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
Novell Files its Opening Appeal Brief in WordPerfect Antitrust Litigation v. Microsoft ~pj Updated
Monday, November 26 2012 @ 01:51 PM EST

Novell has filed its opening appeals brief [PDF] in the Novell v. Microsoft antitrust litigation regarding WordPerfect.

As you know, appeals are about mistakes of law. So the brief lists the mistakes Novell believes the district court judge made, but you see also how Boies Schiller has managed to tell a complicated story in a simple way that is easy to grasp. I assume it's their contribution to the team effort, because that's what David Boies is famous for. Ted Olson put it like this once a couple of years ago: "Above all, Boies is a storyteller. Nobody is better able to describe the gist of a case." And Boies said:

...Boies acknowledges his ability to cut to the chase in jargon-infused litigation. "It is easy to be accurate if you have the freedom to be complicated, and it is very easy to be simple if you have the freedom to shade the truth," he says. "What's hard is to be simple and very accurate, and that takes work to figure out what are the simple truths that are going to sustain your case."
The brief tells this simple story, that what Microsoft did to Novell when it withdrew its APIs left it with a choice between "two different ways to commit suicide" -- use what they had and meet the deadline, with the obvious result of getting bad reviews, or take the time to get it right and miss the release of Windows 95.

It's something that Dr. Roger Noll, one of Novell's experts mentioned [PDF] during the trial (see page 37), but the brief turns it into the theme of the appeal, and it's brilliant, in that it neuters the district judge's misguided opinion that Novell had viable options that it failed to pursue.

When it comes to technology, though, it's possible to be too simple for complete accuracy. In the brief, for one example, the definition of operating system is given as "software that performs the core functions of operating a computer." That doesn't sound right to me. I can't say that's totally wrong, exactly, depending on how you define 'core', but I'd instead define it as Wikipedia does, as "a collection of software that manages computer hardware resources and provides common services for computer programs", acting as an intermediary, in other words, between the hardware and the applications. Maybe that is what they mean by "core", but it's too vague a word, to me, to be precisely right. A judge who isn't a geek might think writing Word documents is "core", for example, and the district court judge seemed mighty confused about the difference between operating systems and middleware. So it matters.

In SCO v. IBM, David Marriott of Cravath explained operating systems the best, I think, in a hearing in July of 2004:

MR. MARRIOTT: An operating system, Your Honor, is at the heart of this case. In fact, a series of operating systems are. Without its software, a computer -- the computer sitting on the desk in front of Your Honor is essentially a useful lump of metal. With its software with the operating system, the computer can perform a number of useful functions. There are basically two kinds of software. There is a system software on the one hand which controls the operation of the computer itself; and there is application software on the other hand which allows the user to perform a particular function.

In the most fundamental kind of -- the system software is the operating system, and it's the operating system that is the issue in this case. The operating system is interfaced between user and the lump of metal on your desk. So when Your Honor, for example, writes a letter using the computer that sits in front of you, you interact with that lump of metal by the operating system. You might use a program, for example, like Microsoft Word. That's an application program that assists the operator and allows you to perform that particular function.

Now, operating systems are originally written in a programming language prepared by human beings which is known as source code. The source code consists of thousands of files, Your Honor, and millions of lines of code. And with the Court's permission, I'd like to illustrate, if I may, a source code by handing a copy of that to the Court.

This, Your Honor, is an illustration of the source code that comprises an operating system. And you'll see the numbering of lines along the left. There are in this particular file 3,070 lines of source code. An operating source code can be comprised of many millions of source codes. The computer has to translate the code into a language which is usable machine language, and it does that by a device called compiler.

And that's not the same thing as middleware. Here's what the definition of middleware is, again from Wikipedia:
In its most general sense, middleware is computer software that provides services to software applications beyond those available from the operating system. Middleware can be described as "software glue". Thus middleware is not obviously part of an operating system, not a database management system, and neither is it part of one software application. Middleware makes it easier for software developers to perform communication and input/output, so they can focus on the specific purpose of their application.
I want to be positive that Boies Schiller understands this before oral argument, because the lower court messed it up quite a lot, because he thought there was no realistic middleware threat. So if you have a better definition, by all means post it. But for heaven's sake, Java is middleware, and for sure Microsoft saw it as a threat. In the trial, Microsoft had one definition of middleware and Novell had another, and you will not be surprised to learn that the judge accepted Microsoft's definition, and you can read all about it in the Findings of Fact's section on middleware on this page.

This is the 3rd appeal in this litigation. The brief is 117 pages long, because the court denied two requests for more space. Otherwise it would have been longer. They asked for 26,500 words, and the court said no. So then they asked for 21,500 words, and the court said no again. So 117 pages had to do. Actually, about 70 of text, the rest being the opening materials like the table of contents and the list of cases and the attachments and certificates of service at the end. It is exactly the same length as SCO's first appeal brief, which was 87 pages but only 70 of text. Here, Novell's is also 70 pages of text. So, yes, they managed to do it.

Here are Novell's main headings, the outline of what happened, and what Novell asserts Microsoft did wrong:

1) Novell's popular WordPerfect and its middleware threatened Microsoft's monopoly in the operating systems market;
2) Microsoft induced Novell to rely on its namespace extension APIs;
3) then Microsoft withdrew support for the APIs to disadvantage its competitors;
4) Microsoft deceived Novell as to the real reason for de-documenting the APIs; and
5) this disadvantaged Novell by delaying its product releases, hence eliminating the threat to Microsoft's monopoly power.
Here's how the introduction describes Microsoft's allegedly anticompetitive conduct:
As in the Government’s case against Microsoft, United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001), Microsoft destroyed a competitive threat not by designing better products or lowering prices, but by impeding and deceiving competitors. Here, Microsoft induced Novell to design its applications to rely on a critical technology (called namespace extension APIs) in the Windows 95 operating system. Almost a year later, Microsoft withdrew support for those APIs when Novell could not timely redesign its applications. As a result, Microsoft left Novell with what economist Roger Noll characterized as “two different ways to commit suicide” – fail to release its product within the critical 90-day period following Windows 95’s release, or release a non-competitive product that sacrificed key functionality.
The brief itemizes some key legal mistakes made by the judge. First, the brief states, he provided no basis for disregarding the evidence Novell presented that Microsoft's common file open dialogue was not a viable competitive option. Second, the district court's "mistaken inferences and conclusions concerning Novell's complaints to Microsoft" provided no valid basis for finding an absence of causation. And finally, the court's conclusion that Quattro Pro caused Novell's delay instead of Microsoft's conduct was a disputed question of fact that should be decided by a jury, not a judge. That last is important, because the judge gave Microsoft the victory on summary judgment, and the rule for that is that you can't rule for a party seeking summary judgment if there are disputed factual issues.

Regarding the judge's mistakes, Novell describes them like this:

The District Court disregarded the governing legal standard, which provides that affirmative conduct is anticompetitive when, as here, it harmed rivals, was not “competition on the merits,” and was reasonably capable of contributing significantly to maintaining monopoly power. Multistate Legal Studies, Inc. v. Harcourt Brace Jovanovich Legal & Prof’l Publ’ns, Inc., 63 F.3d 1540, 1550 (10th Cir. 1995). The District Court did not discuss the Tenth Circuit case law recognizing the applicable standard of liability on the grounds that the facts governing the District Court’s ruling were so “case specific.” JA-199 n.5. But that case law establishes that Microsoft’s conduct was anticompetitive because the reasons Microsoft offered for disadvantaging a competitor were pretextual and the conduct disadvantaged competition and was reasonably capable of contributing significantly to maintaining Microsoft’s monopoly power.

Furthermore, in evaluating the harm caused by Microsoft’s conduct in the operating systems market, the District Court improperly required Novell to prove that its applications, by themselves, would have eliminated (rather than diminished) the applications barrier to entry protecting Microsoft’s monopoly power. The availability of a popular competitive word processor alone, however, would have been important for any operating system to compete with Windows.

In addition, the District Court disregarded the ample evidence of the threats posed by Novell’s franchise applications and middleware, and departed from the D.C. Circuit’s opinion, the law of the case, and decades of precedent, by placing the burden on Novell to recreate the hypothetical marketplace that would have existed but for Microsoft’s improper conduct.

Further, on each ground for its decision, the District Court disregarded the governing standard of review, which permits ruling in Microsoft’s favor only if no reasonable jury could have found for Novell. The District Court repeatedly overlooked key evidence and resolved conflicting testimony in favor of Microsoft rather than Novell. Because a reasonable jury could have found for Novell on the evidence presented, the District Court’s decision must be reversed and the case remanded for trial ....

The District Court’s decision to forego the established legal framework for unlawful monopolization cases resulted in a decision that cannot be reconciled with the law of this Circuit. Under Tenth Circuit precedent, anticompetitive conduct is that which impairs the opportunities of rivals and is not competition on the merits, “if the conduct appears reasonably capable of contributing significantly to creating or maintaining monopoly power.” Multistate, 63 F.3d at 1550 (citation and internal quotation marks omitted). A jury could reasonably conclude from the record below that Microsoft’s affirmative conduct impaired the opportunities of actual and potential rivals and that Microsoft engaged in that conduct solely to disadvantage competitors, not to design a better product or otherwise compete on the merits. Indeed, the District Court’s finding that a reasonable jury could have found Microsoft’s proffered justifications to be pretextual provided sufficient grounds to conclude that its conduct was not “competition on the merits.”

Additionally, the District Court improperly rejected deception as anticompetitive conduct by resolving an evidentiary conflict that properly was for the jury regarding whether Microsoft was aware of Novell’s reliance on the APIs.

The District Court also wrongly held that Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985), required Novell to show that Microsoft terminated its entire relationship with Novell.

Further, the District Court wrongly concluded that eliminating Novell as a competitive threat did not cause harm to competition in the operating systems market. Novell presented ample evidence that established the legitimacy of the threat to Microsoft’s monopoly power from Novell’s applications and middleware. The District Court erred by disregarding this evidence and adopting a test that effectively would allow monopolists to crush nascent threats to their monopoly power seriatim. This test departed markedly from the legal standards applied by this Circuit, the D.C. Circuit in the Government’s case against Microsoft, and even by the District Court and Fourth Circuit on summary judgment.

Finally, the District Court wrongly concluded that there was no “underlying business reality,” JA-226-27, to the claim that Microsoft caused fatal delay in the release of Novell’s product. Here, the District Court substituted its judgment for that of the factfinder, including disregarding the ample evidence that Microsoft left Novell with no viable options when Microsoft de-documented the namespace extension APIs.

They also mention that the judge declared a mistrial after only three days of deliberations. Then, despite him finding that it “appears undisputed that eleven of the twelve jurors would have returned a verdict in favor of Novell on the issue of liability,” the judge granted Microsoft's motion for judment as a matter of law.

If I were to make such a list, it would include all the above and much more. It was quite a performance. It's not every day that you see lawyers rolling their eyes in despair during a trial, but I've never seen such a trial. If you are curious, we have the entire trial transcripts, day by crazy day, which you can find on our docket page beginning here or here. The trial lasted nearly two months, beginning on October 17, 2011, with a mistrial declared on December 16, 2011.

We have two more transcripts to add to our collection, the transcript [PDF] of the hearing held on June 7, 2012 on Microsoft's motion for judgment as a matter of law (where you'll find all the documents filed about that motion) and a previously sealed transcript [PDF] of one day of the trial where Bill Gates was on the stand, on November 21, 2011. Don't get excited. It's on this level of truth:

Q. Did Microsoft tell ISVs about Windows 95 even before there was code running for the new operating system?

A. Yes, we did. We were very open about what we were planning to do. It was very helpful to us to get people's feedback. It shaped how we prioritized the work. The number of things we could have done was so large that there was no way that we could do them all, and that marketplace input shaped what eventually went into the Windows 95 product.

That openness all happened on planet NeverNeverLand, one presumes.

The last four transcripts still under restrictions from this case -- from the hearings dated September 29, 2011, October 6, 2011, October 7, 2011, and October 13, 2011 -- will be available on the 26th, today, so I'll add them when I get them. Sometimes it takes a few days, despite the notice dates, for them to show up.

Here's the filing:

11/21/2012 - [10021391] Appellant/Petitioner's brief filed by Novell, Inc.. 7 paper copies to be provided to the court. Served on 11/21/2012 by ECF/NDA. Oral argument requested? Yes. [12-4143] JMJ

Update: Microsoft's date to reply is now set:
11/26/2012 -    [10022093] joint Appendix filed by Appellant Novell, Inc. Original and 1 copy. Appendix pages: 16,706. Number of volumes: 24. (Volume 24 is provisionally SEALED.) Hardcopy only. Served on 11/21/2012. Manner of Service: FedEx. [12-4143]

11/26/2012 -    [10022095] Minute order filed - Appellee's brief now due 01/23/2013 for Microsoft Corporation. (Deadline set per Circuit Mediation's 9/28/12 letter/order.) (Text Only - No Attachment) [12-4143]

So, January 23rd.

And I have the final four transcripts:

  • 517 Hearing held on September 29, 2011

  • 518 Hearing held on October 06, 2011

  • 519 Hearing held on October 07, 2011

  • 520 Hearing held on October 13, 2011

All PDFs. Enjoy.

Here it is as text, Novell's opening appeals brief:

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

CASE NO. 12-4143

IN THE UNITED STATES COURT OF APPEALS
FOR THE TENTH CIRCUIT

NOVELL, INC.,

Plaintiff – Appellant,

v.

MICROSOFT CORPORATION,

Defendant – Appellee.

_________________

On Appeal from the United States District Court
For the District of Utah
The Honorable Judge J. Frederick Motz
D.C. No. 2:04-CV-01045-JFM

_____________

APPELLANT’S OPENING BRIEF

Jeffrey M. Johnson
James R. Martin
Miriam R. Vishio
DICKSTEIN SHAPIRO LLP
[address, phone, fax]

David Boies
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]

(Additional counsel listed on inside cover)

Oral Argument is requested

SCANNED PDF FORMAT ATTACHMENTS ARE INCLUDED

November 21, 2012

Max D. Wheeler
Maralyn M. English
SNOW, CHRISTENSEN & MARTINEAU
[address, phone, fax]

Stuart H. Singer
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]
Samuel C. Kaplan
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]

R. Bruce Holcomb
ADAMS HOLCOMB LLP
[address, phone, fax]

CORPORATE DISCLOSURE STATEMENT

Pursuant to Fed. R. App. P. 26.1, Appellant Novell, Inc. makes the following disclosure: Novell is owned by The Attachmate Group, Inc. and Attachmate Corporation. The Attachmate Group is wholly owned by Wizard Parent, LLC, which is in turn owned by Thoma Cressey Bravo, Inc. and Thoma Bravo, LLC. The Attachmate Group, Attachmate Corporation, Wizard Parent, Thoma Cressey Bravo, and Thoma Bravo are privately held corporations. No publicly held corporation owns 10 percent or more of the stock of Novell.

i

TABLE OF CONTENTS

Page

CORPORATE DISCLOSURE STATEMENT ..........................................................i

TABLE OF CONTENTS..........................................ii

ATTACHMENTS................................................v

TABLE OF AUTHORITIES .............................. vi

PRIOR OR RELATED APPEALS............................... ix

GLOSSARY ....................................................... x

INTRODUCTION ..................................................1

STATEMENT OF JURISDICTION.............................4

STATEMENT OF THE ISSUE.........................................4

STATEMENT OF THE CASE.........................................5

STATEMENT OF FACTS ...............................6

I. NOVELL THREATENED MICROSOFT’S MONOPOLY POWER
IN THE OPERATING SYSTEMS MARKET ...............................6

A. WordPerfect’s Popularity And Cross-Platform Capability
Threatened Microsoft’s Monopoly Power In The Operating
Systems Market ..............................................8

B. Novell’s Middleware Threatened Microsoft’s Monopoly Power In
The Operating Systems Market......................................10

II. MICROSOFT INDUCED NOVELL TO RELY ON THE
NAMESPACE EXTENSION APIS IN DESIGNING ITS
WINDOWS 95 APPLICATIONS ..................................13
A. The Namespace Extension APIs Enabled Novell’s Applications To
Access The Windows 95 Namespaces And Integrate With The
Windows Shell...................................13

ii

B. Microsoft Induced Novell To Rely On The Namespace Extension
APIs In Designing Its Applications............................16
III. MICROSOFT WITHDREW SUPPORT FOR THE NAMESPACE
EXTENSION APIS TO DISADVANTAGE COMPETITORS ...................18

IV. MICROSOFT DECEIVED NOVELL AND OTHER ISVS
REGARDING THE TRUE REASONS FOR DE- DOCUMENTATION .............................19

V. MICROSOFT’S ANTICOMPETITIVE CONDUCT DELAYED
NOVELL’S PRODUCT RELEASES AND ELIMINATED THE
COMPETITIVE THREAT THEY POSED TO MICROSOFT’S
MONOPOLY POWER..................................22

SUMMARY OF THE ARGUMENT .....................................26

ARGUMENT .........................................................28

I. STANDARD OF REVIEW ..................................28

II. MICROSOFT’S CONDUCT WAS ANTICOMPETITIVE .........................29

A. Microsoft’s Conduct Was Not Competition On The Merits And
Eliminated Novell As A Competitor ...........................31
1. Microsoft Induced Novell’s Reliance On The Namespace
Extension APIs And Thereby Made Novell And Competition
Worse Off Than If Microsoft Had Refused To Make Them
Available From The Outset .......................................31

2. Microsoft’s Conduct Was Deceptive .....................................35

3. Microsoft Altered A Voluntary Course Of Dealing Without A
Legitimate Competitive Reason In Violation Of Aspen Skiing ..........38

B. Microsoft’s Conduct Was Reasonably Capable Of Contributing
Significantly To Maintaining Microsoft’s Monopoly Power In The
Operating Systems Market ..........................................43
1. The District Court Applied An Incorrect Legal Standard...................43

iii

2. The Elimination Of Novell’s Key Franchise Applications Was
Reasonably Capable Of Contributing Significantly To
Maintaining Microsoft’s Monopoly Power..........................45

3. The Elimination Of Novell’s Middleware Was Reasonably
Capable Of Contributing Significantly To Maintaining
Microsoft’s Monopoly Power .................................49

III. NOVELL PRESENTED AMPLE EVIDENCE THAT
MICROSOFT’S ANTICOMPETITIVE CONDUCT CAUSED
NOVELL’S INJURY .............................54
A. The District Court Provided No Basis For Disregarding The
Considerable Evidence That Microsoft’s Common File Open
Dialog Was Not A Viable Competitive Option .....................54

B. The District Court’s Mistaken Inferences And Conclusions
Concerning Novell’s Complaints To Microsoft And Executive
Involvement Provided No Basis For Finding An Absence Of
Causation ..............................................56

C. The District Court’s Conclusion That Quattro Pro, Not Microsoft’s
Conduct, Caused The Delay Improperly Resolved A Disputed
Question Of Fact That Was For The Jury ................................60

CONCLUSION ............................................62

REQUEST FOR ORAL ARGUMENT ..........................................64

CERTIFICATE OF COMPLIANCE..................................65

CERTIFICATE OF DIGITAL SUBMISSION
AND PRIVACY REDACTIONS.................................66

CERTIFICATE OF SERVICE ...........................................67

iv

ATTACHMENTS

No.

Novell, Inc. v. Microsoft Corp., No. 2:04-CV-01045-JFM,
Opinion on Microsoft’s Rule 50(b) Motion (Dkt. #509, JA-197-2291)..........1

Novell, Inc. v. Microsoft Corp., No. 2:04-CV-01045-JFM,
Order on Microsoft’s Rule 50(b) Motion (Dkt. #510, JA-230) ......................2

Novell, Inc. v. Microsoft Corp., No. 2:04-CV-01045-JFM,
Entry of Judgment (Dkt. #511, JA-231)...............................3

v

TABLE OF AUTHORITIES

Page

A.I.B. Express, Inc. v. FedEx Corp.,
358 F. Supp. 2d 239 (S.D.N.Y. 2004) ........................................42

Aspen Skiing Co. v. Aspen Highlands Skiing Corp.,
472 U.S. 585 (1985)......................................passim

Bigelow v. RKO Radio Pictures, Inc.,
327 U.S. 251 (1946)...................................................45

Christy Sports LLC v. Deer Valley Resort Co.,
555 F.3d 1188 (10th Cir. 2009) .....................................passim

Continental Ore Co. v. Union Carbide & Carbon Corp.,
370 U.S. 690 (1962)............................................30

ES Dev., Inc. v. RWM Enters., Inc.,
939 F.2d 547 (8th Cir. 1991) .......................................34

Guides, Ltd. v. Yarmouth Grp. Prop. Mgmt., Inc.,
295 F.3d 1065 (10th Cir. 2002) ...................................28

Instructional Sys. Dev. Corp. v. Aetna Cas. & Sur. Co.,
817 F.2d 639 (10th Cir. 1987) .........................................30

King & King Enters. v. Champlin Petroleum Co.,
657 F.2d 1147 (10th Cir. 1981) ........................................44

Lantec, Inc. v. Novell, Inc.,
146 F. Supp. 2d 1140 (D. Utah 2001),
aff’d, 306 F.3d 1003 (10th Cir. 2002)...........................................31

LePage’s Inc. v. 3M,
324 F.3d 141 (3d Cir. 2003) ..............................................30

MetroNet Servs. Corp. v. Qwest Corp.,
383 F.3d 1124 (9th Cir. 2004) ........................................ 41-42

Multistate Legal Studies, Inc. v. Harcourt Brace Jovanovich
Legal & Prof’l Publ’ns, Inc.
, 63 F.3d 1540 (10th Cir. 1995)......................passim

v

Nobody in Particular Presents, Inc. v. Clear Channel Commc’ns, Inc.,
311 F. Supp. 2d 1048 (D. Colo. 2004)....................................31, 42

Novell, Inc. v. Microsoft Corp. (In re Microsoft Corp. Antitrust Litig.),
699 F. Supp. 2d 730 (D. Md. 2010).....................................5, 44

Novell, Inc. v. Microsoft Corp. (In re Microsoft Corp. Antitrust Litig.),
No. 1:05-cv-01087, 2005 WL 1398643 (D. Md. June 10, 2005),
aff’d, 505 F.3d 302 (4th Cir. 2007)...........................................5

Novell, Inc. v. Microsoft Corp.,
505 F.3d 302 (4th Cir. 2007) ..............................................43

Novell, Inc. v. Microsoft Corp.,
429 F. App’x 254 (4th Cir. 2011).............................................5

Phillips v. Hillcrest Med. Ctr.,
244 F.3d 790 (10th Cir. 2001) ............................................28

Poller v. Columbia Broad. Sys., Inc.,
368 U.S. 464 (1962).............................................................34

Rambus, Inc. v. FTC,
522 F.3d 456 (D.C. Cir. 2008)..........................................36

Reeves v. Sanderson Plumbing Prods., Inc.,
530 U.S. 133 (2000)......................................................29

Safeway Inc. v. Abbott Labs.,
761 F. Supp. 2d 874 (N.D. Cal. 2011)...........................42

Shaw v. AAA Eng’g & Drafting, Inc.,
213 F.3d 519 (10th Cir. 2000) ...........................................28

Strickland v. United Parcel Serv., Inc.,
555 F.3d 1224 (10th Cir. 2009) ...................................28

United States v. Grinnell Corp.,
383 U.S. 563 (1966)...............................................29

United States v. Microsoft Corp.,
253 F.3d 34 (D.C. Cir. 2001)........................................passim

vii

United States v. Microsoft Corp.,
84 F. Supp. 2d 9 (D.D.C. 1999)...........................................6

Verizon Commc’ns Inc. v. Law Office of Curtis V. Trinko, LLP,
540 U.S. 398 (2004).................................................29, 39, 40

Weese v. Schukman,
98 F.3d 542 (10th Cir. 1996) .........................................28

Wichita Clinic, P.A. v. Columbia/HCA Healthcare Corp.,
No. 96-1336-JTM, 1997 WL 225966 (D. Kan. Apr. 8, 1997) ..................31

Statutes

15 U.S.C. § 2...........................................5, 29, 43, 44

28 U.S.C. § 1291................................................................4

28 U.S.C. § 1331.......................................................4

28 U.S.C. § 1337..............................................................4

Rules

Fed. R. Civ. P. 50 ............................................passim

viii

PRIOR OR RELATED APPEALS

Novell, Inc. v. Microsoft Corp., 429 F. App’x 254 (4th Cir. 2011)

Novell, Inc. v. Microsoft Corp., 505 F.3d 302 (4th Cir. 2007)

ix

GLOSSARY

API: Application programming interface

Athena: Codename for Microsoft’s Internet Mail and News client application

Chicago: Codename for Microsoft’s Windows 95 operating system

DOJ: The United States Department of Justice

FOF: Finding of Fact from United States v. Microsoft Corp., 84 F. Supp. 2d 9 (D.D.C. 1999)

ISVs: Independent software vendors

Marvel: Codename for the Microsoft Network (“MSN”) client application

MS-DOS: Microsoft’s Disk Operating System

PC: Personal computer

Win95: Microsoft’s Windows 95 operating system

WPWin: WordPerfect for Windows

x

INTRODUCTION

Novell, Inc. appeals the District Court’s entry of judgment as a matter of law in favor of Microsoft Corporation. Novell alleges that Microsoft unlawfully maintained its monopoly power in the Intel-compatible personal computer (“PC”) operating systems market through anticompetitive conduct that destroyed the competitive viability of Novell’s WordPerfect word processing application and PerfectOffice applications suite. There is no dispute that Microsoft was a monopolist and maintained its monopoly power in the operating systems market. The only questions for the jury were whether Microsoft unlawfully maintained its monopoly power through anticompetitive conduct and whether that conduct proximately caused Novell’s injury.

As in the Government’s case against Microsoft, United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001), Microsoft destroyed a competitive threat not by designing better products or lowering prices, but by impeding and deceiving competitors. Here, Microsoft induced Novell to design its applications to rely on a critical technology (called namespace extension APIs) in the Windows 95 operating system. Almost a year later, Microsoft withdrew support for those APIs when Novell could not timely redesign its applications. As a result, Microsoft left Novell with what economist Roger Noll characterized as “two different ways to commit suicide” – fail to release its product within the critical 90-day period

1

following Windows 95’s release, or release a non-competitive product that sacrificed key functionality. JA-12845 (Noll).

Moreover, Microsoft accomplished the withdrawal of support for the namespace extension APIs by deceiving Novell and other competitors. The District Court concluded there was “sufficient evidence upon which a jury could find that the reasons for the October 3, 1994 decision to withdraw support for the namespace extensions were pretextual.” JA-213.

Eliminating Novell’s products harmed competition in the operating systems market. Microsoft had long recognized the “strategic synergy” between its applications and operating systems businesses and understood that by owning the “key ‘franchises’ built on top of the operating system, we dramatically widen the ‘moat’ that protects the operating systems business.” JA-4293. Word processing was the most important applications category, and Novell’s WordPerfect was the best word processor, with an equivalent number of users to Microsoft Word and far more than any other word processor. By eliminating WordPerfect and Novell’s PerfectOffice applications suite as competitive threats, Microsoft widened the moat protecting its monopoly power in the operating systems market. Further, Novell’s applications offered powerful “middleware” that threatened to reduce the costs of writing applications for Microsoft’s operating system competitors. Microsoft recognized the general threat to its monopoly power in the operating systems

2

market presented by middleware and the specific threat posed by Novell’s middleware.

The District Court disregarded the governing legal standard, which provides that affirmative conduct is anticompetitive when, as here, it harmed rivals, was not “competition on the merits,” and was reasonably capable of contributing significantly to maintaining monopoly power. Multistate Legal Studies, Inc. v. Harcourt Brace Jovanovich Legal & Prof’l Publ’ns, Inc., 63 F.3d 1540, 1550 (10th Cir. 1995). The District Court did not discuss the Tenth Circuit case law recognizing the applicable standard of liability on the grounds that the facts governing the District Court’s ruling were so “case specific.” JA-199 n.5. But that case law establishes that Microsoft’s conduct was anticompetitive because the reasons Microsoft offered for disadvantaging a competitor were pretextual and the conduct disadvantaged competition and was reasonably capable of contributing significantly to maintaining Microsoft’s monopoly power.

Furthermore, in evaluating the harm caused by Microsoft’s conduct in the operating systems market, the District Court improperly required Novell to prove that its applications, by themselves, would have eliminated (rather than diminished) the applications barrier to entry protecting Microsoft’s monopoly power. The availability of a popular competitive word processor alone, however, would have been important for any operating system to compete with Windows.

3

In addition, the District Court disregarded the ample evidence of the threats posed by Novell’s franchise applications and middleware, and departed from the D.C. Circuit’s opinion, the law of the case, and decades of precedent, by placing the burden on Novell to recreate the hypothetical marketplace that would have existed but for Microsoft’s improper conduct.

Further, on each ground for its decision, the District Court disregarded the governing standard of review, which permits ruling in Microsoft’s favor only if no reasonable jury could have found for Novell. The District Court repeatedly overlooked key evidence and resolved conflicting testimony in favor of Microsoft rather than Novell. Because a reasonable jury could have found for Novell on the evidence presented, the District Court’s decision must be reversed and the case remanded for trial.

STATEMENT OF JURISDICTION

The District of Utah had jurisdiction over Novell’s claims under 28 U.S.C. §§ 1331 and 1337. Following the entry of judgment against Novell on July 16, 2012 (JA-230), Novell filed a timely notice of appeal on August 10, 2012 (JA-232-34). This Court has jurisdiction under 28 U.S.C. § 1291.

STATEMENT OF THE ISSUE

Whether Novell presented sufficient evidence at trial from which a reasonable jury could find in Novell’s favor on its claim that Microsoft illegally

4

maintained its monopoly in the operating systems market in violation of the Sherman Act, 15 U.S.C. § 2. See JA-197-231.

STATEMENT OF THE CASE

Count I of Novell’s Complaint (filed November 12, 2004), the subject of this appeal, alleges that Microsoft unlawfully maintained its monopoly power in the operating systems market through anticompetitive conduct that caused injury to Novell’s applications. JA-163-64.

In January 2005, Microsoft moved to dismiss Novell’s Complaint. On June 10, 2005, the District Court denied Microsoft’s motion, and the Fourth Circuit affirmed. Novell, Inc. v. Microsoft Corp. (In re Microsoft Corp. Antitrust Litig.), No. 1:05-cv-01087, 2005 WL 1398643 (D. Md. June 10, 2005), aff’d, 505 F.3d 302 (4th Cir. 2007).

After discovery, the parties submitted summary judgment motions. The District Court entered summary judgment against Novell, reversing its 12(b)(6) ruling by holding that Novell had sold any claims associated with the operating systems market. Novell, Inc. v. Microsoft Corp. (In re Microsoft Corp. Antitrust Litig.), 699 F. Supp. 2d 730, 735 (D. Md. 2010). The Fourth Circuit reversed the District Court on the sale-of-claims issue and remanded, finding that Novell’s Count I claim was “appropriate for trial.” Novell, Inc. v. Microsoft Corp., 429 F. App’x 254, 255, 261-63 (4th Cir. 2011).

5

A nearly two-month trial began on October 17, 2011. At the close of Novell’s case-in-chief, Microsoft moved for judgment as a matter of law. JA-666-741. The District Court declined to grant Microsoft’s motion. JA-13878-82.

The District Court declared a mistrial on December 16, 2011, when the jury could not reach a verdict after only three days of deliberations. JA-16458. Based on its interviews with jurors, the District Court stated: “It appears undisputed that eleven of the twelve jurors would have returned a verdict in favor of Novell on the issue of liability.” JA-198.

In February 2012, Microsoft renewed its Rule 50 motion. JA-1463-1614. On July 16, 2012, the District Court granted Microsoft’s motion. JA-197-230. Novell now appeals.

STATEMENT OF FACTS

I. NOVELL THREATENED MICROSOFT’S MONOPOLY POWER IN
THE OPERATING SYSTEMS MARKET

During the 1990s, Microsoft enjoyed monopoly power in the operating systems market. JA-1625 (FOF ¶¶ 33-35). 2 An operating system (also called a “platform”) is software that performs the core functions of operating a computer.

6

JA-1622 (FOF ¶ 2).

Consumer demand for an operating system is driven primarily by its ability to run applications. JA-1625 (FOF ¶ 37). Applications perform user-oriented tasks (such as word processing) and rely upon the operating system to perform crucial functions. JA-1622 (FOF ¶ 2). “Application Programming Interfaces” (“APIs”) in the operating system enable the application to invoke operating system code that performs a desired function. Id.

To threaten Microsoft’s monopoly power in the operating systems market over the long term, an operating system had to run applications that satisfied user needs. JA-1625 (FOF ¶ 37); JA-12912 (C. Myhrvold). Novell threatened Microsoft’s monopoly power in two mutually reinforcing ways. First, Novell offered the highest quality, most popular competitive application (WordPerfect) in the most valued applications category (word processing) on an array of Microsoft and non-Microsoft platforms. Second, it offered potent “middleware,” a growing category of software that threatened to reduce the industry-wide development costs associated with writing cross-platform applications. These threats existed as the “rise of the Internet” was fueling “the growth of server-based computing, middleware, and open-source software development.” JA-1627 (FOF ¶¶ 59-60); see also JA-4147 (Microsoft CEO Bill Gates stating, “The Internet is a tidal wave. It changes the rules.”). The growth of the Internet represented an “inflection point”

7

in “which categories are redefined and leaders are superseded in the process.” JA-1627 (FOF ¶¶ 59-60). “Working together, these nascent paradigms could oust the PC operating system from its position as the primary platform for applications development.” JA-1627 (FOF ¶ 60).

A. WordPerfect’s Popularity And Cross-Platform Capability
Threatened Microsoft’s Monopoly Power In The Operating
Systems Market

Microsoft recognized the “strategic synergy between the operating system and the software that runs on it.” JA-4293. In its view, if “we own the key ‘franchises’ built on top of the operating system, we dramatically widen the ‘moat’ that protects the operating system business.” Id. These key franchises included “office productivity software.” Id.; see also JA-1976 (Gates stating, “I feel a strong applications business is extremely helpful to our systems strength.”). Thus, “the more market share they can gain in applications, the greater the barrier to entry of new operating systems will be.” JA-12682 (Noll).

In the mid-1990s, the word processor was the most frequently used application. See JA-2820; JA-11946 (Frankenberg); see also JA-11716 (Gibb) (“word processing and spreadsheets were 80, 90 percent of everything that people did”). Moreover, WordPerfect was by far the most significant threat to Microsoft in the word processing market. Its 1994 installed user base was equivalent to Microsoft Word’s (36 percent to Microsoft’s 37 percent), with the next closest

8

competitor at 7 percent. JA-6836. WordPerfect was PC Computing Magazine’s 1994 word processing MVP. JA-12194 (Frankenberg). It was, in Microsoft’s words, “an awesome foe in word processing.” JA-4464.

In 1994, Novell merged with WordPerfect, acquired the Quattro Pro spreadsheet application, and packaged them with other applications to create the product suite called “PerfectOffice.” JA-11928-29, JA-11944 (Frankenberg); JA-5041-42. PerfectOffice was rated as “the best of the high-end office suites.” JA-4124. At that time, 74 percent of users had not yet adopted a suite. JA-5105; JA-9073. PerfectOffice thus had “a very good opportunity” to “capture a significant portion of the new users of office productivity application suites.” JA-11947 (Frankenberg).

WordPerfect also had a history of writing for Microsoft’s operating system competitors. JA-11147 (Harral) (WordPerfect had written to more than a dozen operating systems); JA-5701; JA-2827, JA-2838. WordPerfect would have continued to do so but for Microsoft’s anticompetitive conduct. See JA-11931-33, JA-11969-70 (Frankenberg); JA-11721 (Gibb). Novell prioritized cross-platform development in part “to provide some real competition in the operating system environment.” JA-11930-33 (Frankenberg).

Microsoft perceived the threat posed by Novell. Regarding WordPerfect’s 1994 release for Windows 3.1, Bill Gates said: “I am amazed at their

9

responsiveness. This is very scary and somewhat depressing. This is as much as we plan to do for 1995!!” JA-2964; see also JA-2954 (Gates noting that, after the Novell-WordPerfect merger, “Initiatives to promote anti-Microsoft platforms/API’s/object models become easier to coordinate.”); JA-3973 (Microsoft analysis stating, “The current suite of applications in PerfectOffice are world class and there is a reason for us to follow the progress of this suite very carefully.”).

B. Novell’s Middleware Threatened Microsoft’s Monopoly Power In
The Operating Systems Market

In the 1990s, due to the expense of writing applications for multiple operating systems, ISVs often wrote applications exclusively for Windows to reach the greatest volume of users and offset development costs. JA-1626 (FOF ¶¶ 30, 38). This fact, combined with the importance of applications to a user’s decision to purchase an operating system, is known as the “applications barrier to entry.” JA-1624-26 (FOF ¶¶ 31, 36-39).

To lower the expense, some software developers created “middleware” – i.e., software programs that run on top of operating systems and that make their own APIs available to developers. JA-1622 (FOF ¶ 28); see also JA-11950-51 (Frankenberg); JA-11136-38, JA-11147-50, JA-11164-65 (Harral); JA-12342-44 (Alepin). Middleware promised to lower developer costs by enabling application developers to write to one set of middleware APIs rather than multiple sets of operating system APIs. See JA-1627-28 (FOF ¶¶ 68-69); JA-11136-38, JA-11147-

10

50 (Harral); JA-12336-38 (Alepin); JA-12676-78 (Noll).

“To the extent that developers begin writing attractive applications that rely solely on servers or middleware instead of PC operating systems, the applications barrier to entry could erode.” JA-1624-25 (FOF ¶ 32). The “growth of middleware-based applications could lower the costs to users of choosing a non-Intel-compatible PC operating system like the Mac OS.” JA-1623-24 (FOF ¶ 29). The “more popular middleware became and the more APIs it exposed, the more the positive feedback loop that sustains the applications barrier to entry would dissipate. Microsoft was concerned with middleware as a category of software; each type of middleware contributed to the threat posed by the entire category.” JA-1627-28 (FOF ¶ 68).

PerfectOffice included middleware called AppWare. JA-12344, JA-12346- 48 (Alepin); JA-11167 (Harral). AppWare allowed developers to build stand- alone, cross-platform applications that could also integrate with PerfectOffice applications. JA-5090; JA-5117; JA-12346-47 (Alepin). As Microsoft recognized, “AppWare contains all of the functions of an operating system,” JA-11867 (Silverberg), and was one of its “most serious competitors,” JA-645 (Maritz). See JA-2659 (AppWare was “direct competition to Windows”). AppWare threatened to “reduce Windows or anything underneath it to a commodity,” JA-11867 (Silverberg), and “develop a layer that will provide all of the services required by

11

applications,” JA-645 (Maritz). See JA-2492 (Microsoft analysis stating, AppWare “might be [sic] first viable platform for commercial cross-platform development.”).

Novell additionally engineered its office productivity applications to utilize a single body of “shared code” called “PerfectFit.” JA-11137-38, JA-11140-41, JA-11147-50 (Harral). PerfectFit was middleware. JA-11164-65 (Harral); JA-12343-44 (Alepin). PerfectFit provided Novell’s office productivity applications with critical functions that those applications would have otherwise called upon Windows to perform, including WordPerfect’s user interface, menus, icons, and toolbars. JA-5047; JA-11157-58 (Harral); JA-11952 (Frankenberg).

Novell also enabled third-party ISVs to use its PerfectFit middleware to develop applications that could be ported to any operating system where PerfectOffice was installed. JA-11148-50 (Harral). More than 1,000 of these ISVs became “PerfectFit Partners,” who received dedicated Novell technical support. JA-11156-58 (Harral); JA-11718, JA-11821-22 (Gibb); JA-3437. PerfectFit made available more APIs than Netscape Navigator and Sun’s Java combined. Compare JA-1629-30 (FOF ¶ 77) (stating that Navigator’s and Java’s combined APIs “totaled less than a thousand”) with JA-4445-47 (indicating PerfectFit Shared Code 2.3 had 1555 APIs with more to be added in PerfectFit 95).

12

II. MICROSOFT INDUCED NOVELL TO RELY ON THE
NAMESPACE EXTENSION APIS IN DESIGNING ITS
WINDOWS 95 APPLICATIONS

Microsoft recognized it was important, both to ISVs and Microsoft, that ISVs ship their products within 90 days of the release of Windows 95. JA-5642 (Microsoft email stating that the message to ISVs should be “Windows 95 is going to be huge [sic] if they aren’t shipping within 90 days they will be at a big disadvantage”); see also JA-14202-04, JA-14235 (Struss); JA-14433 (Muglia); JA-2929. Microsoft thus launched the First Wave Program “to get firm commitments from the most important applications’ vendors” to “ship their application within 90 days of the shipment of Chicago” – the codename for Windows 95. JA-2929-30. Microsoft offered to provide documentation, development assistance, and access to early “beta” versions of Chicago. JA-14204-07 (Struss). WordPerfect joined the First Wave Program as one of the “Tier A” or “key” applications that “represent most application software sales.” JA-2858, JA-2935. Microsoft invited WordPerfect to join because it “felt it was critical to have their support for Windows 95 and for Windows 95 to be successful.” JA-14205 (Struss).

A. The Namespace Extension APIs Enabled Novell’s Applications To
Access The Windows 95 Namespaces And Integrate With The
Windows Shell

As part of its promotional efforts, Microsoft counseled WordPerfect and

13

other ISVs regarding what operating system features “a good Chicago app” should seek to exploit. JA-2666; see also JA-2698 (“We’re providing new controls you can (and should) use” and “We’re making it possible for you to extend the shell”). One important feature was Chicago’s new “shell” – i.e., the user interface. JA-14700 (Nakajima); see, e.g., JA-2666-67; JA-2696-719. Windows 95’s shell included the Windows Explorer – a new application that enabled users to browse the computer’s “system resources,” including but not limited to files, applications, hardware, and network drives. See JA-14702-04 (Nakajima). Through the Explorer, users could move files and folders, launch applications to view and edit documents, and access external hardware and network drives. JA-12937 (Ludwig); JA-12351-52 (Alepin); JA-14702-04 (Nakajima).

Windows 95 made various system resources accessible through new “namespaces,” including:

  • The Desktop, which contained all the Windows 95 namespaces and resources the user or ISVs placed on the default Windows 95 screen, JA-11568, JA-11587 (Richardson); JA-11187 (Harral);
  • My Computer, which contained disk drives, JA-11240, JA-11442 (Harral);
  • My Network Neighborhood, which linked to “network places” and drives, JA-11192 (Harral);
  • My Briefcase, which enabled users to store regularly used documents, JA-11193 (Harral);

14

  • The Recycle Bin, which contained deleted files, id.; and
  • The Control Panel, which provided control over printers and sound and graphics cards, JA-13715 (Gates).

The Windows Explorer showed these namespaces in a hierarchical “tree” view in the left hand pane of the Explorer window. JA-11193-94 (Harral). By clicking on a namespace in the left pane, a user could display its contents in the right pane. JA-13713-14 (Gates). As Bill Gates stated, “the hierarchical view (scope pane) view [sic] is critical. The ability to see the real namespace of the system where we are putting everything exists only there – the ability to move things around easily only come [sic] from there. The tree view is central to our whole strategy.” JA-2866.

To satisfy basic expectations of Windows 95 users, WordPerfect’s file open dialog needed the capacity to browse the namespace hierarchy in the same fashion as the Windows Explorer. See JA-11562 (Richardson) (discussing the need to access the namespaces through the file open dialog and explaining Novell “wouldn’t have a Windows 95 product if we didn’t have access to those”). A “file open dialog” is the dialog window that an application or operating system presents to a user when the user selects the “Open” function from a File Menu. JA-11140 (Harral). Microsoft knew that if an ISV “was creating a custom file open,” it was “one of the keys that they recreate the full name space.” JA-15292 (Belfiore). At

15

the same time, WordPerfect needed to maintain key features of past versions of the file open dialog that users had come to expect. JA-11204-05, JA-11277-78 (Harral); see infra pp. 23-25, 54-56.

The namespace extension APIs were a set of Windows 95 APIs that helped applications exploit the namespace technology. See JA-11240-43; JA-11265-67 (Harral). These APIs enabled applications’ file open dialogs to browse the namespaces. See JA-11264-67 (Harral). The namespace extension APIs also enabled Novell to extend the Windows 95 shell by creating and adding custom namespaces that would enable users to make effective use of various functions in Novell’s software within the Windows 95 system. See JA-11220-23 (Harral); JA-11521-25, JA-11545-46 (Richardson). These functions included Novell’s document management system, its QuickFinder search engine, its Internet browser, its email application, and its ClipArt library. JA-11199-200 (Harral); JA-11545-46 (Richardson). The inventor of the namespace extension APIs acknowledged that Novell’s intended custom namespaces were the type of functionality for which those APIs were intended. See JA-14801-02, JA-14806-07, JA-14822-23 (Nakajima).

B. Microsoft Induced Novell To Rely On The Namespace Extension
APIs In Designing Its Applications

Between November 1993 and October 1994, Microsoft actively promoted the Windows 95 shell and the namespace extension APIs to Novell. See

16

JA-11522-26, JA-11674-75 (Richardson) (based on Microsoft’s promotions, Novell “felt we had a green light to create our own namespace browser as well as creating namespace extensions”); JA-5751 (recommending that custom file open dialogs support a “namespace hierarchy that’s the same as the shell”), JA-5753-54; see also JA-15288-89 (Belfiore); JA-2666-67; JA-2696-719; JA-2855; JA-4741; JA-7558-76.

During a November 1993 meeting, in which “half of the conversation concerned these NameSpace extension APIs,” JA-11220 (Harral), Microsoft informed WordPerfect of its decision to document the shell extensions. JA-2666. WordPerfect “talked at length with” Microsoft about its plans to use the namespace extension APIs. JA-11216 (Harral). From that meeting, Microsoft knew that WordPerfect was “very happy about us deciding to document the shell extentions [sic].” JA-2666 (stating that WordPerfect “just aquired [sic] a document management system” and “they will want to plug that in” the shell), JA-2667 (the “shell needs to allow extending the find command” – i.e., the QuickFinder).

Microsoft continued to promote the namespace extension APIs after the November 1993 meeting. See, e.g., JA-2715, JA-2718; JA-2855. In June 1994, Microsoft partially documented – i.e., provided written guidance for development with and use of – those APIs in Chicago’s M6 beta. JA-11234-35 (Harral); JA-3385-404. When asked by Novell, Microsoft stated it would release the full

17

documentation for the APIs in the next beta release. JA-11249 (Harral). By October 1994, Novell’s shared code team had finished 80 percent of the coding to the namespace extension APIs. JA-11253 (Harral). Had Microsoft published the full documentation, as it represented it would to Novell, the shared code team would have completed its work by December 1994. JA-11253-54 (Harral).

III. MICROSOFT WITHDREW SUPPORT FOR THE NAMESPACE
EXTENSION APIS TO DISADVANTAGE COMPETITORS

To Microsoft, affording access to the namespace extension APIs meant “Word and Excel are forced to battle against their competitors on even turf.” JA-2335. On September 20, 1994, Bill Gates attended a software industry conference where he observed Novell demonstrate a prototype of its own cross- platform extensible shell. JA-3651-54; JA-3696; JA-11955-61 (Frankenberg). In an email about the demonstration, Gates discussed the threat posed by Novell: “Novell is a lot more aware of how the world is changing than I thought they were.” JA-3696. For him, Novell’s presentation emphasized “the importance of our shell integration.” Id.

On October 3, 1994, just two weeks after Novell’s demonstration, Gates announced a “bombshell” decision that even Microsoft’s own employees “did not expect,” JA-3694: Microsoft was withdrawing support for the namespace extension APIs, JA-1967. In an email to Microsoft’s executives, Gates explained,

18

I have decided that we should not publish these extensions. We should wait until we have a way to do a high level of integration that will be harder for likes of Notes, Wordperfect to achieve, and which will give Office a real advantage.
Id. Gates acknowledged it was “very late in the day to” be “making changes to Chicago” and the APIs were “a very nice piece of work.” Id. He further acknowledged that Microsoft could not “compete with Lotus and Wordperfect/Novell without” having “the Office team really think through the information intensive scenarios” required by “shell integration work.” Id.

Pursuant to Gates’ edict, Microsoft de-documented the namespace extension APIs by marking them “;Internal,” thereby “hiding” the interfaces, which were not documented at all in the subsequent M7 beta release. JA-3701; JA-12506-07 (Alepin); see also JA-3708-27. Subsequently, Microsoft refused to provide the remaining documentation for the namespace extension APIs that would have allowed Novell to complete its development or to provide any technical assistance for continued use of the APIs. JA-3767-68; JA-3772; JA-11276-77, JA-11296-97 (Harral).

IV. MICROSOFT DECEIVED NOVELL AND OTHER ISVS
REGARDING THE TRUE REASONS FOR DE-DOCUMENTATION

Microsoft subsequently deceived the ISV community regarding the true reason for the decision to de-document and withdraw support for the namespace extension APIs. Brad Silverberg, the executive in charge of Windows 95

19

development, warned in recommending against the decision that Microsoft would face a “firestorm of protest” from ISVs using the APIs – a group he knew to include “WordPerfect, Lotus, Symantec, and Oracle.” JA-3691. Protest by these ISVs would “play out” on “page one of the weeklies” and “lead to calls for the DOJ to investigate.” Id.

Microsoft additionally needed to conceal that certain Microsoft applications continued to rely upon the namespace extension APIs. Marvel, Microsoft’s online services application, see generally JA-5835-57, was so dependent on the APIs that there was “no way” it could “move off the current interfaces and still have” a “chance of shipping with Win ’95.” JA-5755. Thus, Marvel continued to use them. JA-14808 (Nakajima); JA-12959-60 (Siegelman).

Based on past experience, Microsoft knew that ISVs would find Microsoft’s exclusive use of the namespace extension APIs unacceptable. See JA-2353 (ISVs “*really* want extensibility.” “What’s more, they were afraid and angry that Microsoft would use the hooks for its own purposes (apps, mail, etc) but not provide to isv’s. This was a very hot button.”); JA-2453 (WordPerfect complained to Microsoft that it “was an unacceptable situation” for Microsoft to “tie into the shell” if Microsoft would not also “allow ISVs to extend the explorer”); JA-2478 (referring to similar Lotus complaints to Bill Gates). Such conduct would violate a longstanding Microsoft rule against allowing its applications to exploit Windows

20

features that remained undisclosed to ISVs. JA-14768-69 (Nakajima).

Microsoft therefore devised a script employed by its ISV liaisons to deceive ISVs regarding the reasons for its decision to de-document the namespace extension APIs. See JA-3702-06. Echoing a 1993 plan to give false excuses for withholding shell extensibility (i.e., “we couldn’t get it done in time”), JA-2238, Microsoft offered “Compatibility,” “System Robustness,” and “Ship Schedule” as excuses for the de-documentation, JA-3705, even though Gates did not mention these concerns in explaining his decision,3 see JA-1967.

Moreover, even though Marvel continued to use the APIs, Microsoft falsely told ISVs, “All applications within Microsoft which were originally implementing these interfaces have been required to stop.” JA-3703. Microsoft instructed its employees: “PLEASE DO NOT MENTION MARVEL IN ANY OF YOUR CONVERSATIONS” with ISVs. JA-3702. Microsoft also instructed its team to lie if asked directly about Marvel. JA-3705.

In August 1995, Microsoft employee Scott Henson realized that Athena, another Microsoft application under development, was “using the namespace extensions” – in other words, doing “the EXACT thing” that Microsoft “told ISVs

21

they could (and should) not do.” JA-4203. Henson could not “even express how BAD this is!” Id.

V. MICROSOFT’S ANTICOMPETITIVE CONDUCT DELAYED
NOVELL’S PRODUCT RELEASES AND ELIMINATED THE
COMPETITIVE THREAT THEY POSED TO MICROSOFT’S
MONOPOLY POWER

Novell always planned to release a marketable product within the critical window of opportunity that would close 90 days after the release of Windows 95. See JA-8853-54 (“It is critical that WordPerfect Corp. have a version of WPWin that is coded for 32-bit Chicago release within no more than a [sic] two to three months of Chicago’s ship date.”); JA-2771 (same); JA-11213 (Harral). As described below, the de-documentation of the namespace extension APIs prevented Novell from timely constructing a file open dialog that both retained the functionality of past releases and could browse the namespaces.

For many years, WordPerfect had been an industry leader in providing a robust custom file open dialog for its users. JA-11732-33 (Gibb). Users frequently “lived in” the WordPerfect file open dialog because it enabled them to perform important tasks like renaming files, previewing files, and finding content using the QuickFinder search engine, among other things. JA-11732-35 (Gibb); JA-11204-05 (Harral). The file open dialog was a “core differentiator” for the WordPerfect product line. JA-11750 (Gibb).

The de-documentation of the namespace extension APIs left three options

22

for accessing the critical Windows 95 namespaces. JA-11273-75 (Harral). The first was to try to use the incomplete June 1994 documentation to call upon the now de-documented APIs. Id. The second was using the Windows 95 common file open dialog. Id. The third was to attempt to recreate the operating system-level functionality provided by the now de-documented APIs in Novell’s custom file open dialog. Id. None of these options would have allowed Novell to release a competitive product within the critical 90-day window.

The first option proved unworkable because Microsoft refused to answer any questions about the de-documented APIs, see, e.g., JA-3767-68; JA-3772; JA-11276-77, JA-11296-97 (Harral), and Novell and other ISVs reasonably relied on Microsoft’s misrepresentations that the APIs could stop working “in future releases of Windows 95 (or even between interim beta builds)” and that ISVs would “be completely at their own risk” if they attempted to use those APIs. JA-3706; JA-11535-37 (Richardson); see also JA-5755; JA-4203. Novell concluded it was futile to continue using the unsupported APIs. JA-11276-77 (Harral); JA-11535 (Richardson); see also JA-16000 (Bennett) (Microsoft’s expert agreeing that “it’s a bad practice to call on undocumented APIs”).

Novell thus had “two different ways to commit suicide”: use Microsoft’s common file open dialog or recreate the lost functionality in Novell’s custom file open dialog. JA-12845 (Noll). The second option – using Microsoft’s common

23

file open dialog – would have required a disastrous sacrifice of functionality:
the common dialog wouldn’t even give us the level of functionality we had in our last release in Windows or that we had on our DOS card. It was a huge step backwards for us. And we felt it simply wasn’t an option. If we were to go with that option we didn’t really have a product.
JA-11563 (Richardson) (emphasis added); see also JA-11782-83 (Gibb). The file open dialog was a “core differentiator” for WordPerfect. JA-11750 (Gibb); see also JA-11178-82 (Harral). Ultimately, choosing the common dialog “would have been a choice to have disenfranchised our customer base.” JA-11436, JA-11204-05, JA-11277 (Harral).

Microsoft acknowledged the importance of the file open dialog function, as well as the high quality of WordPerfect’s custom file open dialog, both generally and when compared with Windows 95’s common file open dialog and Microsoft Office’s custom file open dialog. See JA-3046 (stating that “File Open is the most frequently used dialog in all of Office” and that “every minor complexity or limitation in the dialog is greatly magnified by its frequent usage”). One 1994 analysis acknowledged that “WordPerfect has been winning rave reviews from the press for its new File Open dialog” and that WordPerfect’s dialog addressed “most of the problems” with Microsoft Office’s file open dialog that the analysis identified. JA-3046-47. The same analysis listed functions in WordPerfect’s dialog that were not present in Windows 95’s common dialog. JA-3059.

24

Novell’s customers’ expectations mandated that it continue to provide the longstanding services it had historically provided to avoid alienating its considerable installed base of users. JA-11204-05, JA-11277, JA-11436 (Harral). Novell therefore determined that using Microsoft’s common file open dialog was not a viable alternative. JA-11177, JA-11436 (Harral); JA-11562-63 (Richardson).

Novell thus tried to timely replicate the operating system level of functionality provided by the de-documented APIs inside WordPerfect’s custom file open dialog – a process that took a full year. JA-11273-74, JA-11278, JA-11281-83 (Harral). Novell developers worked around the clock, JA-11285 (Harral), but by July 1995 it was clear: “There is no conceivable way to have the NSB Code” – i.e., the code for the “Name Space Browser (aka FileOpen Dialog)” – “complete by August 22.” JA-4200. Novell confronted the same delays that Microsoft knew Marvel would face if, like PerfectOffice, Marvel could not use the namespace extension APIs. JA-3694-95; JA-5755; see also infra p. 40.

Ultimately, Microsoft’s de-documentation of the namespace extension APIs caused PerfectOffice and WordPerfect not to ship until well beyond the critical 90-day window after the August 1995 release of Windows 95, and caused Novell to miss the market. JA-11541 (Richardson); JA-11738-39 (Gibb); JA-11427 (Harral).

In 1994, 74 percent of users had not yet adopted a suite. JA-5105. By 1997,

25

however, Microsoft captured 90 percent of the market for office productivity software, JA-4293, thereby widening the moat protecting its monopoly power in the operating systems market, JA-4293.

SUMMARY OF THE ARGUMENT

The District Court’s decision to forego the established legal framework for unlawful monopolization cases resulted in a decision that cannot be reconciled with the law of this Circuit. Under Tenth Circuit precedent, anticompetitive conduct is that which impairs the opportunities of rivals and is not competition on the merits, “if the conduct appears reasonably capable of contributing significantly to creating or maintaining monopoly power.” Multistate, 63 F.3d at 1550 (citation and internal quotation marks omitted). A jury could reasonably conclude from the record below that Microsoft’s affirmative conduct impaired the opportunities of actual and potential rivals and that Microsoft engaged in that conduct solely to disadvantage competitors, not to design a better product or otherwise compete on the merits. Indeed, the District Court’s finding that a reasonable jury could have found Microsoft’s proffered justifications to be pretextual provided sufficient grounds to conclude that its conduct was not “competition on the merits.”

Additionally, the District Court improperly rejected deception as anticompetitive conduct by resolving an evidentiary conflict that properly was for the jury regarding whether Microsoft was aware of Novell’s reliance on the APIs.

26

The District Court also wrongly held that Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985), required Novell to show that Microsoft terminated its entire relationship with Novell.

Further, the District Court wrongly concluded that eliminating Novell as a competitive threat did not cause harm to competition in the operating systems market. Novell presented ample evidence that established the legitimacy of the threat to Microsoft’s monopoly power from Novell’s applications and middleware. The District Court erred by disregarding this evidence and adopting a test that effectively would allow monopolists to crush nascent threats to their monopoly power seriatim. This test departed markedly from the legal standards applied by this Circuit, the D.C. Circuit in the Government’s case against Microsoft, and even by the District Court and Fourth Circuit on summary judgment.

Finally, the District Court wrongly concluded that there was no “underlying business reality,” JA-226-27, to the claim that Microsoft caused fatal delay in the release of Novell’s product. Here, the District Court substituted its judgment for that of the factfinder, including disregarding the ample evidence that Microsoft left Novell with no viable options when Microsoft de-documented the namespace extension APIs.

27

ARGUMENT

I. STANDARD OF REVIEW

This Court reviews the District Court’s grant of judgment as a matter of law de novo. Strickland v. United Parcel Serv., Inc., 555 F.3d 1224, 1228 (10th Cir. 2009). Under this standard, the Court sits in the same position as the trial court and must apply the correct legal standards as if deciding the question in the first instance. Id.; Phillips v. Hillcrest Med. Ctr., 244 F.3d 790, 796 (10th Cir. 2001).

“Judgment as a matter of law is improper unless the evidence so overwhelmingly favors the moving party as to permit no other rational conclusion.” Shaw v. AAA Eng’g & Drafting, Inc., 213 F.3d 519, 529 (10th Cir. 2000). The Court must be “certain the evidence ‘conclusively favors one party such that reasonable men could not arrive at a contrary verdict.’” Weese v. Schukman, 98 F.3d 542, 547 (10th Cir. 1996) (citation omitted). Such evidence must point “but one way” and be “‘susceptible to no reasonable inferences which may support the opposing party’s position.’” Strickland, 555 F.3d at 1228 (citation omitted).

The Court must review the entire evidentiary record, draw all inferences in Novell’s favor, and refrain from making credibility determinations or weighing the evidence. Guides, Ltd. v. Yarmouth Grp. Prop. Mgmt., Inc., 295 F.3d 1065, 1073 (10th Cir. 2002). When faced with conflicting evidence on a particular issue, the

28

Court must resolve all conflicts in Novell’s favor and disregard all evidence favorable to Microsoft except evidence that “is uncontradicted and unimpeached” and “comes from disinterested witnesses.” Reeves v. Sanderson Plumbing Prods., Inc., 530 U.S. 133, 151 (2000) (citation and internal quotation marks omitted).

II. MICROSOFT’S CONDUCT WAS ANTICOMPETITIVE

The law governing monopolization cases is well established. To prove a violation of the Sherman Act, 15 U.S.C. § 2, Novell must show that Microsoft (1) possessed monopoly power in the operating systems market and (2) willfully maintained that power through anticompetitive conduct rather than through development of a superior product, business acumen, or historic accident. Verizon Commc’ns Inc. v. Law Office of Curtis V. Trinko, LLP, 540 U.S. 398, 407 (2004); United States v. Grinnell Corp., 383 U.S. 563, 570-71 (1966). It is undisputed that the first element has been satisfied. JA-1625 (FOF ¶¶ 33-34); see also Microsoft, 253 F.3d at 51.

To satisfy the second element, Novell needed only to adduce sufficient evidence from which a reasonable jury could find that Microsoft maintained its monopoly power in the operating systems market “by anticompetitive or exclusionary means or for anticompetitive or exclusionary purposes.” Aspen Skiing, 472 U.S. at 595-96. A monopolist’s actions are anticompetitive “if they impair opportunities of rivals and are not competition on the merits or are more

29

restrictive than reasonably necessary for such competition,” and “if the conduct appears reasonably capable of contributing significantly to creating or maintaining monopoly power.” Multistate, 63 F.3d at 1550 (citation and internal quotation marks omitted); Instructional Sys. Dev. Corp. v. Aetna Cas. & Sur. Co., 817 F.2d 639, 649 (10th Cir. 1987) (citing 3 P. Areeda & D. Turner, Antitrust Law ¶¶ 625b, 626c, g (1978)); see also LePage’s Inc. v. 3M, 324 F.3d 141, 162 (3d Cir. 2003) (“As the Supreme Court recognized” in Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699 (1962), “courts must look to the monopolist’s conduct taken as a whole rather than considering each aspect in isolation.”).

The District Court failed to apply this standard, under which Microsoft’s conduct was anticompetitive for three independently sufficient reasons. See Argument Parts II.A.1-A.3. Furthermore, while purporting not to resolve the parties’ dispute over whether Novell must show that the anticompetitive conduct “contributed significantly” to maintaining Microsoft’s monopoly power, or was reasonably capable of doing so, JA-225 n.22, the District Court adopted a causation standard that effectively was more stringent than either party’s proposed test. See infra pp. 43-45. Should the Court choose to resolve this dispute, the reasonably capable standard is clearly the correct one. The Tenth Circuit adopted this standard in Multistate, 63 F.3d at 1550, 1553, as did the D.C. Circuit in Microsoft, 253 F.3d at 79. Moreover, whereas Microsoft has never cited a single

30

case in support of its position, Novell has cited numerous cases that use the same or similar standard, including three from courts in this Circuit.4

A. Microsoft’s Conduct Was Not Competition On The Merits And
Eliminated Novell As A Competitor

1. Microsoft Induced Novell’s Reliance On The Namespace
Extension APIs And Thereby Made Novell And
Competition Worse Off Than If Microsoft Had Refused To
Make Them Available From The Outset

Instead of asking whether Microsoft’s conduct was not competition on the merits and impaired the opportunities of its rivals as Multistate requires, the District Court erroneously concluded that Novell either had to prove that Microsoft’s conduct was deceptive or meet a total termination of relationship requirement that it purports to locate in Aspen Skiing. JA-216. The District Court erred in its analysis of these issues (see Argument Part II.A.2-3), but also erroneously overlooked the distinction between a mere refusal to deal akin to the defendant’s conduct in Aspen Skiing and affirmative conduct that imposes costs on a competitor in a manner that harms competition and is not competition on the merits.

This Circuit has recognized the importance of this distinction. In Christy

31

Sports LLC v. Deer Valley Resort Co., 555 F.3d 1188, 1190-91, 1194-98 (10th Cir. 2009), the Court found that a ski resort had no obligation to continue to allow the operation of an independent ski rental business on property it had sold with a restrictive covenant that allowed it to control the use of the property for that purpose. At the same time, the Court recognized the possibility that the change in prior practice “could give rise to an antitrust claim” if “by first inviting an investment and then disallowing the use of the investment the resort imposed costs on a competitor that had the effect of injuring competition in a relevant market.” Id. at 1196; see also Multistate, 63 F.3d at 1553 n.12 (distinguishing between affirmative conduct and a refusal to deal governed by Aspen Skiing).

That is what happened here. Had Microsoft not induced Novell’s reliance, Novell would have had two years from when Microsoft initially stated that it intended to document the namespace extension APIs in November 1993 (see supra pp. 16-18) to produce a competitive product within the critical 90-day window. Instead, as a result of Microsoft’s conduct, Novell traveled for a full year on a path of detrimental reliance under the ultimately thwarted assumption that it would have ready access to the namespace extension APIs and that it would therefore only require a “small amount of work” to complete the file open dialog once Novell received the final documentation. JA-11367; see supra pp. 16-18, 22-25. By encouraging Novell to develop software that depended upon Microsoft’s

32

documentation and support for the namespace extension APIs, and then withdrawing support for them nearly one year later, Microsoft “imposed costs on a competitor” in the form of fatal delay in the release of Novell’s product that destroyed Novell as an effective competitor. Christy Sports, 555 F.3d at 1196.

The District Court did not address the language from Christy Sports or Multistate despite Novell’s prior reliance on it. JA-1796. Microsoft has likened the restrictive covenant in Christy Sports with language in the Windows 95 beta agreement, and a purported industry practice that beta products can change. JA-1939 n.40. The two situations are materially different. The Christy Sports plaintiff “knew from the beginning that it could operate a ski rental business only by permission of” the defendant “on a year-to-year basis.” 555 F.3d at 1196. Accordingly, there was not even an arguable invitation to investment, let alone a combined invitation and disallowance that harmed competition.

By contrast, as one Microsoft executive testified, “the purpose of documenting an API is in effect to put a stake in the ground and say this is something that you as an application developer can count on being available to you as an operating system service today and in the future.” JA-604 (Raikes); see also JA-11730-31 (Gibb) (explaining that Microsoft released betas to give ISVs time to write compatible applications); JA-12254-55 (Henrich); JA-604-05 (Raikes); JA-14254-55 (Struss) (various Microsoft employees offering similar descriptions

33

of betas generally). Microsoft thus specifically intended that ISVs rely on the documentation to develop products for Windows 95, and also induced ISV reliance months prior to providing the preliminary documentation on the expectation that Microsoft would document the namespace extension APIs in the June 1994 beta. See supra pp. 16-18. Moreover, while there was conflicting testimony regarding the industry practice governing betas (that the District Court was required to resolve in Novell’s favor), there was no evidence that the industry practice sanctioned the changing of betas to disadvantage competitors. Further, the test for whether conduct is anticompetitive is not whether it is contractually permissible but instead whether it causes anticompetitive harm and lacks a competitive justification. See Poller v. Columbia Broad. Sys., Inc., 368 U.S. 464, 468-69 (1962) (exercise of individual and legal contract rights violates the Sherman Act when “conceived in a purpose to restrain trade, control a market, or monopolize”); ES Dev., Inc. v. RWM Enters., Inc., 939 F.2d 547, 555 (8th Cir. 1991) (same).

Further, in Christy Sports, there was no argument that competition was harmed in any relevant market. Far from it, the plaintiff failed even to identify a relevant market. 555 F.3d at 1193-94. By contrast, the de-documentation of the namespace extension APIs lacked a legitimate competitive justification, and the combined effect of Microsoft’s invitation and disallowance – i.e., inducing Novell’s reliance on these APIs and subsequently withdrawing support for them –

34

destroyed WordPerfect and PerfectOffice as a competitive force and harmed competition in the operating systems market. See supra pp. 22-26, infra pp. 45-54.

This distinction between affirmative inducement of reliance that harms competition and a refusal to deal also refutes the District Court’s equation of the intentionality of “a decision not to publish the namespace extension APIs in the first place” with a “decision to withdraw support for the namespace extension APIs after they have been published.” JA-215. The latter involved the intentional inducement of reliance, an intentional withdrawal of support for no legitimate reason, and anticompetitive harm that would not have existed had Microsoft refrained from documenting the namespace extension APIs at the outset.

2. Microsoft’s Conduct Was Deceptive

When Microsoft de-documented the namespace extension APIs, it embarked on a campaign of deception that specifically instructed Microsoft employees to mislead Novell and other ISVs regarding the true reasons for the de-documentation and to conceal that certain Microsoft applications continued to use the APIs. See supra pp. 19-22. The District Court correctly held that “a jury could find that the reasons for the October 3, 1994 decision to withdraw support for the namespace extensions were pretextual.” JA-213.

The District Court recognized that deception of a competitor may give rise to an antitrust claim when “the purpose of the deception is to mislead a competitor

35

into taking action (or not taking action) that would substantially change the competitive environment.” JA-216 n.16. “‘Cases that recognize deception as exclusionary hinge’” on “‘whether the conduct impaired rivals in a manner tending to bring about or protect a defendant’s monopoly power.’” Id. (quoting Rambus, Inc. v. FTC, 522 F.3d 456, 464 (D.C. Cir. 2008)). Here, Microsoft’s deception enabled it to de-document the APIs while concealing that its own applications continued to use them. See supra pp. 19-22.

Nonetheless, the District Court found that Microsoft’s deception was not anticompetitive because, it concluded, there was “no evidence” that Microsoft “knew that Novell was using those APIs in the development of its applications and that, by withdrawing support for those APIs, Microsoft knew that Novell would fall behind schedule.” JA-216.

The District Court offered no basis for concluding that knowledge of Novell’s specific plans was required, given that Microsoft needed to deceive the entire ISV community. Moreover, in reaching its conclusion, the District Court impermissibly substituted its judgment for the jury’s. Substantial evidence demonstrated that Microsoft knew Novell was relying upon the APIs and would be delayed or forced to release a non-competitive product if the APIs were de- documented. For example:

  • Two days after Gates’ decision to de-document the namespace extension APIs, Microsoft executive Brad Silverberg informed Gates that, “Other

36

    ISV’s using the extensions are WordPerfect, Lotus, Symantec, and Oracle.” JA-3691. The District Court disregarded this language, which alone permitted a reasonable jury to conclude that Microsoft knew Novell was using the namespace extension APIs.
  • Moreover, at the time of Gates’ decision, Microsoft had known for nearly a year that Novell intended to use the namespace extension APIs. JA-2666-67. WordPerfect’s shared code team also contacted Microsoft’s Premier Support on multiple occasions for assistance concerning the namespace extension APIs before Gates’ decision. JA-11262 (Harral).
  • After the de-documentation, Novell alerted Microsoft on numerous occasions that it required access to the namespace extension APIs, and complained repeatedly about the de-documentation as well as the issue of undocumented APIs generally. See JA-11285-86, JA-11302-03 (Harral); JA-3767-68; JA-3772; JA-11965-66, JA-12181 (Frankenberg).
  • Finally, the District Court relied on an email containing Microsoft’s survey of ISVs taken before Gates’ decision, but ignored Microsoft’s assessment in that email that it was “very likely” Novell had started work with the namespace extension APIs and the response from Novell’s Tom Creighton “that there would ‘be hell to pay in the press’ if” Microsoft “changed the interfaces.” JA-3658. Microsoft also acknowledged in that email that Novell would not want to “tip their hand” regarding their development strategy. Id. The District Court had no basis for disregarding this evidence.

Against this evidence, the District Court relied on industry practice concerning betas as evidence that Microsoft would have lacked knowledge of Novell’s reliance. This inference was untenable given that, among other things, the entire purpose of documenting an API specifically, and betas generally, is to induce reliance, and given the evidence that Microsoft knew of Novell’s reliance. See supra pp. 33-34. At most, it creates an issue of fact for the jury.

Similarly, the District Court’s reliance on evidence from Microsoft’s

37

Brad Struss provided no reason for substituting the Court’s judgment for that of a jury and disregarding the above evidence.
  • Struss testified he did not know whether WordPerfect was part of a group of ISVs who were actively developing with the namespace extension APIs in Fall 1994. JA-14298 (Struss); JA-4202. Based on this evidence alone, a reasonable jury could discount Struss’s recollection of what he understood concerning Novell’s reliance on the APIs.
  • Struss’s email reporting that WordPerfect appeared “OK” with the “namespace extension API changes” was based on a call to WordPerfect in which Microsoft representatives had relied on Microsoft’s deceptive script.5 JA-6926-30. Any lack of protest from WordPerfect could only have reflected its acceptance of Microsoft’s pretextual explanation, not a lack of prior reliance.

In sum, the District Court’s failure to apply the appropriate standard of review in evaluating this evidence is an independent ground for reversal.

3. Microsoft Altered A Voluntary Course Of
Dealing Without
A Legitimate Competitive Reason In Violation Of Aspen Skiing

Under Aspen Skiing, which addresses the circumstances under which a monopolist may be held liable for unilaterally refusing to deal with a competitor, if “a firm has been attempting to exclude rivals on some basis other than efficiency, it is fair to characterize its behavior as predatory.” Aspen Skiing, 472 U.S. at 605

38

(citation and internal quotation marks omitted). The significance of a defendant’s “prior conduct” in analyzing refusals to deal is that it sheds “light upon the motivation of its refusal to deal.” Trinko, 540 U.S. at 409. Thus, in Aspen Skiing, the defendant’s termination of a joint ski ticket with the plaintiff’s smaller mountain was anticompetitive because it “suggested a willingness to forsake short- term profits to achieve an anticompetitive end.” Id. In contrast, in Trinko, the plaintiff did “not allege that Verizon voluntarily engaged in a course of dealing with its rivals,” and so the defendant’s prior conduct shed “no light upon the motivation of its refusal to deal.” Id.

Here, the District Court correctly found that “the text of Gates’ email provides sufficient evidence upon which a jury could find that the reasons for the October 3, 1994 decision to withdraw support for the namespace extensions were pretextual.” JA-213. Once Gates’ admissions established the anticompetitive character of the alteration of the course of dealing, Novell carried its burden of establishing that Microsoft altered its course of dealing “to exclude rivals on some basis other than efficiency.” Aspen Skiing, 472 U.S. at 605 (citation and internal quotation marks omitted).

Ample additional evidence further demonstrated that the de-documentation of the namespace extension APIs ran counter to Microsoft’s legitimate interests in increasing sales of Windows 95 by improving the product, and thus “suggested a

39

willingness to forsake short-term profits to achieve” the “anticompetitive end” that Gates identified. Trinko, 540 U.S. at 409. For example, Microsoft created the First Wave Program to further its interest in having ISVs release new applications within 90 days of Windows 95’s release. See supra p. 13. When Microsoft de-documented the namespaces, however, it knew there was “no way” Marvel could stop using the namespace extension APIs and “still have” a “chance of shipping with Win ’95.” JA-5755. Microsoft must have recognized that if Marvel, a Microsoft application, could not timely overcome the de-documentation, then ISVs would face even greater difficulties in doing so, and therefore the de-documentation would impair Windows 95’s profitability.

Microsoft also viewed ISVs’ ability to extend the shell with the namespace extension APIs to be a competitive advantage for Windows 95 that Microsoft relinquished by de-documenting them. See JA-3494 (identifying as a competitive advantage of Windows 95 over Mac that the Explorer can “be extended by ISVs, so we can expect to see this very powerful application be used as a browser for all kinds of information stored in all kinds of places”); JA-14751-52 (Nakajima) (explaining that enabling developers to customize the Explorer using the namespace extension APIs helped Windows 95 better suit the needs of users); JA-11306 (Harral) (describing how Novell’s planned uses of the APIs would help “make Windows the best version of Windows that it could be”). Microsoft’s later

40

republication of the namespace extension APIs establishes that it had an interest in making them available to ISVs and that the de-documentation ran counter to that interest.6 JA-11286-87 (Harral).

According to the District Court, Novell failed to establish anticompetitive conduct because there “is no evidence that Microsoft withdrew support for the namespace extension APIs for the purpose of terminating its relationship with Novell.” JA-218. By requiring complete termination of a relationship, the District Court erred. The “critical fact in Aspen Skiing was that there were no valid business reasons for the refusal.” Christy Sports, 555 F.3d at 1197; see also Aspen Skiing, 472 U.S. at 608 (“Perhaps most significant, however, is the evidence related to Ski Co. itself, for Ski Co. did not persuade the jury that its conduct was justified by any normal business purpose.”). When the evidence demonstrates the absence of a valid business reason, Aspen Skiing does not require a complete refusal to deal. See MetroNet Servs. Corp. v. Qwest Corp., 383 F.3d 1124, 1132

41

(9th Cir. 2004) (“An offer to deal with a competitor only on unreasonable terms and conditions can amount to a practical refusal to deal.”); Safeway Inc. v. Abbott Labs., 761 F. Supp. 2d 874, 892-95 (N.D. Cal. 2011) (upholding claim based on alteration of course of dealing where defendant’s licensing agreements continued to allow its competitors to use the trademark of its drug in their promotional materials but defendant suddenly raised the price of that drug); A.I.B. Express, Inc. v. FedEx Corp., 358 F. Supp. 2d 239, 250-51 & n.86 (S.D.N.Y. 2004) (defendant refused to deal on same terms as offered over prior five-year course of dealing); Nobody in Particular Presents, Inc. v. Clear Channel Commc’ns, Inc., 311 F. Supp. 2d 1048, 1106-08, 1112-14 (D. Colo. 2004) (radio station denied access to concert promoter even though station still permitted promoter to purchase ads and support).7

42

B. Microsoft’s Conduct Was Reasonably Capable Of Contributing
Significantly To Maintaining Microsoft’s Monopoly Power In The
Operating Systems Market

1. The District Court Applied An Incorrect Legal Standard

In assessing the harm to competition in the operating systems market caused by Microsoft’s conduct, the District Court erroneously departed from the D.C. Circuit’s test and reasoning in Microsoft, as well as the standard the District Court had applied on summary judgment. See Novell, 505 F.3d at 309 (“Novell’s present claims echo the government’s theory” from Microsoft).

Microsoft held that Microsoft harmed competition by injuring two middleware developers, Netscape and Sun, that together “showed potential as middleware platform threats” sufficient to “weaken” or “erode” – but not eliminate – “the applications barrier to entry” that protected Microsoft’s monopoly power in the operating systems market. Microsoft, 253 F.3d at 55, 79, 82 (citing FOF ¶¶ 68- 77). The D.C. Circuit recognized that Section 2 liability does not “turn on a plaintiff’s ability or inability to reconstruct the hypothetical marketplace absent a defendant’s anticompetitive conduct.” Id. at 79. Such a standard “would only encourage monopolists to take more and earlier anticompetitive action.” Id. Neither “plaintiffs nor the court can confidently reconstruct a product’s hypothetical technological development in a world absent the defendant’s exclusionary conduct. To some degree, ‘the defendant is made to suffer the

43

uncertain consequences of its own undesirable conduct.’” Id. (quoting 3 Areeda & Hovenkamp, Antitrust Law ¶ 651c at 78).

Microsoft’s Section 2 liability thus did not depend on showing that Navigator and Java “would actually have developed into viable platform substitutes” for Windows. Id. Instead, the Court recognized that “as a general matter the exclusion of nascent threats is the type of conduct that is reasonably capable of contributing significantly to a defendant’s continued monopoly power.” Id. This is because it “would be inimical to the purpose of the Sherman Act to allow monopolists free reign to squash nascent, albeit unproven, competitors at will – particularly in industries marked by rapid technological advance and frequent paradigm shifts.” Id.

The District Court followed these basic principles in rejecting Microsoft’s contention at summary judgment that its conduct did not cause anticompetitive harm in the operating systems market. See Novell, 699 F. Supp. 2d at 748 (relying on the “hypothetical marketplace” discussion in Microsoft and stating that when “a firm has engaged in anticompetitive conduct, courts should be reluctant to demand too much certainty in proving that such conduct caused anticompetitive harm”); see also King & King Enters. v. Champlin Petroleum Co., 657 F.2d 1147, 1162 n.1 (10th Cir. 1981) (holding in determining antitrust damages that the “most elementary conceptions of justice and public policy require that the wrongdoer

44

shall bear the risk of the uncertainty which his own wrong has created”) (citing Bigelow v. RKO Radio Pictures, Inc., 327 U.S. 251, 265 (1946)).

In granting Microsoft’s Rule 50 motion, however, the District Court did not even mention the “hypothetical marketplace” principle cited in its summary judgment opinion. Instead, as discussed below, it adopted assumptions that would permit the crushing of nascent threats that, as it previously recognized, the antitrust laws are designed to prevent. The District Court’s Rule 50 ruling thus departed from the law of the case as well as Microsoft.

2. The Elimination Of Novell’s Key Franchise Applications
Was Reasonably Capable Of Contributing Significantly To
Maintaining Microsoft’s Monopoly Power

Ample evidence established that, by eliminating WordPerfect as a viable competitive threat, Microsoft’s anticompetitive conduct strengthened the applications barrier to entry and thereby unlawfully maintained its monopoly power in the operating systems market.

Microsoft knew its strength in the applications market protected its monopoly power in the operating systems market. See JA-4293; supra pp. 6-8. At the time of Microsoft’s anticompetitive conduct, the word processor was the most important franchise, and WordPerfect – with the best reviews in the industry and, in Microsoft’s words, “its huge and loyal installed base,” JA-2904 – was by far the most potent competitor to Microsoft in word-processing. See supra pp. 8-10.

45

WordPerfect also ran on an array of operating systems both historically and during the relevant time period. See supra p. 9. The evidence established that WordPerfect would have continued to write for multiple operating systems that could have competed with Windows but for Microsoft’s conduct. See id.; JA-11931-33 (Frankenberg).

Based on this evidence, a reasonable jury could conclude that Microsoft’s destruction of WordPerfect and PerfectOffice was reasonably capable of contributing significantly to maintaining Microsoft’s monopoly power. Indeed, Microsoft recognized during the relevant time period that WordPerfect’s presence (or lack thereof) on an operating system would affect that platform’s ability to compete. See JA-14205 (Struss) (“WordPerfect was a major software application, that we felt it was critical to have their support for Windows 95 and for Windows 95 to be successful.”); JA-2668 (citing WordPerfect’s decision to stop development of its “OS/2 versions” as “a great example of how we kill OS/2” – a competing operating system). After the late release of PerfectOffice, Microsoft Word went from having an even installed base with WordPerfect – which was consistently producing superior products – to a 90 percent share. See supra pp. 8-9, 26. Microsoft thereby “dramatically” widened “the ‘moat’ that protects the operating systems business.” JA-4293. The District Court’s conclusion that Novell’s key franchise did not threaten Microsoft’s monopoly power was based on

46

two fallacious assumptions that improperly shifted the burden to Novell of resolving the uncertainty created by Microsoft’s conduct.

First, the District Court asserted that “Novell recognized that Windows 95 was a superior operating system, constituting a ‘significant step forward’” and thus that there was “no basis for inferring that office productivity applications Novell developed that did not draw upon the superior functionality of Windows 95 would have been as successful as the applications that ran on Windows 95.” JA-221.

In fact, Gary Gibb explained that Windows 95 was a “significant step forward” only when contrasted with Microsoft’s prior product, which was “really old technology” and a “pretty face on a very poor technology architecture.” JA-11722 (Gibb). Windows 95 was approaching “pretty much a part of what everybody else was” and “was not revolutionary in its technology.” Id.; see also JA-2369 (email from Microsoft executive stating, “No one thought” a preview of the Windows 95 user interface was “earth-shattering” and that most felt it reflected “‘obvious’ improvements”); JA-4432 (Windows 95 shell was less extensible than OS/2’s shell).

In violation of the hypothetical marketplace principle (see supra pp. 43-45), the District Court improperly required Novell to speculate about alternative operating systems, the applications Novell would have designed for those systems, and how those hypothetical systems and applications would have affected

47

consumer choices. A reasonable jury could have concluded that Novell’s cross- platform capability and considerable success and “responsiveness” (as Microsoft acknowledged) in designing products prior to Windows 95, JA-2964, established that it could and would have designed competitive applications on non-Windows 95 platforms.

Second, the District Court concluded Novell was not a threat because Microsoft had previously maintained its monopoly power notwithstanding WordPerfect’s prior popularity. JA-221. In addition to violating the Rule 50 standard, this inference in Microsoft’s favor assumed that the competitive world that existed after the rise of the Internet and middleware would have mirrored the world that existed before that time. This assumption contradicted binding findings of fact from the Government’s case against Microsoft. See supra pp. 7-8. That WordPerfect in the past had failed by itself to displace Windows proved nothing about its potential to do so in this new world. At a minimum, the availability of a popular word processor would have been a necessary condition for an operating system to compete with Windows. For this reason alone, Microsoft’s elimination of WordPerfect – the top rival to Microsoft Word – contributed significantly to maintaining its monopoly power.

Moreover, the conclusion that WordPerfect could not have worked with these nascent paradigms to challenge Microsoft’s monopoly power inappropriately

48

required Novell to recreate the hypothetical marketplace that would have existed absent Microsoft’s anticompetitive conduct toward WordPerfect, Navigator, and Java. The District Court further disregarded the evidence that Microsoft itself viewed all three technologies as threats to the maintenance of its monopoly power in the operating systems market. A reasonable jury could have concluded that a popular cross-platform suite (PerfectOffice) that boasted the “best word processor” (WordPerfect), and a cross-platform Internet browser that Microsoft viewed with “dread” (Navigator), combined with middleware technology (Java), would have posed a considerable threat to the applications barrier to entry and Microsoft’s monopoly power. JA-14668 (Ford); JA-15065 (LeFevre); JA-1629-30 (FOF ¶ 77).

3. The Elimination Of Novell’s Middleware Was Reasonably
Capable Of Contributing Significantly To Maintaining
Microsoft’s Monopoly Power

The District Court also disregarded ample evidence of the threat posed by Novell’s middleware. This evidence included Microsoft’s own recognition of the threat posed by Novell’s AppWare. By Microsoft’s own admission, AppWare was one of its “most serious competitors.” JA-645 (Maritz); see also JA-2659 (AppWare was “direct competition to Windows”); supra pp. 11-12. A reasonable jury could readily have concluded that Microsoft’s own assessment reflected the reality of the threat that it faced. The District Court did not address this evidence.

Likewise, PerfectFit made available more APIs than Navigator and Java

49

combined, and enabled ISVs to develop applications that could be ported to any operating system where PerfectOffice was installed. See supra p. 12. Further, more than 1,000 companies became “PerfectFit Partners,” who received dedicated Novell technical support. JA-11156-57 (Harral); JA-11718, 11821-22 (Gibb); JA-3437.

Novell further presented ample evidence that Microsoft’s exclusion of Novell’s middleware harmed competition in the operating systems market in the same way as Microsoft’s exclusion of Navigator and Java. Like Navigator and Java, Novell’s office productivity applications complemented Windows, were very popular, and had the potential to gain widespread use on Windows 95 due to WordPerfect’s existing installed base. See supra pp. 8-9; JA-1628-30 (FOF ¶¶ 69-72, 76-77).

The District Court similarly failed to account for the cumulative nature of the threat posed by middleware that Microsoft recognized and binding findings of fact established. “Microsoft was concerned with middleware as a category of software; each type of middleware contributed to the threat posed by the entire category.” JA-1627-28 (FOF ¶ 68) (emphasis added). By requiring Novell to show that one middleware product, by itself, could have eliminated the applications barrier to entry, the District Court directly contradicted this finding.

The District Court concluded that Novell’s middleware was not a threat

50

based on a fundamental misconception about the technology. That the namespace extension APIs were “‘platform specific’ to Windows” is immaterial to whether Novell could have “effectively ported” its applications to other platforms. JA-222 (citation omitted). No evidence supports the suggestion that developing PerfectOffice with the namespace extension APIs in Windows 95 would have prevented the development of PerfectOffice for platforms without that feature. To the contrary, Novell’s applications had been successfully ported to more than a dozen operating systems. JA-11147 (Harral).

The District Court additionally erred in concluding that Novell’s middleware did not pose a threat because it would not have been ubiquitous on Windows 95 and/or because it did not by itself support full-featured personal productivity applications. JA-222-26. These requirements were based on the erroneous assumption that Novell’s middleware had to have the potential, by itself, to eliminate (rather than erode or diminish) the applications barrier to entry. JA-226; see also JA-12869 (Noll). This test in turn was specifically contradicted by the evidence already discussed and rejected by the two sources that the District Court used to support it.

First, the Court purported to rely on the D.C. District Court’s discussion of the applications barrier to entry to conclude that “elimination (or, at least, near elimination) of the barrier to entry” was required. JA-226. Both the D.C. District

51

Court and the D.C. Circuit, however, reached the opposite conclusion – that middleware had the potential to threaten Microsoft’s monopoly power in the operating systems market in the future by diminishing the barrier to entry even though it was not a reasonable substitute at the time. Microsoft, 253 F.3d at 55, 78-79. Any other conclusion would license Microsoft to crush individual nascent threats while benefiting from the evidentiary uncertainty that its conduct created.

Second, the District Court similarly disregarded Professor Noll’s testimony that, while the existence of a cross-platform application that was present on all or nearly all PCs would be a condition for completely “eliminating the applications barrier to entry,” it “is not correct to say that something less than that couldn’t increase competition” by weakening it. JA-12869 (Noll); see also JA-222.

The suggestion that middleware could only threaten Microsoft’s monopoly power if full-featured applications could be written to it, JA-223-26, also failed to acknowledge that Novell’s products were simultaneously middleware and a set of full-featured personal productivity applications that included the best application in the most important category. See supra pp. 8-12, 45-49. The APIs made available by PerfectOffice’s middleware thus did not have to support another applications suite to satisfy the user’s office productivity needs, because PerfectOffice itself fulfilled those needs. The District Court similarly failed to account for the D.C. Circuit’s conclusion that Microsoft’s conduct toward Navigator and Java

52

harmed competition even though neither had the ability to support full-featured applications. JA-1623 (FOF ¶ 28).

The District Court also mistakenly held that to the extent that Professor Noll’s testimony that diminishing the barrier to entry was sufficient was “based on the premise that other companies would produce similar middleware that, in combination with Novell’s products, would diminish the barrier to entry, there is no evidence that such other products existed.” JA-224. To the contrary, the findings of fact established that “each type of middleware contributed to the threat posed by the entire category,” JA-1627-28 (FOF ¶ 68), and contained extensive discussion of specific middleware products Java and Navigator, and the combined effects of those products “working together.” Microsoft, 253 F.3d at 78-79; JA-1627-30 (FOF ¶¶ 68-77).

Novell also presented evidence of an agreement with Netscape that enabled Novell to bundle Navigator – which eventually contained Java – with its applications. JA-11943 (Frankenberg); JA-3858-86; see also JA-12765-69 (Noll). Professor Noll testified that the result of Microsoft’s conduct was “that 7 million PCs do not have the WordPerfect middleware, and that [sic] do not have Netscape Navigator that would have been loaded with the WordPerfect middleware. So that substantially reduces the degree to which the applications barrier to entry is being undermined by WordPerfect and Novell.” See JA-12769 (Noll). WordPerfect and Navigator’s combined threat is indistinguishable from the threat posed by the Java- Navigator license agreement, except that the Novell agreement would have significantly increased the distribution potential for all three technologies. JA-1629-30 (FOF ¶¶ 76-77).

III. NOVELL PRESENTED AMPLE EVIDENCE THAT MICROSOFT’S
ANTICOMPETITIVE CONDUCT CAUSED NOVELL’S INJURY

A. The District Court Provided No Basis For Disregarding The
Considerable Evidence That Microsoft’s Common File Open
Dialog Was Not A Viable Competitive Option

As explained in more detail above, Novell presented ample evidence that Microsoft’s conduct left Novell with three non-viable options. See supra pp. 22-25. The first option – trying to use the incomplete June 1994 documentation to call upon the now undocumented namespace extension APIs – was reckless and futile. See supra p. 23. Neither the District Court nor Microsoft’s expert witness suggested that Novell should have pursued this option. JA-16000 (Bennett) (agreeing that “it’s a bad practice to call on undocumented APIs”). The second option – using the Windows 95 common file open dialog entailed an unacceptable sacrifice of functionality for WordPerfect. See supra pp. 23-25. The third option – attempting to recreate the operating system-level functionality provided by the now de-documented APIs in the WordPerfect custom file open dialog – caused Novell’s product to miss the critical 90-day window. See supra pp. 25-26.

54

In spite of this evidence, and even though it was the jury’s role to make these factual determinations, the District Court held that if the 90-day window was critical, “Novell clearly would have implemented a different development plan.” JA-228. The District Court, however, did not propose any new suggestions aside from the three non-viable options discussed above. See id. The District Court could not properly speculate under the governing standard of review regarding the existence of a mystery fourth option, let alone conclude that a reasonable jury could only have concluded that that option existed and was competitively viable.

The District Court also ignored the Rule 50 standard in deciding that “Novell could have released those products using Microsoft’s common file open dialog.” JA-228. Novell’s witnesses testified to the contrary. See JA-11561-63 (Richardson) (the common dialog “was a huge step backwards for us” and the third option of replicating the operating system-level functionality “was the only option we had left”); JA-11436 (Harral) (“we could have made the choice to use the common open dialog in 1994 so to ship ’95 but that also would have been a choice to have disenfranchised our customer base”); JA-11750-51, JA-11781-82 (Gibb) (explaining Novell could not cut the file open dialog “back to an extreme because, again, it was core differentiator [sic]” and that use of the common dialog “would be a huge step backwards for our customers”).

In its statement of facts, the District Court stated that Microsoft’s common

55

file open dialog “would have caused Novell to stop providing features that it had provided to its customers in the past, thereby ‘disenfranchis[ing] [its] customer base.’” JA-205 (alteration in original) (citing to Harral, Richardson, and Gibb testimony). Yet, when it concluded Novell “could have released” its products “using Microsoft’s common file open dialog,” JA-228, the District Court devoted only cursory discussion to the issue and did not even discuss the aforementioned testimony of Novell’s witnesses, much less provide a basis for concluding that a reasonable jury could not have credited it. Instead, the court cites two Microsoft witnesses who once worked for Novell – one of whom has since worked for Microsoft for over a decade. JA-228; see also JA-14971 (LeFevre). Under Rule 50, the District Court was required to credit Novell’s witnesses and grant Novell the benefit of all reasonable inferences. It failed to do so in violation of the governing standard.

B. The District Court’s Mistaken Inferences And Conclusions
Concerning Novell’s Complaints To Microsoft And Executive
Involvement Provided No Basis For Finding An Absence Of
Causation

In concluding that there was an absence of urgency at Novell that demonstrated a lack of causation, the District Court relied on a purported absence of (1) complaints “to anyone at Microsoft who could have reversed the decision” and (2) “evidence that any top-level executive was involved in the decision-making process concerning the writing of shared code for WordPerfect.” JA-227. This

56

evidence supposedly demonstrated that “Novell simply did not believe the deadline of bringing applications to market within 90 days of the release of Windows 95” was “material to the success of Novell’s applications.” JA-227 n.23. This analysis was flawed for a number of reasons.

The direct evidence of causation forecloses the District Court’s reliance on a supposed lack of complaints to Microsoft. See Statement of Facts Part V, Argument Part III.A. Regardless of how or if Novell complained or which executives were involved in the decision, Microsoft’s decision left Novell with no viable options for releasing a competitive product within the critical window. See id. Nothing in the antitrust laws requires that a wronged party involve particular executives in a decision or exhaust its administrative remedies with the party that has wronged it. Nor does Rule 50 allow a district court to conclude that the absence of one type of evidence – even if true – means a plaintiff has failed to carry its burden using other evidence.

Moreover, the District Court’s inference “that Novell simply did not believe the deadline of bringing applications to market within 90 days of the release of Windows 95,” JA-227 n.23, contradicted the direct evidence that Novell viewed the deadline as important. See supra p. 22. The District Court also ignored direct testimony concerning the shared code team’s 80-hour work weeks and a “Panic Mode” memorandum stating that there was “no conceivable way” that the dialog

57

could be complete by the target date of August 22, 1995. See supra p. 25; JA-11285 (Harral); JA-4200. Under the applicable standard of review, the District Court could not disregard this evidence of urgency in favor of its own views of how Novell should have handled Microsoft’s conduct.

Regarding complaints, Novell’s CEO testified that he complained “on a number of occasions” to Bill Gates in 1995 about the issue of undocumented APIs but that Gates “refused to talk about it.” JA-11965-66 (Frankenberg). Moreover, a reasonable jury could easily have concluded that a lack of complaints reflected not that the issue was unimportant, but that Microsoft had falsely told Novell that the de-documentation was based on technical justifications. See supra pp. 19-22, 38 & n.4. Novell’s dependent and competitive relationship with Microsoft also counseled against vehement complaints. See JA-11966-67 (Frankenberg) (Novell’s Netware product was “intimately dependent on Microsoft’s cooperation.”); JA-7170 (noting WordPerfect did not “want to tip their hand” regarding its use of the namespace extension APIs); JA-12919 (C. Myhrvold).

The District Court also concluded that Novell’s executives were uninvolved based entirely on the supposed absence of evidence of their involvement. JA-227-28. This was improper under a standard that required all reasonable inferences to be drawn in Novell’s favor. Moreover, ample evidence demonstrated contemporaneous executive involvement in the decision to replicate the namespace

58

extension API functionality. While relying on Novell CEO Bob Frankenberg’s testimony that “some or all” of a group of specific executives would have been involved in the decision, JA-206, the District Court disregarded Frankenberg’s testimony that in early 1995, he learned of “being denied access to interfaces that we previously had access to” from Mark Calkins, Jeff Waxman, or Ad Rietveld – all of whom were part of that group of executives. JA-11965, JA-12077-78, JA-12119 (Frankenberg). The lead engineer in shared code further testified that management knew that the issue was “a big deal.” JA-11354-55. Based on this evidence, a reasonable jury could have concluded that Novell’s highest executives kept themselves informed of the problems Microsoft caused and accepted the developers’ plans to address those problems.

The District Court thus had no basis to conclude that Novell left the decision to develop the custom file open dialog (rather than use Microsoft’s common dialog) to a “middle manager” (a dismissive reference to Gary Gibb) and low level programmers. JA-206-07. Indeed, elsewhere in its opinion, the District Court referred to Nolan Larsen, a Microsoft witness who reported to Gibb, as an “executive of Novell.” JA-205; JA-5271. Further, as Microsoft’s witness confirmed, Gibb (the Director of PerfectOffice Windows 95), Steve Weitzeil (the head of WordPerfect for Windows development), and Tom Creighton (the head of shared code), all favored the decision to develop the custom dialog. JA-14998,

59

JA-15069-70 (LeFevre).

C. The District Court’s Conclusion That Quattro Pro, Not
Microsoft’s Conduct, Caused The Delay Improperly Resolved A
Disputed Question Of Fact That Was For The Jury

The District Court also mistakenly concluded that Microsoft’s conduct did not prevent the timely release of PerfectOffice within the 90-day window because, due to a “mass exodus of programmers” in December 1995, Quattro Pro – PerfectOffice’s spreadsheet application – was not ready for release until 1996. JA-228. In reaching this conclusion, the District Court improperly resolved conflicting evidence in Microsoft’s favor rather than Novell’s, in direct contravention of the applicable standard of review.

The District Court ignored direct testimony from the Director of the PerfectOffice suite (Gibb) that Quattro Pro did not delay the release of PerfectOffice and that, had shared code delivered on time, Novell would have been able to release PerfectOffice within 60 to 90 days of the release of Windows 95. JA-11743, JA-11838 (Gibb); JA-11427 (Harral) (“WordPerfect was not late. Quattro Pro was not late. It was shared code that was late.”). The District Court provided no basis for concluding that a reasonable jury could not have credited this testimony. Moreover, numerous witnesses, called by both parties, confirmed that Gibb was in the best position to know the particular issues that arose in developing PerfectOffice. See JA-11482 (Harral); JA-12179-80 (Frankenberg); JA-14168

60

(Bushman); JA-14600-01, JA-14605 (Larsen); JA-14669 (Ford).

Contemporaneous records confirmed Gibb’s testimony. See JA-9147 (showing that Quattro Pro was “code complete” as of August 23, 1995). Likewise, when Novell’s Board of Directors reviewed the pending decline in value of its Business Applications Division, it did not mention Quattro Pro delays as a contributing factor, but acknowledging that “MS interference,” not Quattro Pro, “has delayed competitive 32-bit apps.” JA-9640.

The District Court cited Frankenberg’s testimony as contrary evidence, JA-228, despite acknowledging during trial that Frankenberg had no real knowledge of Quattro Pro’s development progress and that his testimony merely summarized the document with which he was presented (see DX 230, discussed below), JA-12082; JA-13854. The District Court similarly cites Karl Ford’s testimony, JA-228, although Ford testified that Gibb was in the best position to know, that he merely “heard” Quattro Pro was having issues and “figured that they were late and at risk with the schedule,” JA-14654, JA-14669 (Ford).

Documentary evidence relied upon by the District Court similarly offered no basis for concluding that a reasonable jury could not have credited Gibb’s testimony and the other evidence cited above:

  • The March 1996 “Release To Manufacturing” date for Quattro Pro in DX 231 reflects only that Novell intended to release the entire suite to manufacturing on the same date; it did not reflect that Quattro Pro, rather than shared code, was responsible for the delay. JA-9147. Instead, it

61

    shows that Quattro Pro was code complete in August 1995, well before PerfectFit. Id. No PerfectOffice application (including Quattro Pro) could be released to manufacturing without shared code. See JA-11427 (Harral); JA-8893.
  • DX 230 (JA-9145) is consistent with (1) Gibb’s testimony that any Quattro Pro bug fixes occurring in December 1995 were merely “cleaning and polishing” that would not have delayed the release of PerfectOffice had PerfectFit been ready, JA-11742-43 (Gibb); and (2) the fact that Quattro Pro beta testing (an established purpose of which is to identify and fix bugs, JA-11234 (Harral); JA-12327 (Alepin)) had itself been delayed by PerfectFit. JA-9147 (showing that Quattro Pro beta began on the same date as the PerfectFit and PerfectOffice beta though code complete well before those products); see also JA-11427 (Harral).
  • Despite the concerns about Quattro Pro’s development raised in DX 211 (JA-8893), Novell affirmatively chose a September 1995 release date. JA-11826 (Gibb).
  • DX 221, an email from March 1995 (JA-8988-89), is consistent with Gibb’s recollection that he was nervous “early on” that Quattro Pro might delay the suite but ultimately Quattro Pro was “very conservative in their estimates” and “over delivered” with a code completion date of August 1995, JA-11740, JA-11743 (Gibb); JA-9147.

Accordingly, Novell presented sufficient evidence for a reasonable jury to find that Microsoft’s conduct, not Quattro Pro, delayed the release of PerfectOffice 95. These were jury issues that the District Court improperly resolved on a Rule 50 motion and, therefore, a new trial is required.

CONCLUSION

For the foregoing reasons, Novell requests that the judgment of the District Court be reversed and that this case be remanded for trial.

62

November 21, 2012

Respectfully submitted,

/s/ Jeffrey M. Johnson
Jeffrey M. Johnson
James R. Martin
Miriam R. Vishio
DICKSTEIN SHAPIRO LLP
[address, phone, tax, email]

Max D. Wheeler
Maralyn M. English
SNOW, CHRISTENSEN & MARTINEAU
[address, phone, fax]

R. Bruce Holcomb
ADAMS HOLCOMB LLP
[address, phone, fax]

David Boies
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]

Stuart H. Singer
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]

Samuel C. Kaplan
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]

63

REQUEST FOR ORAL ARGUMENT

Novell respectfully requests oral argument on the issues raised by its appeal. Novell believes these antitrust issues are important and that oral argument would assist the Court in resolving them.

64

CERTIFICATE OF COMPLIANCE

As required by Fed. R. App. P. 32(a)(7)(c), I certify that this brief is proportionally spaced and contains 13,999 words. I relied on my word processor, which is Microsoft Word 2010, to obtain the word count. I certify that this information is true and correct to the best of my knowledge and belief formed after reasonable inquiry.

/s/ Jeffrey M. Johnson

65

CERTIFICATE OF DIGITAL SUBMISSION
AND PRIVACY REDACTIONS

I certify that a copy of the foregoing APPELLANT’S OPENING BRIEF as submitted in Digital Form via the Court’s CM/ECF system is an exact copy of the written document filed with the Clerk and has been scanned for viruses. In addition, I certify that all required privacy redactions have been made.

/s/ Jeffrey M. Johnson

66

CERTIFICATE OF SERVICE

I certify that on November 21, 2012, I caused the foregoing APPELLANT’S OPENING BRIEF to be served on all parties or their counsel of record through the Court’s CM/ECF system.

/s/ Jeffrey M. Johnson

67

[Attachment 1: Judge Motz's Decision; Attachment 2 and 3: Order - see PDF]

1 Please note that some documents in the Joint Appendix were produced with inaccurate dates because such documents, which were restored from back-up tapes, contain an auto-date feature.

2 The District Court ruled that certain findings of fact from United States v. Microsoft Corp., 84 F. Supp. 2d 9 (D.D.C. 1999), cited herein were binding in this case. See JA-380-81. Per the standard of review governing motions for judgment as a matter of law, see infra Argument Part I, Novell is entitled to the benefit of all reasonable inferences from these facts as well as all facts set forth herein.

3 The District Court found that “the text of Gates’ email provides sufficient evidence upon which a jury could find that the reasons for the October 3, 1994 decision to withdraw support for the namespace extensions were pretextual.” JA-213. The District Court further stated that it was “not unreasonable” to view an earlier 1993 plan as “casting light upon the reason for Gates’ decision to withdraw support for the namespace extension APIs on October 3, 1994.” JA-204 n.10.

4 See, e.g., Wichita Clinic, P.A. v. Columbia/HCA Healthcare Corp., No. 96-1336- JTM, 1997 WL 225966, at *7 (D. Kan. Apr. 8, 1997); Nobody in Particular Presents, Inc. v. Clear Channel Commc’ns, Inc., 311 F. Supp. 2d 1048, 1105 (D. Colo. 2004); Lantec, Inc. v. Novell, Inc., 146 F. Supp. 2d 1140, 1145 (D. Utah 2001), aff’d, 306 F.3d 1003 (10th Cir. 2002); see also JA-1807 n.49.

5 This is clear from examination of the full email chain, which begins with circulation of the script at 1:41 a.m. on October 12, 1994, with the instruction “PLEASE READ THIS ENTIRE DOCUMENT CAREFULLY BEFORE YOU DO ANYTHING ELSE!” JA-6926. The 12:03 p.m. email cited by the District Court then reported Microsoft was in the process of notifying the various ISVs, and stated, “So far Stac, Lotus, WP, Oracle, SCC appear to be OK with this.” Id.

6 In a footnote in a separate section of its opinion addressing harm to competition in the operating systems market, the District Court posits that increased applications sales from de-documenting the namespace extension APIs would have offset any profits Microsoft sacrificed in sales of Windows 95. JA-219 n.18. As an initial matter, there is no evidentiary basis for this assertion in the record, and all inferences must be drawn in Novell’s favor. Further, as Professor Noll testified when questioned by the District Court on this point, the hypothesis is irrelevant. JA-12905-09 (Noll). Microsoft’s profits in the applications market cannot immunize Microsoft for the harm to competition it caused in the operating systems market when both resulted from conduct that was not competition on the merits and when Microsoft’s anticompetitive gains in the applications market protected its operating systems market monopoly power. See id.; see also infra pp. 43-54.

7 The District Court also wrongly asserted that Windows 95 “contained Microsoft’s common open file dialog, and Novell never advised Microsoft that this dialog was insufficient for its own purposes.” JA-218. To the contrary, there is ample evidence that Microsoft Office used a custom file open dialog rather than the common file open dialog, JA-15224, JA-15303 (Belfiore), and Microsoft knew of the high quality of WordPerfect’s custom file open dialog and its superiority to Microsoft’s common file open dialog. See supra pp. 24-25. Microsoft thus would have been well aware that its common file open dialog would be insufficient for Novell’s purposes.


  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 )