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
Your contributions keep Groklaw going.
To donate to Groklaw 2.0:

Groklaw Gear

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


To read comments to this article, go here
Santa Cruz to DOJ in '96 & EU in '97: MS forces us to pay royalties forever for their old Xenix code we don't want
Sunday, July 01 2007 @ 05:29 PM EDT

Take a look at this with me, please, will you? It's a letter [PDF] that an outside attorney for Santa Cruz sent to the Department of Justice in 1996, complaining about a 1987 agreement between AT&T and Microsoft which oldSCO says it had inherited, with terms such that oldSCO found itself contractually compelled to include outdated Microsoft Xenix code forevermore in all its own Unix products and to pay royalties on the undesired and irrelevant code. It's Exhibit 6 in the long list of supplemental exhibits SCO filed with a Reply Memorandum [PDF] in support of its Motion for Partial Summary Judgment on its 1st, 2nd and 5th Causes of Action and for Summary Judgment on Novell's 1st Counterclaim [PDF].

As SCO put it, the agreement forced "unwilling parties to license MS software under circumstances where it is not wanted and increase[d] the financial cost of any non-Windows operating system alternative". The result? "Thus, the 1987 MS Agreement is in reality less of a common product agreement than it is a blunt instrument to keep UNIX competition in check clearing the field for Windows."

So for those who are puzzling over Microsoft's motives in doing the latest patent deals, might it be as simple as that, that Microsoft wants to be paid for all products on the market, force competing products to cost more, and to control the competition?

Folks are, after all, agreeing to pay royalties on "Something in Linux" that they claim they don't know about specifically but which might or might not infringe Microsoft patents which might or might not be valid and to accept terms of control. I'd talk about the interoperability agreements, but "Microsoft" and "interoperability" don't really belong in the same sentence. They can control what works forever and ever, by a small proprietary tweak here or there it refuses to share. But the bottom line is the same: no matter what products are sold, theirs or Microsoft's, Microsoft gets a cut. Sounds a lot like the 1987 agreement oldSCO was complaining about to me. There, SCO complained it was compelled to pay for an unnecessary product. Nowadays, Linux vendors are being asked, or else, to pay for what seems an otherwise unnecessary and undesired patent peace promise. It's quite a letter. And anyone thinking of partnering with Microsoft should probably read it first.

Of course, as Exhibit 8 [PDF], a Microsoft 1997 press release tells it while announcing in its own special way the resolution of the complaint SCO filed with the EU Commission over this same issue, SCO had it all wrong -- the Xenix code was a vital and necessary part of Unix, and while it was willing to give up the backward compatibility part of the agreement, it naturally wanted to be paid for its valuable IP. The last paragraph of this exhibit, Exhibit 9 [PDF] tells us what it says really happened: the EU Commission ruled against Microsoft, and after that, Microsoft "released" SCO from the contract except, I gather, for the money part if Xenix code were ever needed, but it was no longer compulsory. SCO apparently tried to take matters further to get paid back for the royalties already paid, but the Commission declined, Microsoft's press release seems to indicate. The contrast between Microsoft's spin and reality was highlighted by Exhibit 10, "A Tale of Two Press Releases," an article mocking Microsoft's press release. The article pretends to be a class in "Microsoft Literature 101". You can still read it here.

You never know when you're going to end up an exhibit in some courthouse.

Why newSCO would include such a document and show it to the court in purported support of its position in the Novell case is beyond me, since it says this:

From its inception, UNIX was promoted as a non-proprietary "open operating system" and was freely licensed by AT&T throughout the computer industry.

That, to my simple eyes, looks like an admission that AT&T's Unix was never patroled by the IP secrecy police the way SCO would like to pretend after the fact.

It also reveals that Kim Madsen got something else wrong, when she said at her deposition that Santa Cruz, prior to the Novell deal, had no license for Unix System V:

Q. So is it your understanding that prior to the execution of the asset purchase agreement, Santa Cruz had no rights relating to UNIX System V?

A: Yes.

It did, though. It had this crazy Xenix license it got from Microsoft by which SCO ended up owing Microsoft in perpetuity for outdated and irrelevant code nobody wanted but for which SCO still had to pay royalties:

SCO's principal product is "SCO OpenServer" ("SCOOS"). SCOOS is a PC operating system based upon UNIX which is designed to operate on computers employing Intel processors....

UNIX is an operating system originally developed by AT&T thirty years ago....SCO's rights to create, distribute and sell UNIX software code at the time it developed SCOOS were acquired through a license chain from (1) AT&T to Microsoft and (2) Microsoft to SCO. Microsoft had acquired a non-exclusive sublicensable license to UNIX from AT&T. Pursuant to its license from AT&T, Microsoft adapted UNIX to function on Intel PC's which used the 286 processor, naming the resulting program "XENIX." XENIX is thus a derivative work of UNIX. Later, in 1987 as a result of a 1987 agreement between Microsoft and AT&T described below, Microsoft developed another version of UNIX for Intel PC's using 386 processors based upon the then current release of UNIX (System V) and XENIX. This product was named "System V/386 Rel. 3.2." System V/386 Rel. 3.2 is also a derivative work of UNIX, dependent upon AT&T's UNIX license to Microsoft. In 1988, Microsoft granted SCO a license to System V/386 Rel. 3.2.

So Ms. Madsen got that part wrong. The deal with Novell wasn't until 1995, long after this MS Agreement. I wonder what she thought SCO used to write OpenServer? That should help you understand why there are rules in court about hearsay and parol evidence that Novell is asking the court to employ to toss out all such "evidence". But I don't forget that SCO's "personalized" version of the Eric Levenez Unix chart that it pulled down off of its website after we showed it was inaccurate purported to show a connection between Xenix/Sinix and Minix and Linux. SCO still pushes the Minix angle and recently it added a snarky remark about him using Sun manuals too. This is like a jigsaw puzzle, the moment when you begin to realize what the big picture might be. Was this the original litigation strategy? For Microsoft to get paid for Xenix code, as well as for SCO to get paid for SCOsource? Sun? That seems unlikely. On the other hand, the entire SCO saga is unlikely. Or was it just to show this supposed connection so SCO could prevail, based on the supposed influence, as in methods and concepts, of Xenix (and Sun manuals?) on Linus via Minix? If so, Groklaw ought to have killed it by getting Eric Levenez and Dr. Peter H. Salus to state for the record that SCO's chart asserting any such connections was inaccurate. Yet, the theme persists.

Inexplicably, SCO attaches both exhibits, 6 and 14, to the same supplemental declaration, without apparently realizing they contradict one another. I seem to read their exhibits more carefully than they do. That's probably because motivated volunteers work harder than paid employees looking at the clock, in my experience.

Exhibit 7 [PDF] is the complaint oldSCO filed against Microsoft in 1997 with the EU Commission. It mirrors the letter in many respects, and it's easier to see why SCO presents it, because the complaint alleges that Santa Cruz owned the copyright to UNIX. But since lawyers write complaints largely based on what the client tells them, as you've seen by some of the things alleged in documents BS&F has filed for its client, what a complaint claims isn't dispositive as to the truth of a matter. We already heard some oldSCO executives testifying for SCO and saying they *thought* they were getting the copyrights, but unfortunately for them, nobody so far can find a clear copyright conveyance document, despite SCO's insistence that the APA, amendments and bill of sale are kinda sorta the same as one.

The complaint details how hard Microsoft fought to force SCO to adhere to the exact terms of the 1987 agreement, by the way, and if you put it and their later press release up side by side, it's comedy gold. There's a Microsoft letter to SCO: "Microsoft expects SCO to adhere to the terms of the 1987 Agreement" Group VP of Microsoft Paul Maritz wrote, and he added that "the 1987 Agreement was explicitly negotiated to be perpetual."

If anyone could OCR this EU complaint, I'd like to put it up also. It took me forever to find the Madsen testimony, because it isn't text yet, only PDF, so I couldn't search for it except by human memory. Future historians will have the same problem, except they'll have nothing like my recall, so if you see something that hasn't been turned into text yet, and there is a lot, thanks to the flurry of filings in Novell, please lend a hand. You could start here.

The complaint seems to have similar material as the letter but it includes information about Sun Microsystem also. For example:

2.5 Sun Microsystems has sub-licensed UNIX from Microsoft. Using its sub-license it also offers a UNIX for Intel PC operating system known as "Solaris X86". Solaris X86 has differences when compared to SCOOS and UnixWare such that a user of Solaris X86 has no assurance that an application program developed for it will operate with SCOOS or UnixWare.

So according to this information, originally both Santa Cruz and Sun were sublicensees of Microsoft, not directly licensees of AT&T. Later, I gather Microsoft got out of that end of things to some degree, because when Sun wanted to license code, including, I have heard, Xenix code, it went to the SCO Group in 2003, not to Microsoft. But who got the Xenix money? Presumably that would still go to Microsoft, I would guess from the outcome of the EU complaint. But speaking for myself, I'd very much like to see any agreement finalizing that EU matter.

The complaint reiterates the nature of the agreement:

4.7 The consequences of these restrictions on competition is enormous. First, they stifle innovation in the development of new forms of UNIX for Intel PCs free of the structures, facilities and code created for 16 bit processors and application programs no longer being sold and which are as many as five generations behind the 64 bit P7. Incorporating these facilities in a program is both unnecessary and costly. Indeed, some of the programs required to be supported have not been sold for nearly a decade. Second, they compel the payment of royalties to Microsoft where none is needed or deserved. Under the Agreement, Microsoft was to be paid $15 per copy royalty for each copy of a program covered by the Agreement which was sold by AT&T or its downstream licensees. By restricting competition in the development and sale of an alternative UNIX based Intel PC program, Microsoft ensured that all such software would be subject to a royalty payment to it. In effect the provision operates like the per processor license agreements which were the subject of the Commission's earlier proceedings against Microsoft. The 1987 MS Agreement forces use of obsolete and redundant Microsoft code in circumstances where it is neither needed nor desired and it provides Microsoft with a royalty for an unnecessary product. Of course, the technical means to develop a new independent UNIX for Intel PC programs have been available at all times; the restriction on pursuing that course ensures that all such software remains under Microsoft's control.

There are small differences between the letter and the EU Complaint. For example, the year of the MS Agreement, according to the latter, is both 1988 and 1987. And at one point it calls AT&T "the new owner of Unix". I just mention this because a letter or a complaint isn't as reliable evidence as a contract. You'll remember the original complaint by Caldera (d/b/a SCO Group) compared Linux to a bicycle compared to the Unix racing car, which was so obviously inaccurate that they pulled it from the next version. And there were many other factual mistakes. Here's my favorite passage from the first complaint:

The General Public License

77. Related to the development of the open source software development movement in the computing world, an organization was founded by former MIT professor Richard Stallman entitled “GNU.”

78. The primary purpose of the GNU organization is to create free software based on valuable commercial software. The primary operating system advanced by GNU is Linux.

79. In order to assure that the Linux operating system (and other software) would remain free of charge and not-for-profit, GNU created a licensing agreement entitled the General Public License (“GPL”).

80. Any software licensed under the GPL (including Linux) must, by its terms, not be held proprietary or confidential, and may not be claimed by any party as a trade secret or copyright property.

Heaven forbid this ever shows up as "evidence" in any court case a decade from now. Let me count the ways this is hilariously wrong: Stallman was never a professor at MIT, the organization he started isn't called GNU, he very urgently would tell you he has nothing to do with the Open Source movement, had nothing to do with Linux's development, doesn't push Linux, wrote the GPL before Linus got started on the kernel, and the purpose of the GNU Project has never been to create software based on proprietary software -- oh, and GPL software doesn't have to be given free of charge or remain not-for-profit. Actually, it doesn't have to start out not-for-profit. Stallman sold emacs to pay bills for some time. [Update: Groklaw member The Cornishman notes that I missed the biggest blooper of all, that it says that GPL'd software can't be copyrighted. All GPL'd software is copyrighted. The GPL license sits on top of the copyright and it relies very much on copyright law for enforcement.] See what I mean about lawyers writing complaints? So while both these documents are of interest, neither is proof positive of anything.

Well. They prove companies make mistakes, just like people do. And they prove to me that one is wise to look before one leaps into Microsoft's arms. They do innovate. But it's in the legal area more so than in the technical, I'd say.

There's one more disturbing element in the complaint. It struck me hard, probably because I've been puzzling over why it is that when Red Hat announced great financials, its stock went down instead of up. IIRC, it's not the first time. Its stock goes down when its profits are up 17% and it beat the market -- well, it's a puzzlement to me. Lots of rumors flying around about talks with Microsoft at the same time, too. So I've been looking to understand. Now, I freely confess, I don't understand the stock market. SCO stock goes up on bad news, it seems to me, and I've never understood that either. I noticed one article attributed the drop in the price of Red Hat stock this time to a report by a Goldman Sachs analyst about expenses, and I did make a mental note that GS is the company that was on the stage with Microsoft and Novell the day they announced their god-forsaken patent deal. But that didn't seem enough of an explanation. Then I read Santa Cruz's EU complaint. Never in a million years would I have thought of something like what I read in this document, where SCO tells the EU Commission, on page 19, what Microsoft allegedly threatened to do to SCO's stock:

(3) Microsoft's position on SCO's Board

Not only does Microsoft hold a 10% share of SCO but one of its officers also occupies a place on SCO's board of directors. Although Microsoft's director is excluded from discussion at SCO relating to competing products, he nevertheless has access to a broad range of information which SCO would not wish to be disclosed to a competitor and it is therefore arguable whether his mere presence on the SCO board constitutes an "abuse" especially when one considers the conflicting fiduciary duties he faces as a director both of Microsoft and SCO.

However the "abuse" is most apparent from the way that Microsoft has sought in the past to use its shareholding in SCO. For example, Microsoft has in the past threatened to force SCO to buy back Microsoft's 10% shareholding at a time SCO did not have the funds available to make such a purchase. When informed that SCO did not have the available funds to make the purchase, Microsoft threatened to dump SCO's stock in the market, in an action which would have been likely to needlessly depress the share price of SCO's stock so as to harm SCO's other shareholders. The threat was withdrawn at the eleventh hour.

I still don't know what the story is with Red Hat, and maybe it is about expenses, but I confess I never would have thought a company would do such a thing as SCO alleged here, simple soul that I am, which is probably part of why I was able to start Groklaw. I was truly innocent as to just how down and dirty some business folks can be. And a sad education it has been. But it started me thinking. Companies don't actually do such things, do they? If I'm understanding this document, you wouldn't actually need to own stock in a company to depress the value of somebody's stock, would you? All you'd need would be to get some friendly analysts and journalists to spread a negative thought about a competitor, maybe even in all innocence. Now, that's a scary new thought for my brain to wrestle with. Hopefully, no one would sink that low.

**************************

EXHIBIT 6

[Brobeck Phleger & Harrison LLP letterhead]

September 19, 1996

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

John Greaney, Esq.
Scott N. Sacks, Esq.
Antitrust Division
Computer & Finance Section
[address]

RE: Santa Cruz Operation/Microsoft

Dear Messrs. Greaney and Sacks:

We are writing to you on behalf of The Santa Cruz Operation, Inc. ("SCO") in advance of our meeting scheduled for 9:30 a.m. October 9. This briefing memorandum contains highly sensitive confidential business information of SCO and should be maintained as a confidential document by the Department.

SCO is a software company headquartered in Santa Cruz, California and is located twenty-five miles south of the Silicon Valley. SCO's principal product is "SCO OpenServer" ("SCOOS"). SCOOS is a PC operating system based upon UNIX which is designed to operate on computers employing Intel processors. Intel processors and processors conforming to the Intel instruction set (so-called Intel "clones" such as those offered by AMD and Cyrix) comprise the vast majority of the PC market. Approximately 90% of all PC's utilize such Intel or Intel clone processors (we refer to both Intel and Intel clones as "Intel PC's").

UNIX is an operating system originally developed by AT&T thirty years ago for what were then known as minicomputers. From its inception, UNIX was promoted as a non-proprietary "open operating system" and was freely licensed by AT&T throughout the computer industry. Unlike proprietary operating systems which were unique to particular hardware vendors such as IBM's MVS or Digital Equipment's VMS,

2

John Greaney
Scott Sacks

September 19, 1996
Page 2

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

UNIX was offered by many hardware vendors and afforded the customer a degree of freedom to migrate among different hardware platforms, all employing UNIX as the operating system, while retaining existing UNIX applications with only small changes. As it has evolved, UNIX has become an extremely advanced operating system providing true multitasking (that is allowing multiple users to access a single processor), tight security (allowing different classes of users to a single computer different degrees of access); advanced networking and communication capabilities; and robustness (low rates of failure or system crashes). Indeed, UNIX was the program standard around which the Internet was originally developed. SCOOS adapts UNIX, originally developed for large systems, and enables it to function as the operating system for an Intel PC.

SCO offers a second UNIX based PC operating system known as "UnixWare." Like SCOOS, UnixWare brings UNIX to the Intel PC platform. SCO acquired the rights to UnixWare in a recent transaction with the original developer of the program, Novell. Because SCOOS and UnixWare have certain differences between them, SCO has plans to merge the two operating sytems into one program, known currently by the code name "Gemini."

SCOOS and UnixWare compete with the other operating systms on the market for Intel PC's including Windows 95, Windows 3.1, Windows NT, IBM's O/S 2 and Novell's NetWare.

SCO's rights to create, distribute and sell UNIX software code at the time it developed SCOOS were acquired through a license chain from (1) AT&T to Microsoft and (2) Microsoft to SCO. Microsoft had acquired a non-exclusive sublicensable license to UNIX from AT&T. Pursuant to its license from AT&T, Microsoft adapted UNIX to function on Intel PC's which used the 286 processor, naming the resulting program "XENIX." XENIX is thus a derivative work of UNIX. Later, in 1987 as a result of a 1987 agreement between Microsoft and AT&T described below, Microsoft developed another version of UNIX for Intel PC's using 386 processors based upon the then current release of UNIX (System V) and XENIX. This product was named "System V/386 Rel. 3.2." System V/386 Rel. 3.2, is also a derivative work of UNIX, dependent upon AT&T's UNIX license to Microsoft. In 1988, Microsoft granted SCO a license to System V/386 Rel. 3.2. Under this license agreement, SCO was permitted to copy System V/386 Rel. 3.2, which of course was almost entirely UNIX code, and to modify

3

John Greaney
Scott Sacks

September 19, 1996
Page 3

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

that code into new products. SCOOS represents years of additions and improvements to the System V/386 Rel. 3.2 software originally licensed to SCO by Microsoft. Among other things, SCO has evolved the System V/386 Rel. 3.2 code to function with modern Intel processors. XENIX and System V/386 Rel. 3.2 were 1987 vintage programs which were designed to permit UNIX to function with Intel 286 and 386 processors. (Both 16-bit processors). SCO has written SCOOS to function with the Intel Pentium, a 32-bit processor, two generations more advanced than the processor for which System V/386 Rel. 3.2 was written. So fundamental are the changes made by SCO, that SCOOS dwarfs in size System V/386 Rel. 3.2 UNIX program licensed from Microsoft. Indeed, SCO's contains nearly five times more code than the System V/386 Rel. 3.2. SCO has converted the program from a character based program to one employing a graphical user interface; added modern networking, Internet, and multiprotocol facilities; and added security features and modern device drivers.

As a result of a chain of several transactions described below, SCO has now acquired ownership of the UNIX program itself. In November 1989, AT&T, the original developer of the UNIX Operating System, spun off the UNIX division as a separate company then known as UNIX System Laboratories, Inc. ("USL"). In June 1993, Novell, the vendor of the NetWare Operating System, acquired USL and hence became the owner of the UNIX program. In turn, in December 1995, Novell sold the ownership of UNIX to SCO. As a result SCO now enjoys the right, as owner of the UNIX program, to exploit that program without the necessity of a license from any other party.

It is SCO's intention to develop a new highly advanced UNIX based operating system for the next generation of Intel processors. Currently, the most advanced Intel processor on the market is known as the "P6." This processor, now only at the start of its product life-cycle, is being sold in very small volumes at extremely high prices. Although they are not the most advanced processor chips currently offered for sale by Intel, various versions of the P5 processor, known as the "Pentium" account for overwhelming portions of current sales. Virtually all Intel PC's sold currently employ Pentium processors. Although SCO's new product, envisioned for the P7 processor, is technically speaking only one generation ahead of the P6, in reality it is two generations ahead of the main stream Intel PC's being sold. SCO's work to create a new UNIX operating system for Intel's P7 based PC's will be a tremendous undertaking, which will

4

John Greaney
Scott Sacks

September 19, 1996
Page 4

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

involve thousands of man years of engineering time. The new product, code named "NGOS" (Next Generation Operating System), will be developed from the ground up, and will be based not upon XENIX or the SCO 1988 licensing agreement with Microsoft (System V/386 Rel. 3.2) but from UNIX itself which SCO now owns.

The 1987 Microsoft/AT&T Agreement

In 1987, Microsoft and AT&T entered into an agreement entitled "Development and License Agreement for Convergence of AT&T's UNIX System V and Microsoft's XENIX Operating System on Intel Microprocessors" (hereinafter the "1987 MS Agreement"). Pursuant to the 1987 MS Agreement, Microsoft was to create a version of UNIX to run on the Intel 386 processor and to be compatible with 186 processors and programs written for the 286 PC's. (The Intel 386 processor is two generations behind the current main stream Pentium. Time has passed it by. It is obsolete and no longer being sold. The 286 is at this point but a historical curiosity. Few 286 PC's even remain in use.) The resulting adaptation of UNIX to run on the Intel 386 was termed under the 1987 MS Agreement "Merged Product." (That "Merged Product" is the System V/386 Rel. 3.2 that Microsoft licensed to SCO in 1988.) The 1987 MS Agreement contemplated that both AT&T and Microsoft would sell the resulting Merged Product. As well it provided for the two companies to develop future evolutions of the first Merged Product (the 386 version) for future releases of UNIX and for future generations of Intel processors. Those future products were never developed pursuant to the 1987 MS Agreement.

Notwithstanding the absence of evolution of the original Merged Product, the 1987 MS Agreement imposes significant restrictions on competition. It prohibits AT&T and its successors from selling any UNIX software for Intel processors (in either executable binary form or source code form) which is not a Product under the 1987 MS Agreement for as long as the 1987 MS Microsoft Agreement remains in force. The restriction on selling executable versions of UNIX for Intel PC's is found at Section 2(c) which reads:

5

John Greaney
Scott Sacks

September 19, 1996
Page 5

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

(c) As to UNIX System Code, or a derivative work thereof, in Executable File form, after one year from acceptance of the initial Merged Product, MS and AT&T shall, except as hereinafter provided, market and distribute only Binary Compatible Product for Intel Microprocessor Based General Business Computer Systems.

"Binary Compatible Product" is defined in the Agreement as a "Product" which, in turn, is defined as the "Merged Product" or derivative works thereof which are governed by the 1987 MS Agreement. Binary Compatible Products are also required to run and support a listed group of application programs written for 286 Intel processor machines.

The restriction on source code distribution is similar and found at Section 2(d):

(d) After ninety (90) days from acceptance of the initial Merged Product, any source code license granted by AT&T for UNIX System Code for an Intel Microprocessor, or any source code license granted by MS for a derivative work of UNIX System Code for an Intel Microprocessor shall be for Product only. Source code licenses granted by either party prior to the ninety first (91st) day after acceptance of the initial Merged Product shall continue in full force and effect.

Again, "Product" is a defined term in the Agreement which covers the "Merged Product" and derivative works thereof.

As a consequence of these restrictions, AT&T and its successors are prevented from offering any UNIX product for Intel PC's that is not based upon the original Microsoft "Merged Product" developed under the 1987 MS Agreement. This is to say, these restrictions compel AT&T and its successors to sell only Merged Product or derivative works based upon the 1987 Merged Product for so long as the contract remains in force.

6

John Greaney
Scott Sacks

September 19, 1996
Page 6

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

The consequences of these restrictions on competition are enormous. First, they stifle innovation in the development of new forms of UNIX for Intel PC's free of the structure, facilities and code created for 16 bit processors and application programs no longer being sold and which are as many as five generations behind the 64 bit P7. Incorporating these facilities in a program is both unnecessary and costly. Indeed, some of the programs required to be supported have not been sold for nearly a decade. Second, they compel the payment of royalties to Microsoft where none is needed or deserved. Under the 1987 MS Agreement, Microsoft estalished a continuing obligation to have paid to it a $15 per copy royalty for each copy of a program covered by the 1987 MS Agreement which was sold by AT&T or its downstream licensees. By restricting competition in the development and sale of an alternative UNIX based Intel PC program, Microsoft insured that all such software would be subject to a royalty to it. In effect, the provision operates like the peer processor license agreements which were the subject of the Department's earlier proceedings against Microsoft. The 1987 Ms Agreement forces use of Microsoft code in circumstances where it is not needed or desired and it provides Microsoft a royalty for an unnecessary product. Of course, the technical means to develop a new independent UNIX or Intel PC program have been available at all times; the restriction on pursuing that course insures that all such software remains under Microsoft's foot.

The anticompetitive effect of these restrictions is magnified by the term provisions of the Agreement which keep the Agreement in force, and thus the restrictions and royalty provisions in force, until such time as neither party (AT&T and its successors or Microsoft) has commercially released a new generation product for a new Intel processor or new release of UNIX for a period of two years. (See Section 14) The 1987 MS Agreement in every practical respect is thus evergreen. It will continue with its restrictions in force under its express terms forever unless both parties have failed to offer products for new Intel processors or new variations of UNIX. A two year hiatus in the offer of new UNIX software products for new Intel processors or new releases of UNIX, necessary to release AT&T or its successors from the 1987 MS Agreement is in all commercial respects equivalent to termination of business. It is well known in the electronics industry that such a failure to advance product offerings would cause the customer base to migrate irrevocably to other competitive up-to-date operating system products -- in this circumstance undoubtedly those offered by Microsoft.

7

John Greaney
Scott Sacks

September 19, 1996
Page 7

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

Effect on SCO

The anti-competitive restrictions in the 1987 MS Agreement significantly imperil SCO's development of NGOS. First, the restrictions by their express terms prevent effective product innovation to create a new product for the 64 bit P7 processors. Rather than being free to take the UNIX software that it now owns and develop a revolutionary new UNIX operating system to run on the P7 processor, the terms of the 1987 MS Agreement constrain SCO to manipulate and adapt Microsoft's original work done for 16 bit processors. The effect hobbles innovation since it compels continued use of features and constructs no longer relevant ten years after the 1987 MS Agreement was made.

By handcuffing all future product evolution of UNIX for Intel PC's to its original work done for the 386 chip in 1987, Microsoft seeks for itself a competitive advantage in its offer of its own programs Windows 95 and Windows NT which compete with UNIX. Although, in theory, Microsoft is subject to the same restrictions under the terms of the 1987 MS Agreement -- it, too, can only offer UNIX for Intel PC's which is based upon or a derivative work of Merged Product -- the restrictions have no meaning in reality for Microsoft. After it developoed the so-called Merged Product for the 386, Microsoft decided not to bring it to market. It has not offered for sale any UNIX for Intel PC product for years. Hence it is not restricted at all. The restriction operates only upon Microsoft's competitor SCO and upon programs that compete with Microsoft.

The restrictions in the 1987 MS Agreement significantly impact SCO's NGOS in a second way. By undertaking the expense and burden of developing a new product, SCO should not be subject to payment of royalty to Microsoft. By enforcing a provision which requires that new products continue to be a derivative work of the Microsoft 386 "Merged Product," and therefore subject to royalty, no matter how irrelevant the 386 program version is four chip generations and ten years in the future, Microsoft imposes a significant financial penalty on a competitor and, thereby, a competitive price advantage for its Windows operating system products.

By restricting the development of a PC operating system that directly competes with its products and saddling competitive product offerings with unnecessary links to old software and making them subject to an unnecessary royalty, Microsoft acts unfairly to perpetuate its monopoly in the supply of PC operating systems.

8

John Greaney
Scott Sacks

September 19, 1996
Page 8

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

We noted above the manner in which the restrictions in the 1987 MS Agreement operate with similar effect as per processor agreements which were the subject of the Department's prior actions and resulting consent decree. Both the per processor agreements and the 1987 MS Agreement force unwilling parties to license MS software under circumstances where it is not wanted and increase the financial cost of any non-Windows operating system alternative.

The 1987 MS Agreement as well has effects which duplicate, in this UNIX setting, other past anti-competitive DOS/Windows practices, covered by the previous action and Consent Degree. Sections IV B of the Consent Decree prohibits Microsoft from entering into license agreements which restrict OEM's from sale or distribution of non-Microsoft operating systems. The 1987 MS Agreement has exactly this effect. SCO, under its terms would be prohibited from offering new UNIX products free of the obsolete Microsoft code. The entire field of UNIX for Intel PC's, thus, becomes one where every product is a product subject to Microsoft royalty payments. Indeed the provisions of the 1987 MS Agreement are even more pernicious than those covered by the previous decree covering DOS and Windows because in this case the prohibition against using non-Microsoft software both stifles innovation for competitive UNIX products and forces inclusion of obsolete Microsoft software making UNIX products less desirable.

The term of the restrictions in the 1987 MS Agreement similarly runs afoul of the spirit of the Consent Decree in Section IVA. There Microsoft is prohibited from license agreements with terms longer than one year. Here Microsoft has created a structure, that in theory is perpetual. Indeed, the 1987 MS;s prohibition against new product development is far worse than that found to be copyright misuse in Lasercomb America, Inc. v. Reynolds, 911 F.2d 970 (4th Cir. 1990). There imposition of a 99 year restriction on development of competitive software, as a condition of the grant of a software license was found to be a form of copyright misuse. The circumstances here are threefold more exaggerated in their anti-competitive effects: (i) Microsoft possesses monopoly power which was not the case with the Lasercomb licensor; (ii) the "license" of Microsoft software is not desired by the licensee for new products, but it forced; and (iii) the term of the restriction is potentially longer.

In addition to the operation of the 1987 MS Agreement so as to enhance Microsoft's monopoly power in the market for PC operating systems, it can easily be

9

John Greaney
Scott Sacks

September 19, 1996
Page 9

PRIVILEGED & CONFIDENTIAL
CONTAINS BUSINESS SECRETS

seen that there is no efficiency enhancing integration of economic activity which flows from the Agreement. For purposes of developing a common version of UNIX, an agreement between AT&T and Microsoft to create such a work together, simpliciter, would have sufficed. Extending the arrangement beyond that boundary, so as to bar creation of competing works, does not legitimately further that purpose. On that ground alone the arrangement violates the antitrust laws. More than that, the 1987 MS Agreement actually operates to suffocate the "common" version of UNIX. Where Microsoft does not sell the "common" UNIX product, but competes against it with Windows, saddling all future versions with useless code and an unnecessary royalty furthers the demise of the "common" UNIX product to the benefit of Microsoft's own products. Under the present circumstances, Microsoft derives greater profit from the failure of UNIX for PC's in the marketplace, than success of any "common" UNIX product. Thus, the 1987 MS Agreement is in reality less of a common product agreement than it is a blunt instrument to keep UNIX competition in check clearing the field for Windows.

We look forward to our meeting on October 9 to discuss these matters further.

Respectfully submitted,

[signature]
Jeffrey S. Kingston

JSK:nm

10


  View Printable Version


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 )