is so critical I would slip to get
2. Patents. I raised this issue as a serious issue at one
of our reviews. I want to make sure people
understand. I want soemthing where anyone doing our APIs
or using our formats will infringe on our ideas.
I have received nothing on this. I think there are
misunderstandings about what can and cant be done. This
should be treated seperately but I want to know who the
champion is, what the status is. How many
patents would a chicago clone on top of Unix violate? If
not a lot then I will help you be more creative.
What lawyer has been picked to really be your partner in
thinking this through in a brilliant way and make
sure the documentation is done right? I want to involve
some thinking nathan has been sharing with me on
this as well so make sure you are communicating with him.
In particular I want to do something about
patenting .exe format and encrypting drivers.
3. Product partitioning. I am excited about the thinking I
have done on this! We want to revenue maximize
with this product. I have a very specific plan for this
that will have limited but some impact on the product.
It is a plan that is good for customers, oems, isvs and
Microsoft. Essentially there are 3 places that things
a. The kernel piece
b. The power piece
c. One of the addons.
The kernel piece is basicly only licensed to Oems. It can
run Windows applications but it is missing key
stuff that the vast majority of Windows users will want.
It does not run as fast as a Windows with b. We
can license this to oems for somewhat less than the
current oem price for DOS + Windows. The simplicty
of this product will emphasize the simplicity for a
consumer user but only the most simple minded customer
would get by with just this piece. We would prefer to err
on the side of this piece being too small rather
than too big. The code between these 2 pieces will have to
mate in a very complex way for technical
reasons. The manual for this piece is simple - just enough
to let you run a single windows application. It
may have its own unique simple shell.
The power piece has to appeal to almost everyone and yet
be unique enough that no one will claim to be
able to do everything we do in this piece. It will speed
things up, it will have the desktop,... it will see to
users at retail for something like $79-$99 and we will
make sure that almost everyone knows this is a
crucial piece for them. We allow OEMS to also buy this
piece and bundle it if they want but it is not subject
to much volume discount and we are fine if they dont since
we understand the retail channel for windows
-this would mean the oem fee for just this piece would be
around $35-$50. The piece has lots and lots of
stuff in it.
Whatever is left over or too big we put on the CD (which
needs a spec at some point) as an extra buy
I think perhaps a face to face meeting on this last topic
is required since its a VERY VERY important topic -
hundreds and hundreds of millions are at stake. $/windows
has been going down and this is our chance to
help everyone at get $/windows to go up at the same time -
the power of an innovative approach. Its also a
complex topic with lots of implications for oem/retail. We
should include steveb in this if he is free.I am
ready to meet anytime.
"we also haven't given them chicago code, and, we changed nt apis at the last minute and didn't tell them." Chicago is what Windows 95 was called as it was being developed. So. This is what Novell is complaining about and what Microsoft says didn't happen, or only happened for valid business reasons. But its significance to me is that Microsoft's lawyer at trial reportedly said that the APIs were still in there, and Novell could have used them. But what apparently really happened with NT is that the change was stealth. Novell only found out by having their stuff break. This doesn't prove it happened again with Chicago, only that Microsoft withheld code from Novell, but it's to me a very disturbing email thread. And I can't help but notice that Bill Gates was a participant in this thread.
From: Brad Silverberg
To: bobkr; bradsi; jonro; rogersw; toma
Subject: RE: Novell developers
Date: Wednesday, September 22, 1993 7:52AM
paranoia runs deep.
To: bobkr; bradsi; jonro; rogersw; toma
Subject: RE: Novell developers
Date: Tuesday, September 21, 1993 9:47PM
you must be pulling my leg in asking what novell is suspicious about, right?
they think we're out to cheat them. we, of course, did this with wfw (in their eyes). we also haven't given them chicago code, and, we changed nt apis at the last minute and didn't tell them.
in the new novell math, this equals the fact that ms is not a good partner.
note that i didn't say this was logical...
>From itgmsmlbradsi Tue Sep 21 20:04:34 1993
To: bobkr, jonro, rogersw, toma
Subject: RE: Novell developers
Date: Tue, 21 Sep 93 18:45:00 PDT
X-Mailer: Microsoft Mail V3.0
what are they suspicious about? what are their concerns? did they think we'd turn down their offer and we called their bluff?
>>From: Bob Kruger
To: bradsi; jonro; rogersw; toma
Subject: RE: Novell developers
Date: Tuesday, September 21, 1993 4:45PM
I contact Novell following our mtg this afternoon. The gears are turning in Provo.
This guy, Jeff Turner, is a very suspicious hombre. We'll get what we need, but it's going to be time-consuming.
>>> From: Brad Silverberg
To: Bob Kruger; Jonathan Roberts; Rogers Weed; Tom Anderson
Cc: John Ludwig
Subject: FW: Novell developers
Date: Tuesday, September 21, 1993 14:15
if we get their free and no strings, we should ship it.
>>>From: Bill Gates
To: bradc; bradsi; jimall
Cc: jonl; paulma
Subject: Novell developers
Date: Tuesday, September 21, 1993 10:39AM
At Agenda Drew Major and Kyle Geiger were present to receive a hindsight award (10 years after the start of a product - charles got one for word and nisho for model 100).
I went up to them and congratulated them. They were super friendly as always. We agreed that the press situation was very bad. The said they totally disagreed with what Mary Burnside said about us mot providing adequate support for NT.
I said when will you make the client side software properly reentrant. Kyle said he was back at work now and was going to focus in on making the client better with Windows. I said only if there stuff was free and with no strings could we ship it. I said in any case we will keep our efforts going. They agreed that made sense.
They said Richard King who is back after 2 years absence is a good guy but that Kyle himself would try and get more technical dialog. I said I didn't know any specifics but that I thought that some of their plans for the client side were overengineering and were guaranteed to create conflicts.
This is 1995, note. I love the question: "Is this anti-competitive?" You think? Well, I expect lawyers will be happy to argue it one way and the other and then a third way, if need be. But what do *you* think?
Novell Legal Department Memorandum
From: Ryan Richards
Date: March 29, 1995
Subject: Spokane Community College Deal with the Devil
The information contained in this memorandum is PRIVILEGED
AND CONFIDENTIAL and is intended only for the
use of the individual or entity named above. If you are
not the intended recipient, you are hereby notified that
dissemination, distribution or copying of this document or
communication of its content are prohibited. If you have
received this document in error, please immediately notify
us so that it can be returned to us. Thank you.
Redacted By Court Order
Comes v. Microsoft
Mark has a good friend at SCC, Stan McDonald, who is one
year away from retirement. He
told Mark in confidence, that MS had come in and offered
SCC free MS software for faculty,
students and labs if they would dump WP and quit teaching
it in the classes. SCC took the
offer. Stan felt bad that SCC had done this deal, but he
won't say anything further because he
doesn't want to jeopardize his retirement.
Mark doesn't know anyone at SCC who will offer any more
help and he is afraid if we do
anything based alone on what Stan told him, we would put
Stan at risk.
1) Is this anti-competitive?
2) Could we get an investigator in to independently
verify this information so we would not
have to use Stan as a direct source?
([Ed: 11-page memo to Bill Gates, Steve Ballmer, Mike Maples, Pete Higgins, Jon Lazarus from Phillipe Goetschel/Jim Conner/Mark Kroese/Dave Malcom. Dated November 2, 1993, Subject: "Office Friendly Program Update"]
So, when Microsoft's attorney tells the jury that WordPerfect couldn't compete with a complete office suite, I can't help but wonder whether that isn't exactly the point, judging from this email. It appears, to my non-lawyer eyes, like Microsoft was deliberately setting it up that way, so competitors would look like they were not as good. The next exhibit increases my suspicion.
The objective of the Office Friendly Program is to evangelize some of our most visible IDG features to non-competing ISVs. The intention is to create a MS Office applications 1ook that would make our direct competitors’ applications look different and inconsistent. To date, we have discussed the program preliminarily with 18 software companies of which 15 have shown a high degree of interest. The plan is to invite them to our suite at Comdex and discuss remaining questions and contractual issues with them. We expect to have about 10 contracts signed by early December. By the end of CQl 94 we estimate to have 3-6 "Office Friendly" compliant products in the market. Over the next 12 months we plan to recruit an additional 40 ISVs for the program....
This program will not be available to any product which competes with MS Office or its component applications. It is only available to products which add value to Office by helping MS create the industry’s only "Super Suite". Participation is offered on a product-by-product basis, so a company which enrolls a non-competing product cannot include Office’s features in a competing application....
The features included in this program are relatively easy to implement, and they require no special technology. The primary deliverables are feature specifications and a copy of Word 6, which will serve as an illustration of the proper implementation. We will also include copies of our toolbar button faces and a copy of our 3D DLL....
Terminating the Program
We believe this program will add value to MS Office, but we have planned for the possibility that it will be terminated in the future. MS will reserve the right to stop selling the Feature Set Licenses at any time, which will prevent new vendors from joining the program. Current participants may continue to use the features indefinitely, but we do not believe this will harm our interests, because their products do not compete with Office. MS will also reserve the right to stop issuing logo licenses and to cancel current licenses upon 30 days notice. Products which are already displaying the logo may continue to do so for 90 days.)
Some Ideas on Exploiting Product Synergy in
Word The document comments on Microsoft's success in "stealing
significant market share away from WordPerfect", then suggests tighter
integration with other programs or functions such as email and forms so that
Word itself does not suffer the same fate as WordPerfect when faced with another competitor. (You can find particulars of how the market looked before Microsoft stole market share away from WordPerfect in this
exhibit [PDF] from 1989.)
The section in "Hyper-integration with Excel and Office" on page 3 of #5572
specifically suggests the possibility of using private interfaces between Office
applications to improve "integration".
So "synergy" was a deliberate strategy, a business model.
Mike Mathieu, Word Program Management, 7/1/93
During the past few years, we have been able to steal
significant market share away from WordPerfect. An
important lesson to learn here is how relatively easy it was
for us to do this. It should serve as a warning for Word.
In today's market, word processing might be called mission
critical, but no specific word processor can be called
mission critical. As long as there are low transition
barriers, it will be all too easy for a small, new, low-
priced competitor to come along, errode prices, and steal
business from us.
An important goal in the 3-5 year timeframe is to
increase the "mission-criticalness" or Word--at least to the
same level that an app like Excel is considered mission
critical. In the past, file formats, macros, unique
features, and app-specific user interface conventions were
all barriers to transition. In today's competitive world we
must go far beyond these former barriers and exploit synergy
with business system components that will help entrench
Word. This document serves as a starting point for the
general discussion of how to better entrench Word, and
offers some ideas on the potential areas for developing
synergy with other produces.
Besides a true word processor, the next most common
editor used by the average person is their email editor. If
we can make Word so that it can take the place of the email
editor, then users will be more satisfied--they get a more
feature rich mail editor and it's totally consistent with
their word processor--and we will be able to leverage off of
the company's choice of email system (a mission critical
system) and link it to purchases of Word. Today we don't
really integrate any better with MS Mail than somebody like
The question then becomes whether we exploit private
interfaces in Mail to make this happen, or whether we rely
on public interfaces--which our competitors can use for the
same purpose. When the email system comes with the OS, this
becomes a bigger problem. One possible idea is that this
extra level of integration only comes when using the
distinct Microsoft Mail product, as opposed to the mail
client built into Windows (Is the current plan still to ship
two separate mail clients?)
Some specifics to consider are:
Using Capone's custom forms feature to register
certain Word templates as a form (Who handles textizing?)
Use Capone's (more robust?) routing features,
rather than DAD app-specific routing.
Using Word's File Open dialog to go into the LMS
file and open any Word documents, or even preview file
attachments in other previewable formats. In general, be
able to use the LMS as a regular extension of the file
system. e.g. Find File, links, Save As, Move, Create
There are several plans circulating around now about
ways to take advantage of Word's forms and layout
capabilities as a component of a more generalized enterprise
forms and document database system. The idea is to provide
some unified front end browsing tool that lets you get into
a database, and then use a variety of forms tools like
Access, VB, or Word, to display the data in the most
It is interesting to note that Microsoft Eforms really
has no special level of integration with Word or other
desktop apps. We'd have to be able to create a new
positioning, perhaps based upon the notion of a document
database, where Word could serve as one of the value-added
forms components for this mission critical document sharing
system. (In this sense, Word would be a superior competitor
to the front end document forms client that Notes provides.
It would be up to another group to come up with the
approriate backend system to compete.) We'd want Word
to be able to tie document fields into backend databases,
provide features like lookups, and some sort of security or
data protection scheme.
The idea here is that once a user has bought into
Windows, aside from the ability to run other Windows app,
the shell itself is probably the most important component of
their user experience. (This in itself should serve as a
warning to the Chicago team. Too big of a change would
lower the barriers for competing shells.) Having Word
integrate perfectly with the shell would give us an
advantage over those word processors who didn't integrate as
well. As in the email case, this area of integration is not
able to provide our apps with a unique advantage that we
could sustain. As soon as we integrated nicely with the
shell, another app could some [sic] along and do the same
thing. This would just extend the feature wars into
integration feature wars, rather than making Word more
mission critical. (Of course, if we're in the lead in the
integration feature wars, this is a way that we could stay
ahead of competition without having so much obvious "feature
bloat" or reduced ease of learning.)
Some obvious targets for shell integration includes
making MS Write's file format identical to Word's file
format; providing document previewing filters; OLE drag and
drop scrap support; OLE Automation support; an Explorer
handler dll to provide enhanced browsing into components of
a Word document (e.g. master document pieces.)
Word as an Online Document Writer
Particularly with the growth of email, the use of rich
online documents is likely to grow significantly in the next
five years. This area talks about a set of related ideas
like making Word into a replacement for the Multimedia
Viewer, or making Word into the authoring system for Help
files, and perhaps even a display engine for help files.
Today there is a whole cottage industry of products
either built around Word or around other word processors to
allow someone to easily create help files. The manual
process is tedious, and so software developers look to these
tools to ease the process of creating the help files.
Creating a help file today requires a RTF file with special
tags to indicate specific objects. This file is then put
through a help compiler, finally producing a .hlp file that
the help engine can use. Since help files are mission
critical to software products, we could make Word a perfect
help content editor, as well as a great viewer. Hlp format
could be a standard output format by combining the
processing of the help compiler with the interfaces of
Word's export converters. Having Word read hlp files would
help solidify the hlp format as the standard format for
online documents (do we want that, or do we want .doc files
to be this standard?) Some additional issues to consider
Would we stop improving or selling the Help
Would ISV's balk at the idea of having to buy Word
to make help files?
Would customers want Word to support hlp files, or
just want Word to be able to easily create documents with
easy popups, secondary windows, hypertext jumps, and per-
The notion of using Word as a replacement for the
Multimedia Viewer implies some pretty big changes in the way
we sell and distribute Word. For instance, today we give
away MM Viewer in order to sell the content. We've never
done anything like that with Word.
If we could come up with a business model that allowed
us to give away some viewer-like version of Word, that could
help make Word's document format the de facto standard for
online information. (There are tons of CD's and online
services around today, and they all use different formats
for the data. We need to think about potential advantages
to owning the standard format for online documents--besides
the fact that Word itself would be the best content editor,
and that Word would have the highest fidelity display of
We could perhaps invest in a model similar to that of
Folio Views, where there is a full version of Views used to
create the content. Then, individual users can use Views
Lite to view the content and make edits to their own private
"shadow" file (this is essentially the same thing as a
"changes" file that I've described before in relation to
replication.) Perhaps this is a separate product that just
happens to be very similar to Word proper?
Standard Format for Information Distribution
This is an overriding theme in several of the sections
discussed in this document. By providing a standard format,
Word will also become the standard authoring tool, as well
as the standard browser and content delivery platform.
Word, by typing to workflow, and document production
systems, will become a standard tool for the distribution of
information--in effect, becoming the modern day, GUI
replacement for the old standard ASCII file. This would
include higher end, online information sources like e-books
We should also consider how Word might fit into our
plans for an online information service. Could Word serve
as the standard information browsing front end, and allow
for a textizing engine that provides for users without Word?
Tight Integration with a Search Engine
The search engine in Cairo is one of the focal points
of the product as a step towards information at your
fingertips. We need to think about ways that we can let
Word integrate better with either the Cairo search engine,
or third party engines, to further support Word's position
as the leading front end for browsing online information.
This should include a standard UI (same as Cairo, or
more value added? How do we differentiate Word?) for
searching and navigating the information hierarchy, and a
smoothly integrated interface for viewing the information.
Hyper-integration with Excel and Office
It's not clear what this means in terms of features,
but the idea is to use (private?) interfaces between Word
and Excel (and the rest of the Office?) to provide a
tremendous level of integration. This might be describing
the Integrated Office, or it might just be extensions to the
apps in the current Office. Some things to think about are:
Use Excel engine to provide a great Table object
Take more advantage of system support for
messaging. (Distributed processing? App-to-app
Implement commands which take advantage of
features in other applications.
Implement features as objects or shared components
so other apps can use them.
Complete support for Object Basic
OLE 3.0 (whatever we need it to be)
Platform (or Microsoft) standard tabular and
textual interchange formats
Component architecture taken to it's logical
extreme (explosion in variety of shared components.)
Consistency with object oriented shell and support
for OFS. (Cairo)
Native drawing tools shared
HyperIntegration and an extreme component architecture
could have a significant impact on how we market our
software as a company. This will need a lot more thought as
the time approaches.
From: Todd Millett
Again, note the date is January of 1996 and it's a Novell guy complaining about missing documentation. So Novell did complain. And what was this person provided in response to this email? Judging from the next exhibit, I'd guess nothing.
Date: 1/10/96 2:39pm
Subject: Windows 95 Group Policy Support
We are trying to get group policies to work with our client. However, there is no documented method for our client to pass group information to POLEDIT.EXE. When we looked at the GROUPPOL.INF and GROUPPOL.REG files, it became obvious that there is some interface that we are not aware of, which has not been documented.
The GROUPPOL.REG file sets a value under the Network Provider key:
the value is:
I assume this tells GROUPPOL.DLL which entry point to call to get the group and user information.
How can we implement this? Where is it documented? Is this function supposed to implemented in the Network Provider? It's not documented anywhere in the NetWork Provider spec.
Any documentation you could give us on the interface between POLEDIT.EXE, GROUPPOL.DLL, and the Network Provider DLL would be appreciated.
Update: It turns out there were complaints prior to the above email, because Millett emailed Microsoft more than once, beginning in July of 1995, and back then he was promised materials, but then he wrote again in August of 1995, and then again in the January 1996 email, so apparently no satisfaction was achieved by then. Here's the exhibit with the fuller thread, #2375:
CONFIDENTIAL So someone at Novell did complain.
From: Todd Millett
To: Internet: email@example.com
Date: 8/2/95 4:10pm
Subject: Windows 95Password Provider Header
To whom it may concern:
Below is the text of 2 messages sent previously regarding header files and
for implementing a Windows 95 Password Provider. To date, we have had no
but we still need this information. Can someone please respond with the
We have the 950 DDK, and in the network.doc file, located on that CD, these
constants are mentioned, as are the API's we need to implement:
and PPChangePassword. There is a whole chapter (Chapter 6) on this subject,
contains the same information that was in the "Utilizing the Windows 95
Control Panel" doc.
It is also well documented in the NETWORK.HLP file on the DDK CD.
In short, we have everything but the headers and libraries to actually implement
functionality. If these constants and API's have been removed, why are they so
documented? Also, if they have been removed, how do we integrated password
with Windows 95?
Any help would be appreciated.
>>> yvesm 07/20/95
I believe the constants PS_ONOFF, PWDCHANGE_MASTERPWD_NOTIFY have been
removed as well as the APIs you are referring about. The document
windows 95 password has not been updated yet.
Also you'll find the latest files in the 950 DDK that we sent to
Novell last week. However, I'll make sure you receive the updated version of
document on how to the windows 95 password.
From: Todd Millett [SMTP:TODD_MILLETT@novell.com]
Sent: Wednesday, July 19, 1995 9:56 AM
Cc: Carla Heesch@novel1.com; WIN95@nove1l.com
Subject: Password Provider constants
We are trying to implement a Windows 95 password provider for our
Netware client. We have the document "Utilizing the Windows 95 Password
Pane1", which documents the API's we need to support. However, we can't
various constants which are referenced (such as PS_ONOFF,
PWDCHANGE_MASTERPWD_NOTIFY), nor the API's themselves, referenced in any header
we have. We have looked in
NETMPR.H, but they are not there.
Could you please direct us to the proper header and library we should be using
these calls, or send them if they are not included in the standard SDK or
From: Greg Whitten [gregw]
"I want Wordperfect to understand that we consider
Opendoc as a competitor to MS Systems. We will not help them." I'd say that's clear as a bell. Jim Allchin is a recipient of this thread.
This is late 1993, November, so this is just prior to the Novell purchase, but it shows their policy at the time, which likely continued thereafter. "It is critical
maintain ownership of the interfaces and not code this
Opendoc." Ownership of the interfaces. There it is in black and white. And it's talking about WordPerfect, not NetWare.
Sent: Monday, November 22, 1993 10:51 AM
To: Bob Atkinson; Bob Muglia; Jim Allchin
Cc: Mike Maples; Paul Maritz; Tony Williams
Subject: RE: WP, OpenDoc
We have to take the hard line. We can't give OpenDoc even
an inch as easy runtime interface to OLE 2.0. We don't
to lose control of the API that ISV's are writing to. MFC
is a special case since it belongs to Microsoft. If we
OpenDoc is an OK runtime, then they will use that position
to get ISV's to write to that interface plus they now get
platform benefits, etc.
I did not understand the part of your comment about the
OpenDoc interfaces not being new and by implication
from OLE 2.0. Is Apple using OLE 2.0 interfaces or are
they wrapping them?
Three more things.
Nagel has been jerking us around. We should be taking a
harder line against Apple, IBM, and Novell and especially
any ISV's in their coalition. I.e., an ISV's
participation in CIL could disqualify them from OLE
pre-release information. We
should let WP know that in uncertain terms along with all
the other hard messages we given them.
We should start up the vaporware PR campaign for OLE 3.0
(1994) and OLE 4.0 (1996). I would like to see a brief
comparison of OpenDoc vs. OLE 2.0. I have not seen a
proposal for an OLE 3.0 feature set. Tony and Bob should
together an initial list for distribution.
I suggested quite a long time ago that we should do an OLE
assessment day for OLE 2.0 usage within the company. This
needs to happen for the compound document and automation
parts. We need to get clear feedback about OLE 2.0 from
our internal clients and what should be added, fixed to
increase the quality of our implementations, and to spread
understanding about what can be done with all the OLE 2.0
interfaces and its object model. I don't know of a better
for DAD to state their opinions and team about OLE and for
us to take the feedback and make OLE better. I don't have
anyone to set this up. Someone else should do it. This
should happen sooner than later. It should have happened
the PDC so that we could be more concrete about where OLE
would be going.
From: Bob Atkinson
To: Bob Muglia; Jim Allchin
Cc: Greg Whitten; Mike Maples; Paul Maritz; Tony
Subject: RE: WP, OpenDoc
Date: Mon, Nov 22, 1993 9:44AM
I still think there is a less hardcore position that has a
of value in causing churn inside OpenDoc. I think that in
writing it off entirely at this stage you are being too
and white about it. That we'd be so summary with WP and
wishy-washy with Apple as we are presently doing is beyond
Apple is clearly out to get us, and there's nothing that
disuade them, while it's still plausble that WP is just
| We need to be very careful about following a direction
| using opendoc to build OLE apps. Opendoc
| defines a different set of interfaces. It is critical
| maintain ownership of the interfaces and not code this
Absolutely true: we *must* retain ownership of interface
definitions. Despite their hype, though, most of OpenDoc.
largely not new interfaces. Yes, they have a couple of
that we don't have, and to retain control I was proposing
(sorry, didnt make that clear in the last mail) that as
of what we we do we'd set up and define "OLE" interfaces
interoperability standard for these features. Theres only
couple of 'em.
| Further, I am convinced from my conversation with Mark
| that Opendoc will fail, at least on Windows. They are
| further away then they claim and it is not at all clear
| they can get OLE interop working without our help.
| OLE interop, they are toast. We have significant
| with OLE, all we need to do is execute on our deliveries
| continue our ISV evangelism.
I know that it will come late. But complete failure, I'm
all so sure. Not at all. Witness the amazing amount of
reception that a piece of trash like SOM gets. 60%
can still be a big headache
| Specifically, I do not want to begin any conversations
| merging opendoc and OLE.
I agree. This would be a total mistake. But this is not at
what I'm suggesting. Nobody claimed that we merged OLE and
| We are in a position of strength.
| If we execute on the base we've established, we are
| positioned to maintain our leadership position for many
| to come. Screwing around with opendoc can only hurt us
| the ISVs who have followed our lead with OLE.
If we "screw around" then we've failed. Indeed we must do
without compromising OLE's leadership. But if we offer and
a position that is both technically truthful and socially
palatable to WP, we can end up with them getting on board
just the place we want them: as solid supporters of OLE.
| Instead, I think we need to follow-up with Wordperfect
| from a much more hardcore position. Although I need to
| more details on the politics of the situation from
| crew, I want Wordperfect to understand that we
| Opendoc as a competitor to MS Systems. We will not help
Indeed it certainly is a systems competitor today. But
mispositioning is the point: I think I'm offering a way
can be a lot less confrontational ("OpenDoc is fine, we
think you've haven't understood where it fits in") even if
the end the two approaches end up with the same actual
| make Opendoc work. They also need to understand that if
| pursue bento as their storage format, they will have
| significant disadvantage in the Cairo world where
| transparently converted to native OFS storage.
Curious: do we yet have any real numbers on the
advantage that, say, Word6 would see?
| I essentially told Mark all of that when I saw him at
| but I want to follow-up with his management so that it
| clear that wordperfect has gotten the message.
And take a look at this next one, which although it is about DR DOS, not WordPerfect, shows that a Microsoft employee said, in 1994, that the company apparently did deliberately code incompatibilities to thwart rivals.
The code's sole purpose was to "bonk" DR DOS. He found it "while considering a similar option." Novell is claiming that Microsoft bonked it with regard to its other code, WordPerfect, too, and in this same time period. Do you find that hard to believe? Of course, a trial is about proof and a decision will be made based on the evidence offered at the trial itself. But looking at these exhibits certainly tilts the field, to me anyway. Feel free to find more exhibits. Many eyeballs make the work easier.
From: Rick Segal
Subject: Dos & Windows
Date: Wednesday, January 26, 1994 10:41PM
--- Forwarded Message ---
#: 107191 (P) S4/HOT!/NOT! Software (CANOPUS)
Sb: #107063-MSFT's Software
Fm: Tony Ingenoso[PSP] 73764,3632
To: Rick Segal[Microsoft] 76276,2706
>> Microsoft does not code incompatabilities into their products.
Tread lightly here Rick. We just discovered some code in Windows whose sole
purpose seems to be to bonk DR-DOS.
It has to do with scanning low memory for the string "CON " to see if
found 10 times (the code is buggy too BTW).
This bonks any OS where the internal tables have been relocated above the 512K
mark (like DR-DOS could do).
While considering a similar option we ran into this "feature". We
around it and Windows works just fine...
Here's another, 2382:
CONFIDENTIAL Update 2:
From: Brad Young
To: TOBY COREY, WILLY DONAHOO, Fortune.SJF-MAIL1.MEPST...
Date: 8/10/95 8:03 am
Subject: Client32 Beta 1 issue
This is a brief note to inform you of a possible delay with Client32 Beta 1.
Issue: There may be a possible delay with Beta 1, due to a compatibility problem between Windows 95 gold code, and Client32.
Fidelity Investments notified Novell yesterday that after installing the "released code" version of Windows 95, Novell's Windows 95 (Alpha II) client no longer functioned. Engineering's latest available version of Windows 95 is BUILD950.RC4 (Build 950, release candidate 4). Fidelity has a later version that RC4, which does not function with the client.
Richard Jones has made multiple attempts to obtain the latest code from both John Ludwig and Yves Mitchelli (sp) at Microsoft, but his messages have not yet been returned.
Engineering will obtain a version RC6 (release candidate 6) on Thursday 8/10, and will begin testing with this version to identify the incompatibility. Engineering will also contact Fidelity to verify that the version they are seeing problems with is RC6.
If the issue cannot be readily identified and resolved, then the Beta 1 date could be negatively impacted.
I will post you a status of this issue Friday Morning.
CC: PRV-MAILS.RICHARD JONES, JWERNER, HUB-FORTUNE2.SFO ...
Here's another email thread, from 1995, which includes Bill Gates, in which upper management folks discuss how to deal with the Internet. Gates views it as a threat to their position of power, specifically controlling APIs.
From: paulma So, Gates viewed APIs as a form of power. When he is on the stand talking about the email and why he removed things Novell felt it needed, I'd want to ask him to explain this: "Given that we are looking at the Internet destroying our
position as the
setter of standards and APIs do you see things we should
be doing to use
ACT assets to avoid this?"
Sent: Wednesday, April 12, 1995 8:41 AM
To: bens; johnlu
Subject: FW: Internet (aka Web Windows)
From: Rick Rashid
Sent: Tuesday, April 11, 1995 6:04 PM
Subject: FW: Internet (aka Web Windows)
Here's the mail I referred to. BTW, Dan Ling and I
brainstormed some more after our meeting. We came up with
additional ideas and we're both going to think some more
about it this evening. One area that this "Web Windows"
excel in would be in providing "shared" application
spaces. It would be very possible for a server to send
its graphics data
more than one user simultaneously and get input from
either or both. This would be a way to provide support for
shared applications which now are next to impossible on
Maybe a brainstorming session with some key people would
be in order?
From: Rick Rashid
To: Bill Gates; Craig Mundie
Cc: Nathan Myhrvold; Russell Siegelman
Subject: RE: Internet
Date: Tuesday, April 11, 1995 3:01PM
OK - Here is a crazy idea (which frankly has nothing much
to do with ACT but does address the question of Internet
As an example of Bill's point, Dan Ling showed me this
morning that Satan (the new Internet "demon" :-)) actually
HTML as a user interface. Other software may begin to do
this. Its easy. It produces a reasonable looking machine
independent UI rather quickly. Its dangerous from our
perspective of wanting to make and preserve valuable
There are plenty of other examples.
On the other hand, the way people are increasingly using
HTML as an "interactive" network interface is extremely
Today HTML interaction is largely limited to point and
click URL selection and forms. While an amazing amount can
done that way, it is extremely simplistic and not very
satisfying to someone who wants a really interactive
At the same time, we have the existing systems like
Citrix, JoshK's work, etc where the output of a Window's
transported across a network and displayed -- the moral
equivalent of an X-Terminal. These systems have the
that they provide all of the richness of Windows
interfaces remotely. For relatively uncomplicated things
(such as Word
Processors) the Citrix solution can work well even at
relatively low 14.4 and 28.8 dialup rates -- at least
that's what they
claim. Certainly ISDN or direct Internet access can be
So here's the idea: make Windows the standard interactive
application interface for the Internet.
A URL could point to either a running or explicitly
spawned application with the protocol being a GDI-based
playback could be accomplished either through an explicit
playback engine embedded in a browser (on a Mac or Unix
system, for example) or directly through Window's actual
GDI using a pretty approach. You could go as far as the
people have in providing a complete remote "virtual
windows machine" through your browser (the Citrix people
claim to have
experimented with this) or -- more likely -- you would
provide a more special purpose way for individual
applications to have
their display windows mapped remotely. We could either
develop this technology ourselves or license work others
ATTORNEYS EYES ONLY
done (say Citrix) as a starting point.
However it was done -- and there are a number of options --
the net effect would be to encourage application
easily adapt their Windows apps (e.g. multimedia reference
titles) directly for the Internet and it would encourage
community to use Windows as the standard for interactive
network applications. NT provides a number of security
that could be taken advantage of to allow "secure"
applications to operate on a server and this could itself
be viewed as a
NT advantage. The browser software would be distributed
for free. MSN would provide equivalent services (perhaps
through the Internet) using its own NT servers, etc. I
would certainly be a way to get out in front in a new
area -- interactive
internet applications and services -- with a strategy which
maximizes use of our existing resources.
| From: Bill Gates
| To: Craig Mundie
| Cc: Nathan Myhrvold; Rick Rashid; Russell
| Subject: Internet
| Date: Monday, April 10, 1995 3:00AM
| X-Mailer: Microsoft Mail V3.0
| X-Ms-Attachment: WINMAIL.DAT 957 00-00-1980 00:00
| Given that we are looking at the Internet destroying our
position as the
| setter of standards and APIs do you see things we should
be doing to use
| ACT assets to avoid this?
| I admit I find it hard to focus lots of resources on
trials and things
| when the Internet is taking away our power every day
and will have
| eroded it irretrievably by the time broadband is
pervasive on the course
| we are on right now.
ATTORNEYS EYES ONLY
Sent: Wednesday, April 12, 1995 9:04 AM
To: bens; paulma; adamr
Subject: RE: Internet (aka Web Windows)
lots of overlap with conferencing too. we are already
signed up to do the work to remote a window to another
both point-to-point and multipoint scenarios. and the
office team is going to modify office to take greater
advantage of that
over time. we should certainly make sure that conference
objects can be embedded in web places and vice versa. this
would make every windows app today instantly "internet
capable" which is a nice benefit.
we should focus on some scenarios with mass appeal so that
this model becomes pervasive. 1:1 conferences aren't going
to turn the internet on its head, the web is not a 1:1
place. we need to think thru how we do 1:many scenarios.
if I connect to the ms web page, maybe i see a list of
free office training sessions. i can join one, and on my
screen i get [to]
see a live demo/walkthru of some office scenarios, with a
voiceover using either conferencing voice support or
progressive audio stuff.
Here is an interesting exhibit, where a Microsoft employee asks if he can make an undocumented Windows call, and he specifically states that while Microsoft publicly claimed that there were no "special hooks" available to Microsoft Apps Division programmers that were not publicly available, in fact "they do".
>we are currently working out the details of this new policy. You will receive information about this in the near future.
What is so interesting about this email thread, aside from showing that Microsoft has not always been truthful in public statements, according to this employee, who states he was aware of the duplicitous public position, but the dates. The question was sent in 1989, but it was responded to in November of 1991.
From markwa Tue Aug 29 17:03:20 1989
Date: Thu Nov 07 16:40:01 PDT 1991
David, it sounds like you're the guy to answer the question below.
>From jonm Tue Aug 29 17:03:22 1989
Date: Tue Aug 19 17:00:30 1989
Hi . . . could I ask another question about Windows? This question is indirectly related to the SDK, but if you think someone else would be more appropriate to ask, please pass it on...
The question is: do you think it is feasible to document "DefineHandleTable" for ISV programmers to use?
"DefineHandleTable" is an undocumented Windows call which is used by the Apps Division. It allows great speedups when using moveable memory, because it permits the program to find the current address of a moveable segment (or detect it it's paged out in EMM) without having to make a Windows call.
I am designing features in Microsoft C which will allow the Apps Division to use our standard retail tools, while increasing the level of support for ISV Windows programming. One of the options I am evaluating is to make a system using "DefineHandeTable" available to ISV's. (There are other good options as well.)
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. Therefore, it might be embarrassing to document "DefineHandleTable" at this late stage, as part of a system for ISV's to use. Could you give me some feedback on whether this is an option?
From markwa Wed Aug 30 08:20:48 1989
Subject: Re: Floating Point Exception App
Date: Thu Nov 07 16:40:01 PDT 1991
WinMail 1.21 brucen Thu Nov 07 16:36:32 1991Page: 31