FindLaw |

Legal News & Information

TABLE OF AUTHORITIES iv

GLOSSARY ix

INTRODUCTION 1

SUMMARY OF ARGUMENT 3

ARGUMENT 6

I. The District Court’s Tying Ruling Should Be Reversed 8

A. Windows and IE Are Not "Separate Products" under This Court’s Test 10

B. The Alleged Tie Did Not Have an Anticompetitive Effect on the Putative "Web Browser" Market 15

II. The District Court’s Monopoly Maintenance Ruling Should Be Reversed 17

A. Microsoft Does Not Possess "Monopoly Power" 18

1. Products Need Not Be Complete Substitutes To Be Included in the Same Product Market 18

2. Plaintiffs Have Not Satisfied Their Burden of Establishing Significant Barriers to Entry 20

3. Microsoft’s Behavior Confirms That It Faces Intense Competition 22

B. Microsoft Did Not Engage in Anticompetitive Conduct 24

1. Microsoft Did Not Exclude Navigator from the OEM Channel 26

Predatory Product Design 27

Microsoft’s OEM License Agreements 28

No Restriction on Distribution of Navigator 32

2. Microsoft Did Not Exclude Navigator from the IAP Channel 33

3. Plaintiffs’ Other Allegations Do Not Amount to a Section 2 Violation 34

Microsoft’s Agreements with Certain ICPs and ISVs 35

Microsoft’s Java Implementation 35

Microsoft’s Agreements with Apple 37

C. Plaintiffs Have Not Shown the Requisite Causal Link between Microsoft’s Conduct and Its Position in Operating Systems 37

III. The District Court’s Attempted Monopolization Ruling Should Be Reversed 40

A. Microsoft’s June 1995 Discussions with Netscape Did Not Constitute Attempted Monopolization 40

B. Microsoft’s Competition with Netscape Following the June 1995 Discussions Did Not Constitute Attempted Monopolization 43

1. Microsoft’s Desire To Keep Navigator from Becoming the "Standard" Web Browsing Software Is Not a Specific Intent To Monopolize 43

2. Plaintiffs Have Not Shown a Dangerous Probability of Monopolization 45

IV. The Relief Awarded by the District Court Cannot Stand 46

A. Microsoft Was Entitled to an Evidentiary Hearing 47

B. The Breakup of Microsoft Is Legally Indefensible 53

1. Plaintiffs Have Conceded Away the Basis for a Breakup 55

2. The Breakup Is Not Narrowly Tailored To Redress the Conduct Found To Be Anticompetitive 56

3. The Breakup Is Unjustified Given the Vagaries of the Legal Rules Applied to Microsoft’s Conduct 57

4. The Breakup Would Impair Important Public Benefits 57

5. The District Court Ignored Severe Implementation Problems Inherent in the Breakup 59

6. The Breakup Is Punitive in Concept and Effect 59

C. The Other Relief Awarded by the District Court Is Punitive and Unjustified 61

1. Confiscation of Microsoft’s Valuable Intellectual Property 62

2. Interference with the Design of Microsoft’s Operating Systems 64

V. Microsoft Was Severely Prejudiced by the District Court’s Handling of the Case 66

VI. The District Judge’s Repeated Extrajudicial Statements and Discussions Provide an Independent Basis for Vacating the Judgment 72

CONCLUSION 75

(Authorities on which Microsoft chiefly relies are marked with asterisks.)

Am. Can Co. v. Mansukhani, 814 F.2d 421 (7th Cir. 1987) 48

Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985) 24

* Ass’n for Intercollegiate Athletics for Women v. NCAA, 735 F.2d 577 (D.C. Cir. 1984) 43

Atchinson v. District of Columbia, 73 F.3d 418 (D.C. Cir. 1996) 68

Berger v. United States, 255 U.S. 22 (1921) 74

* Berkey Photo, Inc. v. Eastman Kodak Co., 603 F.2d 263 (2d Cir. 1979), cert. denied, 444 U.S. 1093 (1980) 57, 63

Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209 (1993) 22

Brown Shoe Co. v. United States, 370 U.S. 294 (1962) 20

Caribbean Broad. Sys., Ltd. v. Cable & Wireless PLC, 148 F.3d 1080 (D.C. Cir. 1998) 68

Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752 (1984) 17-18

Davoll v. Webb, 194 F.3d 1116 (10th Cir. 1999) 48

Dial A Car, Inc. v. Transp., Inc., 82 F.3d 484 (D.C. Cir. 1996) 45

Eastman Kodak Co. v. Image Technical Servs., Inc., 504 U.S. 451 (1992) 18, 22

Eli Lilly & Co. v. Generix Drug Sales, Inc., 460 F.2d 1096 (5th Cir. 1972) 48

Fonar Corp. v. Domenick, 105 F.3d 99 (2d Cir.), cert. denied, 522 U.S. 908 (1997) 30

Fuentes v. Shevin, 407 U.S. 67 (1972) 51-52

* Gilliam v. ABC, 538 F.2d 14 (2d Cir. 1976) 30

Granny Goose Foods, Inc. v. Bhd. of Teamsters, Local No. 70, 415 U.S. 423 (1974) 47

Great Escape, Inc. v. Union City Body Co., 791 F.2d 532 (7th Cir. 1986) 44

H.J., Inc. v. ITT, 867 F.2d 1531 (8th Cir. 1989) 27

Hartford-Empire Co. v. United States, 323 U.S. 386 (1945) 59, 61

Hilyer v. Howaat Concrete Co., 578 F.2d 422 (D.C. Cir. 1978) 71

* ILC Peripherals Leasing Corp. v. IBM, 458 F. Supp. 423 (N.D. Cal. 1978), aff’d sub nom. Memorex Corp. v. IBM, 636 F.2d 1188 (9th Cir. 1980), cert. denied, 452 U.S. 972 (1981) 20, 27

In re Fine Paper Antitrust Litig., 685 F.2d 810 (3d Cir. 1982) 66

* In re IBM Peripheral EDP Devices Antitrust Litig., 481 F. Supp. 965 (N.D. Cal. 1979), aff’d sub nom. Transamerica Computer Co. v. IBM, 698 F.2d 1377 (9th Cir.), cert. denied, 464 U.S. 955 (1983) 19-20, 22, 24, 28, 44

Int’l Boxing Club of New York, Inc. v. United States, 358 U.S. 242 (1959) 61

Int’l Longshoremen’s Ass’n v. Philadelphia Marine Trade Ass’n, 389 U.S. 64 (1967) 64

Int’l Salt Co. v. United States, 332 U.S. 392 (1947) 56

* Intergraph Corp. v. Intel Corp., 195 F.3d 1346 (Fed. Cir. 1999) 34

Jack Walters & Sons Corp. v. Morton Bldg., Inc., 737 F.2d 698 (7th Cir.), cert. denied, 469 U.S. 1018 (1984) 10

Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2 (1984) 9, 10, 15

Liteky v. United States, 510 U.S. 540 (1994) 73, 74

M & M Med. Supplies & Serv., Inc. v. Pleasant Valley Hosp., Inc., 981 F.2d 160 (4th Cir. 1992) 46

MCI v. AT&T, 708 F.2d 1081 (7th Cir.), cert. denied, 464 U.S. 891 (1983) 44

Multi-Med. Convalescent & Nursing Ctr. of Towson v. NLRB, 550 F.2d 974 (4th Cir. 1977) 69

Multistate Legal Studies, Inc. v. Harcourt Brace Jovanovich Legal & Prof’l Publ’ns, Inc., 63 F.3d 1540 (10th Cir. 1995) 45

Neumann v. Reinforced Earth Co., 786 F.2d 424 (D.C. Cir. 1986) 29

Olympia Equip. Leasing Co. v. W. Union Tel. Co., 797 F.2d 370 (7th Cir. 1986), cert. denied, 480 U.S. 934 (1987) 24

* Omega Envtl., Inc. v. Gilbarco, Inc., 127 F.3d 1157 (9th Cir. 1997), cert. denied, 525 U.S. 812 (1998) 26

Roland Mach. Co. v. Dresser Indus., Inc., 749 F.2d 380 (7th Cir. 1984) 26

Roy B. Taylor Sales, Inc. v. Hollymatic Corp., 28 F.3d 1379 (5th Cir. 1994) 16

Ryko Mfg. Co. v. Eden Servs., 823 F.2d 1215 (8th Cir. 1987), cert. denied, 484 U.S. 1026 (1988) 26

S. Card & Novelty, Inc. v. Lawson Mardon Label, Inc., 138 F.3d 869 (11th Cir. 1998) 15

* S. Pac. Communications Co. v. AT&T, 740 F.2d 980 (D.C. Cir. 1984), cert. denied, 470 U.S. 1005 (1985) 3, 7, 75

Schine Chain Theatres, Inc. v. United States, 334 U.S. 110 (1948) 56

Scott v. Davis, 404 F.2d 1373 (5th Cir. 1968) 50

* Sims v. Greene, 161 F.2d 87 (3d Cir. 1947) 50

Smith Mach. Co. v. Hesston Corp., 878 F.2d 1290 (10th Cir. 1989), cert. denied, 493 U.S. 1073 (1990) 15

Socialist Workers Party v. Ill. State Bd. of Elections, 566 F.2d 586 (7th Cir. 1977), aff’d, 440 U.S. 173 (1979) 48

Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447 (1993) 47

Stitt Spark Plug Co. v. Champion Spark Plug Co., 840 F.2d 1253 (5th Cir.), cert. denied, 488 U.S. 890 (1988) 26

United Shoe Mach. Corp. v. United States, 347 U.S. 521 (1954) 54

United States v. Am. Tobacco Co., 221 U.S. 106 (1911) 51

United States v. Am. Airlines, Inc., 743 F.2d 1114 (5th Cir. 1984), cert. dismissed, 474 U.S. 1001 (1985) 5, 40, 41, 42

* United States v. Baker Hughes Inc., 908 F.2d 981 (D.C. Cir. 1990) 20, 21

United States v. Cont’l Can Co., 378 U.S. 441 (1964) 19

* United States v. Cooley, 1 F.3d 985 (10th Cir. 1993) 73

United States v. E.I. du Pont de Nemours & Co., 177 F. Supp. 1 (N.D. Ill. 1959), vacated, 366 U.S. 316 (1961) 50

United States v. E.I. du Pont de Nemours & Co., 366 U.S. 316 (1961) 53

United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377 (1956) 19

* United States v. Microsoft Corp., 147 F.3d 935 (D.C. Cir. 1998) passim

* United States v. Microsoft Corp., 56 F.3d 1448 (D.C. Cir. 1995) 73, 74, 75

United States v. Microsoft Corp., 97 F. Supp. 2d 59 (D.D.C. 2000) 54

United States v. Microsoft Corp., 87 F. Supp. 2d 30 (D.D.C. 2000) 17, 18, 25, 26, 57

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

United States v. Microsoft Corp., 1998-2 Trade Cas. (CCH) ¶ 72,261 (D.D.C. Sept. 14, 1998) 10

United States v. Nat’l City Lines, Inc., 134 F. Supp. 350 (N.D. Ill. 1955) 56

United States v. Paramount Pictures, Inc., 334 U.S. 131 (1948) 49

United States v. Syufy Enters., 903 F.2d 659 (9th Cir. 1990) 23

United States v. U.S. Gypsum Co., 340 U.S. 76 (1950) 56

United States v. United Shoe Mach. Corp., 391 U.S. 244 (1968) 54

United States v. United Shoe Mach. Corp., 266 F. Supp. 328 (D. Mass. 1967), rev’d, 391 U.S. 244, 252 (1968) 54

United States v. United Shoe Mach. Corp., 110 F. Supp. 295 (D. Mass. 1953), aff’d, 347 U.S. 521 (1954) 53-54, 60

United States v. Ward Baking Co., 376 U.S. 327 (1964) 51

* WGN Cont’l Broad. Co. v. United Video, Inc., 693 F.2d 622 (7th Cir. 1982) 30-31

28 U.S.C. § 455 6, 73

Fed. R. Evid. 602 67

Fed. R. Evid. 703 50

Phillip E. Areeda et al., Antitrust Law (1996) 10, 45, 57, 63

Ken Auletta, World War 3.0: Microsoft and Its Enemies (2001) 2-3, 46-47, 59, 72-73, 74-75

Ken Auletta, Final Offer, New Yorker, Jan. 15, 2001, at 40 46, 74

Joel Brinkley & Steve Lohr, U.S. v. Microsoft (2000) 72

Joel Brinkley & Steve Lohr, Retracing the Missteps in the Microsoft Defense, N.Y. Times, June 9, 2000, at A1 53, 74

CNN Morning News, Jan. 16, 2001 72

Editorial, Judge Jackson’s Remarks, Wash. Post, Jan. 16, 2001, at A20 73

Editorial, So Much for the Rule of Law, N.Y. Post, Jan. 10, 2001, at 32 73

Lino A. Graglia, Is Antitrust Obsolete?, 23 Harv. J.L. Pub. Pol’y 11 (1999) 54

James V. Grimaldi, Microsoft Judge Says Ruling at Risk, Wash. Post, Sept. 29, 2000, at E1 74

Manual for Complex Litigation (Third) (1997) 67

McCormick on Evidence (1999) 70

Report of the Judicial Conference Committee on Codes of Conduct (Sept. 1992) 73

Joseph Gregory Sidak, Debunking Predatory Innovation, 83 Colum. L. Rev. 1121 (1983) 28

Lawrence H. Summers, "The Wealth of Nations" (Remarks, May 10, 2000) 23

Charles A. Wright & Victor J. Gold, Federal Practice and Procedure (1990) 69

"APIs" Application Programming Interfaces. APIs are interfaces exposed by an operating system or other platform software that can be invoked by other software to obtain a wide range of system services, such as displaying text on the video monitor or saving a document to the hard disk.

"DOJ" Antitrust Division of the U.S. Department of Justice.

"FF" Findings of fact issued by the district court on November 5, 1999 and reported at 84 F. Supp. 2d 9 (D.D.C. 1999).

"IAPs" Internet Access Providers. Plaintiffs use the term IAPs to refer collectively to both ISPs and OLSs.

"ICPs" Internet Content Providers. ICPs are entities that provide content and applications to users of the Internet by maintaining Web sites.

"IE" The Internet Explorer components of Windows operating systems.

"ISPs" Internet Service Providers. ISPs provide their subscribers with a connection to the Internet via telephone, cable or satellite, typically in exchange for a monthly fee.

"ISVs" Independent Software Vendors. ISVs are entities engaged in developing and marketing software products, including applications, tools and utilities.

"Java" The term Java applies to three separate but related concepts: (i) a high-level programming language developed by Sun, (ii) a set of class libraries that provide services to Java programs, and (iii) a runtime called a Java Virtual Machine that executes Java programs. The Java programming language is an object-oriented variant of the C++ programming language that has been simplified to eliminate features like memory pointers that cause common programming errors.

"JVM" Java Virtual Machine. Software that executes programs written in the Java programming language after they have been compiled into intermediate instructions called Java bytecodes. JVMs are included with popular operating systems from vendors like Apple, Hewlett-Packard, IBM, Microsoft and Novell.

"Linux" An operating system whose kernel was created in 1991 by Linus Torvalds, a Finnish graduate student, and subsequently expanded and improved on a cooperative basis by software developers around the world.

"MS Br." Brief for Defendant-Appellant Microsoft Corporation.

"Navigator" Web browsing software developed by Netscape. Netscape and Navigator are now owned by AOL Time Warner.

"Netscape" Netscape Communications Corp. A company co-founded by Marc Andreessen and Jim Clark that developed the first commercially successful Web browsing software. Netscape is now owned by AOL Time Warner.

"NSP" Native Signal Processing. NSP is the name Intel gave to a group of software products designed to consume additional microprocessor cycles and thereby spur demand for more advanced Intel products. NSP narrowly refers to the ability of Intel microprocessors to process certain digital signals internally rather than through the use of an external device known as a digital signal processor or DSP. As originally designed by Intel in 1995, NSP software was incompatible with both Windows 95 and Windows NT.

"OEMs" Original Equipment Manufacturers. OEMs are manufacturers or assemblers of personal computers.

"OLSs" Online Services. OLSs provide their subscribers with a connection to the Internet as well as proprietary content and services like e-mail and personal Web pages.

"PC" Personal Computer.

"Pls. Br." Brief for Appellees United States and the State Plaintiffs.

"Windows 95" A client operating system developed by Microsoft that was commercially released in August 1995. Windows 95 incorporated the functionality of MS-DOS and Windows 3.1 in a 32-bit operating system with a range of new features and functionality. Windows 95 was updated with various interim releases, called OSRs, that were provided by Microsoft to OEMs.

"Windows 98" A client operating system developed by Microsoft that was commercially released in June 1998. Windows 98 is an improved version of Windows 95 that includes a range of new features and functionality. Windows 98 has been updated by interim releases provided to OEMs.

"Windows NT" An operating system developed by Microsoft that was first commercially released in 1992. Windows NT has a fundamentally different architecture than Windows 95, Windows 98 and Windows Millennium. The three principal versions of Windows NT (3.1, 3.51 and 4.0) were targeted primarily at business customers. Windows NT is the predecessor of Windows 2000.

"Windows Update" A feature of Windows 98 that automatically detects when new enhancements to the operating system have not been installed on a user’s computer and allows the user to download such enhancements easily from a Microsoft Web site.

Plaintiffs acknowledge that much of the conduct they challenge was lawful and procompetitive. On the central issue in the case, plaintiffs concede that Microsoft did not violate the antitrust laws by "offering IE to OEMs in a bundle with Windows at no extra charge." Pls. Br. at 63-64. Indeed, they admit that the district court found Microsoft’s conduct "in developing a Web browser and offering it to OEMs and users with Windows to be lawful." Id. at 66. With regard to the challenged distribution agreements, plaintiffs concede that "[d]eveloping and improving [IE], making it widely available, and encouraging its use based on the merits of the product constituted competition on the merits." Id. at 62. And, with regard to Java, plaintiffs acknowledge that Microsoft "develop[ed] an attractive alternative Java implementation" for Windows. Id. at 66-67. Whatever remains of plaintiffs’ case after these concessions does not amount to a Sherman Act violation, and is certainly not sufficient to justify breaking up Microsoft and imposing other extreme relief.

Despite acknowledging the procompetitive nature of much of the challenged conduct, plaintiffs’ over-arching theme is that Microsoft’s conduct "would be inexplicable for a non-monopolist." Id. at 42. Contrary to plaintiffs’ repeated assertions, e.g., id. at 3, 36, 65, the challenged conduct made perfect "business sense" for anyone seeking to develop and market operating systems that appeal to users and ISVs. In response to demand for Internet-related functionality, all major operating systems—not just Windows—now include Web browsing software. Schmalensee ¶¶ 216-23; Allchin ¶¶ 261-77. In fact, Netscape correctly predicted in January 1995—before any of the challenged conduct occurred—that all vendors would "eventually incorporate some Web browser functions into their operating systems as standard features" and "offer this functionality at little or no additional cost to customers." DX 68 at 69242. Microsoft announced its decision to "include built-in access to the Internet" in Windows long before Netscape released its first version of Navigator. MS Br. at 21. Even plaintiffs grudgingly concede that "improving the quality of Windows by including Web browsing or other functionalities without separate charge might have made business sense." Pls. Br. at 62; see also United States v. Microsoft Corp., 84 F. Supp. 2d 9, 45 (D.D.C. 1999) (FF 140).

After investing "more than $100 million each year" to improve the IE components of Windows, Pls. Br. at 15, it also made perfect business sense for Microsoft to enter into agreements designed to achieve widespread distribution of IE, thereby helping to attract ISV interest in the new platform capabilities IE provided. Such cross-marketing agreements are commonplace in all industries, not just in software. 1/4/99 pm Tr. at 83-91; 2/9/99 am Tr. at 31-37; GX 1256 at 294, 295; DX 1427; DX 2082. Microsoft has been following this same business model—improving its platforms to meet demand for additional functionality, distributing those improvements broadly and encouraging ISVs to utilize those improvements—for 20 years.

Instead of addressing the fatal defects in their case, plaintiffs drape themselves in the district court’s findings of fact, intimating that this Court is constrained to accept those findings without subjecting them to scrutiny. In considering this invitation to rubberstamp what the district court did, the Court should bear in mind the district court’s goal in drafting its findings. As the district judge told Ken Auletta, a reporter for the New Yorker writing a book about the case, he separated his findings of fact from his conclusions of law (thus enabling him to import numerous conclusions into his findings) in order to insulate his ruling from reversal:

There may have been another motive to split the facts from the law, whispered lawyers on opposite sides of the case: Judge Jackson was trying to box in the Court of Appeals. Guilty, Jackson later admitted to me, "The general rule of law is that the Court of Appeals is generally expected to defer to the trial judge as to matters of fact—unless the findings are clearly erroneous . . . . What I want to do is confront the Court of Appeals with an established factual record which is a fait accompli. And part of the inspiration for doing that is that I take mild offense at their reversal of my preliminary injunction in the consent-decree case, where they went ahead and made up about ninety percent of the facts on their own."

Ken Auletta, World War 3.0: Microsoft and Its Enemies 230 (2001).

The district judge’s repeated public comments (see MS Br. at 146-50) also demonstrate an animus towards Microsoft so strong that it inevitably infected his rulings. For example, Auletta reports that the district judge compared Microsoft executives to unremorseful gang members convicted of brutal, execution-style murders, Auletta, supra at 369-70, and criticized Microsoft’s CEO for having "‘a Napoleonic concept of himself and his company,’" id. at 397. Indeed, in the public eye, the district judge has joined the fray as an adversary of Microsoft.

By contrast, the district judge told Auletta that he holds government lawyers in particularly high regard, giving credence to the charges they level against a defendant:

"I trust the lawyers from the Department of Justice. I trust civil servants and the U.S. attorneys. I’m not hostile to the government . . . . In criminal cases, by and large, my experience is that when the government charges someone, they are probably guilty. I give the benefit of presumptive innocence, but I know of no case of a wrongful conviction."

Id. at 44.

Such comments put the proper gloss on the district court’s findings. In another case in which there was reason—but far less reason—to question whether a district judge improperly decided a matter based on his personal views, this Court stated: "[I]n light of the special circumstances of this case, we have reviewed the District Court’s findings against the record with particular, even painstaking, care." S. Pac. Communications Co. v. AT&T, 740 F.2d 980, 984 (D.C. Cir. 1984), cert. denied, 470 U.S. 1005 (1985). Painstaking review is warranted here as well.

1. Plaintiffs continue to retreat from their tying claim. Although they pay lip service to the notion that Windows and IE are "separate products," plaintiffs make no effort to satisfy this Court’s test. In fact, plaintiffs no longer argue that IE can or should be removed from Windows, recognizing that other parts of the operating system and numerous Windows applications rely on IE. Instead of attempting to show that the many benefits of Microsoft’s integrated design could be duplicated by combining an operating system with a standalone Web browser such as Navigator, plaintiffs argue something completely different: that Microsoft violated Section 1 because it failed to permit its distributors—OEMs—to hide "user access" to the built-in Web browsing functionality in Windows. That argument flies in the face of this Court’s June 1998 decision and tying law generally. See United States v. Microsoft Corp., 147 F.3d 935, 941 n.3 (D.C. Cir. 1998) ("[B]y allowing OEMs to conceal IE, rather than to refuse it, the remedy fits poorly with the Department’s tying theory."). Plaintiffs’ failure to show that the alleged tie foreclosed competition in the "Web browser" market is also fatal to their tying claim. Microsoft’s lack of success with IE 1 and IE 2 is dispositive on this point. Although included in every copy of Windows 95 distributed through the OEM channel over a twelve-month period, IE 1 and IE 2 struggled to obtain a 5% usage share, while Navigator enjoyed a usage share exceeding 80%.

2. Plaintiffs have not shown that Microsoft possesses "monopoly power," i.e., the power to control price or exclude competition. Plaintiffs also have not established that Microsoft engaged in "anticompetitive" conduct, or that the challenged conduct contributed significantly to the maintenance of Microsoft’s supposed monopoly. The two principal pillars of plaintiffs’ monopoly maintenance claim—tying and exclusive dealing—were also the bases for plaintiffs’ Section 1 claims. Plaintiffs cite no authority for their assertion that conduct can be "exclusionary" under Section 2 even though it did not "unreasonably restrain trade" under Section 1. Nor do plaintiffs address the district court’s findings that Netscape was able to distribute 160 million copies of Navigator in 1998 alone and had access to "every PC user worldwide," which undermine any claim of foreclosure. Finally, plaintiffs failed to prove that there would be any more competition in the putative market for "Intel-compatible PC operating systems" if Microsoft had not engaged in the challenged conduct.

3. Plaintiffs base their attempted monopolization claim primarily on the June 1995 discussions between Microsoft and Netscape, which they concede resulted in nothing. Plaintiffs cite only one case, United States v. American Airlines, Inc., 743 F.2d 1114 (5th Cir. 1984), cert. dismissed, 474 U.S. 1001 (1985), in which a court found that making a proposal that is then rejected by the other party can constitute attempted monopolization. Microsoft’s efforts to persuade Netscape to utilize Internet-related functionality in Windows 95 bear no resemblance to the naked price-fixing proposal in American Airlines. Plaintiffs also ignore Barksdale’s admission that Microsoft was eager at the time for Netscape to release a version of Navigator for Windows 95 because Microsoft believed it would increase demand for the new operating system. Plaintiffs go on to argue that Microsoft’s "post-June 1995" conduct likewise constituted attempted monopolization. Yet they concede that Microsoft’s goal in competing with Netscape was to prevent Navigator from becoming the "standard" Web browsing software, not to obtain "monopoly power" in the alleged "Web browser" market. Plaintiffs also have not established a dangerous probability of monopolization, especially now that Navigator is in the hands of AOL Time Warner, a formidable competitor.

4. The extreme relief imposed by the district court without conducting an evidentiary hearing is unsustainable. Plaintiffs’ various arguments in defense of the district court’s decision to impose extreme relief based on uncross-examined declarations and excerpts from newspaper and magazine articles are specious. The district court was not at liberty to deny Microsoft an evidentiary hearing on disputed factual issues or to base permanent injunctive relief on rank hearsay. More fundamentally, neither the facts nor the law justifies breaking up Microsoft or imposing punitive "conduct" remedies that would result in confiscation of Microsoft’s intellectual property and unprecedented governmental regulation of the design of Microsoft’s operating systems. In addition to these defects, if the Court rejects any of the district court’s liability rulings, the entire decree must be vacated because it is impossible to determine which provisions were intended to redress which supposed violations.

5. Once plaintiffs were permitted to expand their case dramatically, the district court abused its discretion in rushing the case to trial and limiting both sides to twelve witnesses. Contrary to plaintiffs’ assertions, Microsoft never agreed to the district court’s unorthodox procedures. Nor can plaintiffs deny that it is impossible to ascertain whether the district court improperly relied on inadmissible hearsay because there are no record citations in its findings.

6. The district judge’s public comments about the merits of the case and his ad hominem attacks on Microsoft are indefensible. Plaintiffs cannot seriously contend, for example, that it was permissible for the district judge to compare Microsoft to "gangland killers" in comments made to a reporter while the case was still pending. Such comments (i) violate the Code of Conduct for United States Judges, (ii) are clear grounds for disqualification under 28 U.S.C. § 455(a), and (iii) provide an independent basis to vacate the judgment.

Plaintiffs argue that Microsoft waived any challenge to the district court’s findings because it did not "assert specifically that any fact found by the district court is clearly erroneous." Pls. Br. at 3. They acknowledge, however, that Microsoft expressly stated that the district court’s findings are subject to review for clear error, id. at 46, and their brief makes abundantly clear that they understand Microsoft is challenging the accuracy of many findings, particularly those relating to market definition, monopoly power, product design and Netscape’s distribution of Navigator through the OEM and IAP channels. Plaintiffs fail to cite a single case which holds that a defendant waives its challenge to factual findings if it does not expressly invoke the "clearly erroneous" mantra each time it takes issue with a finding. To the contrary, when an appellant expresses disagreement with a finding of fact, this Court reviews the evidence discussed by the district court and the parties and rejects the finding if the Court is "left with the definite and firm conviction that a mistake in factfinding has been committed by the trial court." S. Pac. Communications, 740 F.2d at 998.

Given the district court’s failure to cite any evidence in its findings of fact, plaintiffs have done nothing to assist this Court’s review by simply regurgitating the findings in their statement of case and only sporadically citing record evidence. Making matters worse, plaintiffs’ description of the facts is sometimes so incomplete as to be misleading. For example, in discussing Intel’s NSP software, Pls. Br. at 39-40, plaintiffs fail to mention that NSP was incompatible with Windows 95 and Windows NT, MS Br. at 55-57; DX 978. Plaintiffs also state that Microsoft delayed the release of Windows 98 to permit inclusion of IE 4, Pls. Br. at 23, but they fail to acknowledge that IE 4 was made available on October 1, 1997, eight months before the June 1998 release of Windows 98, 84 F. Supp. 2d at 13 (FF 8); Chase ¶ 136; 2/4/99 pm Tr. at 11-12. In other instances, plaintiffs’ description of the facts is simply false. For example, they assert that "[e]ven as late as June 1995," Microsoft was planning to include only "low-level Internet ‘plumbing’" with Windows 95. Pls. Br. at 20 (emphasis added). The cited finding, however, says something different: "[A]s late as the fall of 1994, Microsoft was planning to include low-level Internet ‘plumbing,’ such as a TCP/IP stack, but not a browser, with Windows 95." 84 F. Supp. 2d at 49 (FF 156) (emphasis added). And the two cited exhibits, GX 124 and GX 125, are dated April 1994 and June 1994, respectively. GX 124 (April 20, 1994 e-mail); GX 125 (June 10, 1994 e-mail); see also MS Br. at 20-24 (Microsoft’s decision to include IE in Windows predated Netscape’s emergence as a competitor).

For Microsoft to prevail, however, this Court need not reject a single finding of fact. As Microsoft explained in its opening brief, MS Br. at 68, even if one were to accept the district court’s one-sided findings in toto, the district court’s conclusion that Microsoft violated the Sherman Act and corresponding state-law provisions is unsustainable as a matter of law.

Plaintiffs continue to de-emphasize their tying claim, which was the focus of their complaints and their sole basis for urging that the case be rushed to trial. See 5/22/98 Tr. at 5-12, 20-24, 30-33. Plaintiffs no longer seriously contend that IE is a "separate product" that can be removed from Windows. Instead, they now claim that Microsoft violated the antitrust laws because it did not permit its distributors, OEMs, to remove from Windows "the means of user access" to some of the Web browsing functionality provided by IE. Pls. Br. at 104. By arguing that only "user access" to this functionality should be made removable, plaintiffs acknowledge that IE provides important functionality to other parts of the operating system (e.g., the Windows Update feature) and to applications running on Windows. Plaintiffs cite no case upholding a tying claim based on a defendant’s failure to permit its distributors to hide "user access" to valuable features of a product that benefit consumers.

Recognizing that they cannot establish that Windows 98 is an unlawful tie under controlling precedent, plaintiffs focus on Microsoft’s inclusion of IE 1 and IE 2 in Windows 95 in the twelve months between August 1995 and August 1996, something barely mentioned in their complaints. Pls. Br. at 101, 102, 105. Plaintiffs argue that "[w]hatever the implications of [this Court’s decision] for the versions of Windows and IE that are physically linked, it provides no basis to reverse the district court’s determination that Microsoft unlawfully tied IE 1 and IE 2." Id. at 102. That plaintiffs are reduced to arguing about obsolete versions of IE that have not been part of Windows for more than four years speaks volumes about their tying claim.

Although IE was less integrated in the initial releases of Windows 95 than it was in subsequent releases, Microsoft’s inclusion of IE 1 and IE 2 in the operating system had no antitrust significance because it resulted in no foreclosure of competition in the putative "Web browser" market. As the district court found, in late 1995 and early 1996, IE’s usage share was approximately 5%, whereas Navigator’s usage share exceeded 80%. 84 F. Supp. 2d at 98-99 (FF 360); see also Barksdale ¶ 222; Warren-Boulton ¶ 137; Fisher ¶ 230; Allchin ¶ 322; Chase ¶ 136. The fact that so many users in 1995 and 1996 chose to use Navigator rather than IE despite the presence of an IE icon on the desktop of every copy of Windows 95 distributed through the OEM channel demonstrates that the alleged tie had no competitive impact whatsoever in the market for the putative tied product. Schmalensee ¶¶ 524-25. Even plaintiffs’ own economist, Franklin Fisher, admitted that IE 1 and IE 2 were competitively insignificant because so few consumers used them. 6/3/99 pm Tr. at 47-48.

Absent some exclusionary effect, the inclusion of IE 1 and IE 2 in Windows—Microsoft’s first steps in integrating Web browsing functionality into the operating system—cannot violate Section 1, much less provide a basis for breaking up the company four years later. See Jefferson Parish Hosp. Dist. No. 2 v. Hyde, 466 U.S. 2, 21 n.34 (1984) ("The legality of petitioners’ conduct depends on its competitive consequences, not on whether it can be labeled ‘tying.’"). All that IE 1 and IE 2 establish is that, contrary to plaintiffs’ claim, distribution of Web browsing software in the operating system is not uniquely effective in garnering usage.

Plaintiffs contend that the district court properly applied the "conventional" Jefferson Parish test rather than the test articulated by this Court in its June 1998 decision. Pls. Br. at 98. Although the district court initially recognized that this Court "articulate[d] a framework for determining whether an integration amounts to a single product for purposes of evaluating a tying claim," United States v. Microsoft Corp., 1998-2 Trade Cas. (CCH) ¶ 72,261, at 82,675 (D.D.C. Sept. 14, 1998), plaintiffs argue that this Court’s decision provided no "direct guidance on how to apply Section 1 to tying arrangements," Pls. Br. at 101.

To begin with, the Supreme Court held in Jefferson Parish "that products are separate for tie-in purposes if there are separate markets for each product." Jack Walters & Sons Corp. v. Morton Bldg., Inc., 737 F.2d 698, 703 (7th Cir.) (emphasis added), cert. denied, 469 U.S. 1018 (1984); see also X Phillip E. Areeda et al., Antitrust Law ¶ 1745d2, at 211 (1996) ("[A] distinct market for the tied item does not imply separate products absent widespread sales of the tying item in unbundled form."). There is virtually no consumer demand for operating systems with no Web browsing functionality. Because Internet access is the leading reason consumers purchase personal computers today, all major operating systems now provide Web browsing functionality. Schmalensee ¶¶ 216-23; Allchin ¶¶ 261-77; see also Rose ¶¶ 19, 21.

More importantly, plaintiffs do not dispute that this Court previously rejected the DOJ’s assertion that Jefferson Parish’s "consumer-demand" test applies to integrated products. 147 F.3d at 946. This Court stated that applying the "consumer-demand" test to integrated products would stymie innovation by thwarting companies’ "legitimate desire to continue to integrate products that had been separate—and hence necessarily would have been provided in distinct markets." Id. at 953. Plaintiffs say nothing about the chilling effect that applying Jefferson Parish to integrated products would have on innovation in a wide range of industries. See MS Br. at 70. They do not deny that under the "consumer-demand" test, every addition of new functionality previously provided by a standalone product could be challenged as an unlawful tie, thus making the permissibility of such innovations a question for judges and juries.

Plaintiffs’ efforts to minimize the significance of this Court’s decision are unavailing. First, although this Court’s decision involved the "construction of a consent decree," Pls. Br. at 101, plaintiffs do not deny that this Court relied on antitrust cases and the Areeda treatise in fashioning its test. See 147 F.3d at 948-50. Second, plaintiffs argue that this Court’s decision has "only limited application here" because it centered on "the combinations of Windows 95 with IE3 and IE4," whereas this case also involves IE 1 and IE 2. Pls. Br. at 101. Plaintiffs do not dispute, however, that the focus of this case has always been on Windows 98, and that IE is even more deeply integrated into Windows 98 than it was into Windows 95. See 11/16/98 pm Tr. at 46. Third, plaintiffs emphasize that this Court’s decision was based on "a ‘limited record,’" Pls. Br. at 105 (quoting 147 F.3d at 950 n.15), but the applicable legal framework does not change depending on the extent of the record. Fourth, plaintiffs suggest that less deference should be given to product design decisions in the software industry because software "is a uniquely malleable product." Id. at 67. Yet this Court established its test fully cognizant that "[s]oftware code by its nature is susceptible to division and combination in a way that physical products are not." 147 F.3d at 951.

As a fallback, plaintiffs argue, as they did below, that Windows and IE are separate products under this Court’s test. Pls. Br. at 104-07. Had the district court agreed, it presumably would have so ruled. Plaintiffs come up short under this Court’s test in any event because they cannot show that the many benefits of Microsoft’s integrated design could be duplicated by combining an operating system "with a stand-alone browser such as Netscape’s Navigator." 147 F.3d at 950. In fact, plaintiffs do not even attempt to meet that standard.

Plaintiffs do not deny that Microsoft’s design of Windows to include Web browsing software produced significant, tangible benefits to users, ISVs and OEMs. See Allchin ¶¶ 105-41 & App. A. To the contrary, plaintiffs concede that the district court found that Microsoft’s inclusion of Web browsing functionality in Windows benefited consumers. Pls. Br. at 66, 103 (citing 84 F. Supp. 2d at 55, 110-11 (FF 186, 408)). They also accept that there are significant benefits to "using HTML rendering and other technologies used in browsing to provide non-browsing functions" in Windows 98, such as the user interface, the HTML Help system and the Windows Update feature. Id. at 66. And plaintiffs acknowledge that "IE expose[s] APIs" on which ISVs rely in developing applications for Windows, id. at 122, including applications that display information in HTML, 147 F.3d at 950-51. Rather than attempting to show that any of these benefits could be duplicated by combining an operating system with a standalone Web browser such as Navigator—as required by this Court’s test—plaintiffs proclaim with remarkable circularity that the benefits can be obtained by combining Windows with its IE components. Pls. Br. at 107; see also MS Br. at 76-77.

Plaintiffs also argue that "there was no reason why Microsoft could not provide a ‘browserless’ version of Windows." Pls. Br. at 103. They contend that "[a] ‘browserless Windows’ is Windows with no accessible browsing functionality, regardless of how that is accomplished." Id. at 104. According to plaintiffs, "[a]dding or removing the means of user access to a given function, by whatever means, amounts to adding or removing the software product," even if no software code is removed. Id. Thus, under plaintiffs’ view, IE would be "removed" from Windows even though all of the relevant software code remains in the operating system and continues to expose APIs to applications and provide important functionality to other parts of Windows. In other words, plaintiffs contend that Microsoft violated Section 1 because it failed to provide a version of Windows in which "user access" to some of its Web browsing functionality was hidden or disabled. Id. at 103. Plaintiffs cite no authority for the startling proposition that the antitrust laws required Microsoft to offer consumers a product with useful functionality hidden.

Plaintiffs’ focus on hiding "user access" is a departure from the position they took at trial. For instance, the DOJ alleged in its complaint that it is "technically feasible and practicable to remove Microsoft’s Internet browser software from Windows 98," but that "OEMs are prevented from doing so by Microsoft’s contractual tie-in." DOJ Compl. ¶ 20 (emphasis added). Similarly, plaintiffs’ technical expert, Edward Felten, insisted that his so-called "prototype removal program" showed "that Microsoft can deliver a version of Windows 98 from which the IE Web browser has been removed, and they can deliver that in a way which does not affect the non Web-browsing functionality of Windows 98." 6/10/99 am Tr. at 9 (emphasis added). By arguing that only "user access" to Web browsing functionality can or should be removed, plaintiffs now acknowledge that Microsoft’s so-called "Internet browser" should remain in Windows because it provides essential functionality to both applications and other parts of the operating system.

This Court has already stated that requiring Microsoft merely "to hide the allegedly tied product" by, for example, running the Add/Remove Programs utility, "suggests the oddity of treating as separate products functionalities that are integrated in the way Windows 95 and IE are." 147 F.3d at 952 n.18. As this Court previously observed, plaintiffs do not seek to "remove the IE software code, which indeed continues to play a role in providing non-browser functionality for Windows." Id. at 941. "In fact, browser functionality itself persists," and can be summoned up "by running any application (such as Quicken) that contains the code necessary to invoke the functionality." Id. (emphasis in original). As in the prior case, all plaintiffs seek is to "enable the OEMs to make user access to IE more difficult." Id.

Because little, if any, software code is removed in the process, eliminating "user access" to Web browsing functionality does nothing to alleviate the purported "collateral harm" that the district court found results from inclusion of IE in Windows. 84 F. Supp. 2d at 53, 111 (FF 173, 410). (This supposed "harm" simply reflects the engineering tradeoff inherent in adding functionality to Windows: it entails consumption of additional computer resources. See MS Revised Proposed Findings of Fact ¶¶ 594-95.) As plaintiffs’ own technical expert, Glenn Weadock, admitted, "merely removing the means of access to Web-browsing software does very little to address" concerns about "computer resource use (disk space, memory, CPU power, et cetera)." 11/16/98 pm Tr. at 23-24.

The absurdity of treating "user access" to Web browsing functionality as the tied product is also highlighted by the relief awarded. Although the district court’s breakup order assigns IE to the so-called "AppsCo," Pls. Br. at 120, plaintiffs still cannot identify the software code or files that they claim constitute IE, id. at 103-04. Presumably, the district court intended for "AppsCo" to receive something more than an icon, which is, after all, nothing but a small picture of a blue "e." And the restrictions on the ability of the so-called "OpsCo" to develop modified versions of IE are incomprehensible if there is no delineation of what software code or files constitute IE. See MS Comments on Pls. Revised Proposed Final J. at 7-8.

Plaintiffs’ arguments about hiding "user access" to functionality do nothing to undermine the fact that Microsoft has, once again, "clearly met the burden of ascribing facially plausible benefits to its integrated design as compared to an operating system combined with a stand-alone browser such as Netscape’s Navigator." 147 F.3d at 950. On this question—which is dispositive under this Court’s test—plaintiffs are silent.

Plaintiffs incorrectly assert that Microsoft disputes only the "separate products" element of their tying claim. Pls. Br. at 97. Microsoft clearly argued in its opening brief that plaintiffs’ tying claim fails for an additional reason: the alleged tie did not foreclose competition in the "Web browser" market. MS Br. at 65, 69, 80-83. Moreover, because IE is included in Windows at no charge, there is no forced purchase of a separate tied product, as required by tying law. Id. at 80-81 (citing cases).

As the Supreme Court recognized in Jefferson Parish, "every refusal to sell two products separately cannot be said to restrain competition." 466 U.S. at 11. Thus, the legality of an alleged tie-in "depends on its competitive consequences, not whether it can be labeled ‘tying.’" Id. at 21 n.34; accord S. Card & Novelty, Inc. v. Lawson Mardon Label, Inc., 138 F.3d 869, 874-75 (11th Cir. 1998); Smith Mach. Co. v. Hesston Corp., 878 F.2d 1290, 1295 (10th Cir. 1989), cert. denied, 493 U.S. 1073 (1990). If the "competitive consequences" of an arrangement are not those to which the tying rule is addressed, "then it should not be condemned irrespective of its label." Jefferson Parish, 466 U.S. at 21 n.34. As plaintiffs themselves acknowledge, "the prohibition on tying is concerned with foreclosure of competition on the merits in the tied product." Pls. Br. at 99. Microsoft’s inclusion of IE in Windows did not foreclose competition on the merits in developing and marketing Web browsing software. Unlike Jefferson Parish, where a surgical patient could be anesthetized by only one anesthesiologist, Windows users are free to install Navigator on their computers and use it instead of, or in addition to, IE.

Plaintiffs do not contest the testimony of Netscape’s CEO, James Barksdale, that Navigator is "perfectly interoperable" with Windows. 10/22/98 pm Tr. at 51-52. Nor do plaintiffs deny that OEMs are free to preinstall Navigator on their computers and that many OEMs in fact do so. Pls. Br. at 74. And plaintiffs do not dispute that a user who obtains a personal computer on which Navigator is not preinstalled can use IE (or other software) to download a free copy of Navigator from the Internet. As Microsoft explained, MS Br. at 81, there is no financial disincentive for such a user to switch to different Web browsing software, or to use both IE and Navigator, because both are available free. Chase ¶¶ 162-65. Even if a Windows user does not want to download Navigator from the Internet for some reason, he or she would have no difficulty obtaining a copy of Navigator from another source. Schmalensee ¶¶ 373-78. As Netscape has repeatedly boasted, Navigator is "ubiquitous." 10/21/98 pm Tr. at 14. In short, plaintiffs did not—and cannot—establish that Microsoft’s inclusion of IE in Windows prevented Netscape from distributing Navigator broadly and seeking to persuade consumers to use it. Cf. Roy B. Taylor Sales, Inc. v. Hollymatic Corp., 28 F.3d 1379, 1383 n.23 (5th Cir. 1994).

The low usage shares of IE 1 and IE 2 demonstrate that Microsoft’s inclusion of Web browsing software in Windows does not itself produce competitive success, much less foreclose competition on the merits in the putative "Web browser" market. See generally Schmalensee ¶ 524. Every copy of Windows 95 that was distributed through the OEM channel between August 1995 and August 1996 included either IE 1 or IE 2, Allchin ¶ 322, and OEMs were not permitted to modify Windows by removing the IE icon from the Windows desktop, Rose ¶¶ 25-31; Kempin ¶ 69. Although Microsoft shipped tens of millions of copies of Windows 95 during that period, "less than 5% of people accessing the Web were using Internet Explorer." Allchin ¶ 322. IE’s usage share continued to languish in single digits until Microsoft dramatically improved the quality of the Web browsing software included in Windows 95. See Schmalensee ¶¶ 288-89; Chase ¶¶ 135-36. Plaintiffs have never overcome the intuitively obvious fact that it was the improved quality of IE, not its inclusion in the operating system, that was principally responsible for Navigator’s declining usage share.

Plaintiffs admit that the principal basis for the district court’s monopoly maintenance ruling was its finding that Microsoft improperly foreclosed Netscape from the OEM and IAP channels of distribution. E.g., Pls. Br. at 79 (district court found that only the "OEM and IAP components" of Microsoft’s conduct independently violated Section 2). The district court held that Microsoft foreclosed the IAP channel by entering into promotion and distribution agreements with IAPs and foreclosed the OEM channel by tying IE to Windows (i.e., not permitting OEMs to modify the operating system to remove "user access" to Web browsing functionality). United States v. Microsoft Corp., 87 F. Supp. 2d 30, 39-42 (D.D.C. 2000).

The district court, however, rejected plaintiffs’ claim that Microsoft’s agreements with IAPs unreasonably restrained trade in violation of Section 1. And it is equally clear, as explained above, that Microsoft’s inclusion of Web browsing software in Windows did not amount to an unreasonable restraint of trade either. Plaintiffs cite no authority for their assertion that conduct that does not unreasonably restrain trade under Section 1—because it is not exclusionary—can nevertheless be "anticompetitive" under Section 2. Nor do they respond to Microsoft’s argument, based on settled Supreme Court case law, that "‘[c]oncerted activity subject to § 1 is judged more sternly than unilateral activity under § 2’ because ‘[c]oncerted activity inherently is fraught with anticompetitive risk.’" MS Br. at 97 (quoting Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752, 768-69 (1984) (emphasis added)). Stripped of its two principal pillars—foreclosure of the OEM and IAP channels—there is little left of plaintiffs’ monopoly maintenance claim.

Plaintiffs agree that "monopoly power is the ‘power to control prices or exclude competition.’" Pls. Br. at 51 (quoting Eastman Kodak Co. v. Image Technical Servs., Inc., 504 U.S. 451, 481 (1992)). Microsoft does not possess such power in a properly-defined product market.

The district court adopted the narrowest possible product market: "Intel-compatible PC operating systems." 87 F. Supp. 2d at 36. The district court thus excluded not only competitive operating systems such as Apple’s Mac OS, but also the two middleware technologies—Navigator and Java—that were the focus of plaintiffs’ monopoly maintenance claim and that were found to be the most serious competitive threats to Microsoft’s operating systems. 84 F. Supp. 2d at 28-30 (FF 68-77). Plaintiffs do not cite a single case in which the most serious competitive threats to an alleged monopoly product were excluded from the relevant product market. The antitrust laws do not countenance such a restrictive approach to market definition.

Although they state that Navigator and Java "threatened the Windows monopoly," Pls. Br. at 53, plaintiffs argue that the district court properly excluded middleware from the relevant product market because middleware is "a complement to, rather than a substitute for, an operating system," id. at 54. Plaintiffs acknowledge, however, that one of the principal functions of an operating system is to "serve[] as a platform for applications." Id. at 9. And they similarly argue that the platform characteristics of Web browsing software are a "critical dimension" of its value. Id. at 90. As a result, middleware "is not only a complementary product, but also is a potential substitute for a crucial function of the operating system: [it] could serve as a platform for software applications developers." AOL, CCIA, ProComp & SIIA Br. at 15. Indeed, that is the central tenet of plaintiffs’ monopoly maintenance claim.

Plaintiffs argue that the district court "properly excluded middleware from the market" because "no middleware product currently serves the same purposes as the Windows operating system." Pls. Br. at 53. The Sherman Act does not require, however, that "products be fungible to be considered in the relevant market." United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 394 (1956). If the law were otherwise, "only physically identical products would be part of the market," id., which would result in too narrow a market in which to measure an alleged monopolist’s power. See also United States v. Continental Can Co., 378 U.S. 441, 457 (1964) ("complete inter-industry competitive overlap need not be shown").

Courts have rejected efforts to exclude from the relevant product market partial substitutes that significantly constrain a defendant’s power over price. For example, in an antitrust action brought by Transamerica against IBM, the court rejected Transamerica’s attempt to limit the relevant product market to complete computer systems (i.e., the combination of hardware and software needed to satisfy a customer’s data processing requirements):

Transamerica’s definition of the systems market is, however, too narrow. Only suppliers of complete systems are included, suppliers of parts of systems are ignored. Firms which offer the user parts of systems have become a major competitive force in the computer industry, and, because they significantly constrain IBM’s power to control the price of the systems it sells, a market definition that ignores them is incorrect.

In re IBM Peripheral EDP Devices Antitrust Litig., 481 F. Supp. 965, 977-78 (N.D. Cal. 1979) [hereinafter Transamerica], aff’d sub nom. Transamerica Computer Co. v. IBM, 698 F.2d 1377 (9th Cir.), cert. denied, 464 U.S. 955 (1983). As the Transamerica court properly observed, "A market definition should ‘recognize competition where, in fact, competition exists,’ and should include all significant competition even though that competition differs in form or nature." Id. at 978 (footnotes omitted) (quoting Brown Shoe Co. v. United States, 370 U.S. 294, 326 (1962)). "By assuming that only manufacturers offering complete systems compete with IBM," the court stated, Transamerica ignores a "significant constraining force." Id. Plaintiffs ignore a significant constraining force on Microsoft by arguing that only complete operating systems (indeed, only operating systems that run on a particular type of microprocessor) compete with Windows.

The relevant market inquiry "boils down to whether there are products that restrain a defendant’s ability to act without regard for other manufacturers and suppliers." ILC Peripherals Leasing Corp. v. IBM, 458 F. Supp. 423, 428 (N.D. Cal. 1978), aff’d sub nom. Memorex Corp. v. IBM, 636 F.2d 1188 (9th Cir. 1980), cert. denied, 452 U.S. 972 (1981). Competing middleware technologies constrain Microsoft’s actions and thus must be included in the relevant product market. MS Br. at 84-88. As Microsoft has explained, the only sensible product market definition includes all platforms for developing and running applications. Id. at 87

Plaintiffs do not deny that they bore the burden of proving that there are significant barriers to entry into the putative market. See United States v. Baker Hughes Inc., 908 F.2d 981, 987-89 (D.C. Cir. 1990). Although plaintiffs now insist that they "alleged and proved other barriers," Pls. Br. at 57, the district court’s findings leave no doubt that it found only one barrier here—the applications barrier to entry, see 84 F. Supp. 2d at 19-23 (FF 34, 36-50).

The nature of the purported applications barrier to entry is clear from plaintiffs’ brief: consumers currently prefer Windows over competing platforms because it supports a large array of compatible applications (the so-called "network effect"). Pls. Br. at 57, 58; see also 84 F. Supp. 2d at 19-20, 23 (FF 37, 50). As plaintiffs acknowledge, a critical attribute of an operating system is its ability to run applications. Pls. Br. at 9. Windows is successful because Microsoft has excelled at developing innovative platform technologies, distributing those technologies broadly and persuading ISVs to create applications that utilize them. Maritz ¶¶ 124-86. Although plaintiffs quote this Court’s statement that "‘[a]ny market condition that makes entry more costly or time-consuming . . . should be considered a barrier to entry,’" Pls. Br. at 58 (quoting S. Pac. Communications, 740 F.2d at 1001), they fail to cite a single case in which consumer demand for an attribute of a product (here, support for a variety of applications) alone was found to constitute a barrier to entry.

Plaintiffs’ defense of the district court’s "empirical evidence" that supposedly supports the applications barrier to entry is equally unpersuasive. See 84 F. Supp. 2d at 22-23 (FF 45-50). Plaintiffs note that the district court rejected Microsoft’s assertion that "the Linux ‘open source’ OS has created a ‘serious platform threat to Windows,’" Pls. Br. at 58, but they do not dispute that Netscape was able to mount a "serious platform threat" to Windows within a year of its formation as a six-person startup, MS Br. at 94. In addition, plaintiffs do not deny that Linux, which was initially written by a Finnish graduate student in his spare time, was able to enter the putative market and now competes with Windows, particularly as a server operating system. 84 F. Supp. 2d at 23 (FF 50). Even if Linux has not yet made significant inroads as a desktop operating system, it "can nevertheless exert competitive pressure on that market." Baker Hughes, 908 F.2d at 988.

Lastly, plaintiffs do not even try to reconcile their assertion that ISVs have no incentive to write applications for platforms other than Windows with their proffered basis for breaking up Microsoft—i.e., their contention that after the breakup, "AppsCo" would have an incentive to port its applications "to other platforms, such as Linux or sub-PC-operating systems installed on personal digital assistants or wireless telephones." Pls. Br. at 121. If the only thing preventing Microsoft from "respond[ing] to ordinary market incentives to port its powerful and widely used applications" to other platforms such as Linux is "a stake in the Windows monopoly," id. (emphasis added), then the thousands of ISVs around the world with no such stake are presumably acting in accordance with those "ordinary market incentives" already.

Plaintiffs do not dispute that Microsoft—in their view a monopolist protected by a virtually insurmountable entry barrier—devotes billions of dollars annually to research and development, more than all its competitors. MS Br. at 89. They instead assert that "[a] monopolist has powerful incentives to invest in R&D to improve its product." Pls. Br. at 61. To the contrary, monopoly is "expected to result in technological stagnation. Without a competitive spur, the monopolist is thought not to have any incentive to innovate." Transamerica, 481 F. Supp. at 976-77.

Plaintiffs also do not dispute that the district court found that they failed to show that Microsoft charged a "monopoly price" for Windows. See 84 F. Supp. 2d at 27 (FF 65). Moreover, in the absence of any finding that Microsoft has control over productive capacity, plaintiffs cannot contend—and the district court did not find—that Microsoft can "raise price by reducing output," Eastman Kodak, 504 U.S. at 487 (Scalia, J., dissenting), the hallmark of monopoly power. See Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 233 (1993) ("Supracompetitive pricing entails a restriction in output.").

Microsoft invests heavily in research and development and keeps the price of Windows low because it must contend not only with numerous existing and potential platform competitors, but also with prior versions of Windows. Because software never wears out, Microsoft must continue to improve Windows and price it attractively to give existing customers an incentive to purchase a new computer or otherwise obtain a new version of the operating system. See MS Br. at 91. "If a dominant supplier acts consistent with a competitive market," as Microsoft does, "the purpose of the antitrust laws is amply served." United States v. Syufy Enters., 903 F.2d 659, 668-69 (9th Cir. 1990).

Plaintiffs place great emphasis on Microsoft’s "market share." Pls. Br. at 55-56. Given the nature of competition in the software industry, it is not surprising—and not at all indicative of "monopoly power" in the antitrust sense—that one firm has a large share of sales. 84 F. Supp. 2d at 25-26 (FF 59); Schmalensee ¶¶ 32-36, 63-73. Software firms compete for leadership in categories such as word processors, spreadsheets, personal finance software and operating systems, as well as in brand new categories. Schmalensee ¶ 32. If successful, software firms often obtain a high share of a particular category and, with it, high profits given the very low marginal cost of reproducing software. Schmalensee ¶ 32. As then Treasury Secretary Lawrence Summers recently explained, "the only incentive to produce anything" in the "new economy" is the prospect of earning the high profits that arise from risk taking and success in achieving market leadership "because without that power the price will be bid down to marginal cost and the high fixed costs cannot be recouped." Lawrence H. Summers, "The Wealth of Nations" (Remarks, May 10, 2000). This is not to say, however, that category leaders can cease innovating or that they can raise the price of their products without fear of being displaced. Schmalensee ¶¶ 32, 36. To the contrary, history shows that category leaders that fail to innovate or to price their products attractively are frequently displaced, oftentimes by so-called "niche players" or by technological advances that redefine an entire category or render it obsolete. Schmalensee ¶¶ 66, 71-73; Maritz ¶¶ 75-105.

Because of this dynamic competition, the rate of innovation in the software industry is remarkably high, and has been an engine of growth and increased productivity for the Nation’s economy. See MS Br. at 11-12. What the Transamerica court said about the computer industry in 1979 rings true about the software industry today:

This is not the story of a stagnant, dominated industry. There is no doubt that the pace of technological progress in the computer industry is extraordinary. Commentators are fond of saying that had the auto industry kept the same pace over the last 30 years, a Rolls Royce would cost $2.50 today and would have an EPA gas rating of 2,000,000 miles per gallon.

481 F. Supp. at 982. By giving undue weight to plaintiffs’ "market share" evidence and adopting a static view of the marketplace, the district court ignored competitive realities that require Microsoft to innovate and price its software attractively.

In arguing that Microsoft engaged in anticompetitive conduct in violation of Section 2, plaintiffs cite no cases involving conduct similar to the conduct at issue here. Rather, they rely on abstract definitions of "anticompetitive" or "exclusionary" conduct set out in cases such as Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585 (1985). See Pls. Br. at 47-48. Courts have cautioned, however, against reading cases like Aspen too broadly. E.g., Olympia Equip. Leasing Co. v. W. Union Tel. Co., 797 F.2d 370, 379 (7th Cir. 1986) (Posner, J.), cert. denied, 480 U.S. 934 (1987). Standing alone, the language in Aspen on which plaintiffs rely provides an insufficient basis to distinguish lawful from unlawful conduct—and insufficient guidance to businesspeople (and their counsel) attempting to conform their conduct to the dictates of the law. Plaintiffs fail to address any of the Section 2 cases that articulate controlling legal standards applicable to the specific kinds of conduct challenged in this case, even though such cases exist and were relied on by Microsoft. MS Br. at 101-02, 109-10 (citing cases involving technological tying and exclusive dealing claims under Section 2).

Microsoft’s conduct was not "anticompetitive" or "exclusionary" for one simple reason: nothing Microsoft did excluded Netscape—the focus of the case—from the marketplace. Although plaintiffs do not dispute this crucial fact, they attempt to sidestep it by claiming that "Microsoft took actions to exclude Navigator" from the OEM and IAP channels of distribution. Pls. Br. at 63. But, even assuming that Navigator was excluded from those two channels—which it was not—Netscape was able to distribute huge quantities of Navigator to consumers through numerous other channels. See MS Br. at 97. Unable to refute this fact, plaintiffs set up a straw man: they argue that Microsoft violated Section 2 even though "it did not completely foreclose Netscape’s distribution of Navigator." Pls. Br. at 68 (emphasis in original); see also id. at 42-43, 73-76, 78-79. Microsoft has never contended that Section 2 requires total exclusion from the marketplace. MS Br. at 97, 107-10. What is required is a showing of substantial exclusion, and that is what is missing from this case. The district court did not find that Microsoft excluded Netscape from any percentage of the marketplace, much less a substantial percentage. Instead, the district court expressly acknowledged that Netscape was able to distribute Navigator "to every PC user worldwide." 87 F. Supp. 2d at 53 (emphasis added).

Confronted with the district court’s finding that "Netscape was able to distribute 160 million copies of Navigator" in 1998 alone despite Microsoft’s supposed foreclosure of the OEM and IAP channels, id., plaintiffs argue that Microsoft’s conduct was anticompetitive because it supposedly prevented Navigator from being widely used. Pls. Br. at 31, 63, 75-76, 78. The antitrust laws, however, ensure competitors access to the marketplace; they do not guarantee that consumers will use their products. Consumers may choose not to use a particular product for a variety of reasons unrelated to a competitor’s access to distribution channels. If competitors can reach the ultimate consumers of their products by distributing directly to them, competing for the services of existing distributors or developing alternative distribution channels, the "[a]ntitrust laws require no more." Omega Envtl., Inc. v. Gilbarco, Inc., 127 F.3d 1157, 1163 (9th Cir. 1997), cert. denied, 525 U.S. 812 (1998); see also Stitt Spark Plug Co. v. Champion Spark Plug Co., 840 F.2d 1253, 1257-58 (5th Cir.), cert. denied, 488 U.S. 890 (1988); Ryko Mfg. Co. v. Eden Servs., 823 F.2d 1215, 1233-35 (8th Cir. 1987), cert. denied, 484 U.S. 1026 (1988); Roland Mach. Co. v. Dresser Indus., Inc., 749 F.2d 380, 393-95 (7th Cir. 1984).

In any event, Navigator is widely used. The district court found that Navigator’s installed base of users increased from 15 million in 1996 to 33 million in December 1998, 87 F. Supp. 2d at 53; 84 F. Supp. 2d at 103 (FF 378), the period in which Microsoft was supposedly engaged in a "multifaceted campaign of exclusionary conduct," Pls. Br. at 62. Netscape’s CEO, Barksdale, testified to even higher figures, estimating in October 1998 that there were "somewhere between 40 and 70 million" Navigator users. 10/21/98 pm Tr. at 13. There is also no truth to plaintiffs’ suggestion that the OEM and IAP channels are the only channels effective in generating usage of Web browsing software. Pls. Br. at 15, 63. Netscape informed AOL during AOL’s due diligence investigation that "currently about 50% of our installed base" of users obtained Navigator by downloading it from the Internet. DX 2538 at 36457; see also DX 2810 at 79. As Netscape’s co-founder, James Clark, testified, the Internet is the most efficient distribution system for software in the world. DX 2562 at 41-42. Plaintiffs ignore this evidence.

Leaving aside plaintiffs’ failure to show that Netscape was excluded from any portion of the marketplace—which by itself is fatal to their Section 2 claims—plaintiffs’ specific allegations of anticompetitive conduct cannot withstand scrutiny.

Plaintiffs assert that Microsoft "effectively exiled Netscape from the OEM channel." Pls. Br. at 62. First, they challenge the manner in which Microsoft included Web browsing software in Windows 98, id. at 66, claiming that Microsoft is guilty of "predatory product design," id. at 67. Second, plaintiffs contend that Microsoft’s OEM license agreements improperly prohibited "OEMs from modifying Windows in any unauthorized way." Id. at 68. Third, plaintiffs contend that Microsoft "offered OEMs valuable incentives and discounts" to "limit distribution of Navigator." Id. at 17-18. None of these contentions has merit.

Predatory Product Design. Plaintiffs concede that "Microsoft’s conduct in developing a Web browser and offering it to OEMs and users with Windows [was] lawful." Id. at 66. They also concede that Microsoft’s provision of Web browsing software with Windows benefited consumers. Id. Recognizing that courts generally refuse to second-guess engineers’ product design decisions in antitrust cases, e.g., ILC Peripherals, 458 F. Supp. at 439, plaintiffs now insist that they are challenging only "very limited aspects of Microsoft’s product design," Pls. Br. at 66, an assertion that cannot be reconciled with their request that Microsoft be broken up and subjected to wide-ranging product design regulations.

Plaintiffs’ "predatory product design" claim fails for the same reasons that their "technological tying" claim fails under Section 1. Plaintiffs cite no authority for their suggestion that a different, more stringent standard should apply to product design decisions under Section 2 than applies under Section 1. To the contrary, one of the cases plaintiffs cite, H.J., Inc. v. ITT, 867 F.2d 1531, 1542 (8th Cir. 1989), applied the Section 1 standard to a tying claim under Section 2. Nor do plaintiffs articulate a standard that could be applied in the future to distinguish beneficial innovations from "predatory" design changes or enable software developers to determine what functionality may be added to leading products and whether a mechanism must be provided to remove "user access" to that functionality. See Pls. Br. at 66. Lastly, plaintiffs do not deny that there are technical reasons why Windows 98 invokes its IE components by default in a handful of instances. See MS Br. at 82-83; Pls. Br. at 108-09. Plaintiffs thus provide no basis to outlaw the design of Windows 98, even if courts are permitted under the Sherman Act to evaluate product design decisions under some vague and amorphous standard.

The law is settled that design changes that improve a product cannot violate Section 2, regardless of the defendant’s intent. E.g., Transamerica, 481 F. Supp. at 1005 ("Mallard was a design adopted primarily to preclude PCM competition, but it was a superior design, and its effect on competition was negligible. A finding adverse to IBM on this aspect of its conduct would amount to punishment for intent alone."). To hold otherwise would chill technical innovation. That is why courts have roundly rejected allegations of "predatory innovation" or "predatory design." See Joseph Gregory Sidak, Debunking Predatory Innovation, 83 Colum. L. Rev. 1121, 1121 n.2 (1983) (citing cases). Plaintiffs do not dispute that Microsoft’s design of Windows to include Web browsing functionality improved the operating system and benefited consumers. Pls. Br. at 66. Plaintiffs contend that Microsoft lacked "a legitimate business justification" for its design of Windows, id. at 67, but it is plaintiffs’ suggestion—that Microsoft was required to create a mechanism for hiding "user access" to useful functionality in the operating system—that makes no business sense.

Microsoft’s OEM License Agreements. Plaintiffs do not dispute that "Microsoft’s license agreements have never prohibited OEMs from pre-installing programs, including Navigator, on their PCs and placing icons and entries for those programs on the Windows desktop and in the ‘Start’ menu." 84 F. Supp. 2d at 63 (FF 217). Nor do they dispute that Microsoft’s license agreements do "not forbid the OEM to set Navigator as the default browsing software." Id. Plaintiffs instead claim that Microsoft’s license agreements are unlawful because they do not permit OEMs to delete or modify parts of Windows without authorization. Pls. Br. at 16-17. In particular, plaintiffs challenge the so-called "Windows Experience" provisions of Microsoft’s license agreements, which require that "the very first time a consumer turns on his or her computer, Microsoft’s Windows operating system . . . be allowed to go through the initial startup sequence as designed, developed and tested by Microsoft and to display the Windows desktop without any aspect of that screen having been deleted by the OEM." Kempin ¶ 38.

a. Plaintiffs concede that Microsoft’s inclusion of Web browsing functionality in Windows did not violate the antitrust laws. Pls. Br. at 102. They nevertheless argue that Microsoft was required to permit OEMs, which act as Microsoft’s distributors, to delete "user access" to that functionality. By claiming that Microsoft must permit OEMs to make unauthorized modifications to its copyrighted operating systems, plaintiffs seek to deprive Microsoft of its rights under federal copyright law.

Contractual provisions limiting the ability of a licensee to modify copyrighted works without the licensor’s permission are common in the software industry, and are found in the OEM license agreements of other leading operating system vendors. Kempin ¶ 25; see, e.g., DX 1661 at 4; DX 1777 at 102. From an antitrust perspective, Microsoft’s possession of this legal right contributes to its "efficiency in the same way that possession of any other asset does." Neumann v. Reinforced Earth Co., 786 F.2d 424, 427 n.3 (D.C. Cir. 1986). As this Court has explained, "the legitimate assertion of a legal right against a competitor, however damaging to the competitor, or even to competition, that may be, does not constitute monopolizing or attempting to monopolize within the meaning of the Sherman Act." Id.

According to plaintiffs, "Microsoft neither articulated a tenable copyright theory that supports its asserted defense, nor offered evidence to support such a theory." Pls. Br. at 69. That assertion is demonstrably false. From the outset of the case, Microsoft argued that plaintiffs’ "challenges to Microsoft’s license agreements with computer manufacturers are without legal merit by reason of Microsoft’s rights under the federal copyright laws." Answer to DOJ’s Compl. at 34. As Microsoft argued in seeking summary judgment, "[t]he ‘bundle of rights’ possessed by the holder of a valid copyright includes the right to prevent unauthorized modifications to its copyrighted work." MS Mem. in Supp. of Mot. for Summ. J. at 54. At trial, Microsoft introduced into evidence Certificates of Registration from the United States Copyright Office for both Windows 95 and Windows 98. 2/24/99 am Tr. at 63-66; DX 813; DX 814; see also 84 F. Supp. 2d at 66 (FF 228); Kempin ¶ 2; 2/25/99 am Tr. at 57-58. The admission of these copyright registrations "shift[ed] to [plaintiffs] the burden of proving the invalidity of the copyright[s]." Fonar Corp. v. Domenick, 105 F.3d 99, 104 (2d Cir.), cert. denied, 522 U.S. 908 (1997). Plaintiffs never contended—much less proved—that Microsoft’s copyrights are invalid.

Rather, plaintiffs argue that copyright law permits a licensee to make "minor modifications" to a copyrighted work without authorization from the copyright holder. See Pls. Br. at 69-72. According to plaintiffs, "deleting a single icon from Windows 95" and "providing users with additional information in the boot sequence of Windows" are "minor modifications [that] would hardly frustrate the objectives of the Copyright Act." Id. at 72. Leaving aside the irony that plaintiffs contend Microsoft violated Section 2 and should be broken up because it did not permit OEMs to make what they now call "minor modifications" to Windows, plaintiffs are wrong on both the facts and the law. The Second Circuit’s reference in Gilliam to "mutilation of [the] work" (quoted at Pls. Br. at 70) was not contained in the discussion of the copyright infringement claim, but rather in the discussion of the separate claim under the Lanham Act. Gilliam v. ABC, 538 F.2d 14, 23-24 (2d Cir. 1976). And, contrary to plaintiffs’ assertion, Pls. Br. at 70, the Seventh Circuit in WGN did not find that the "teletext" included by WGN in the vertical blanking interval of its 9:00 p.m. news broadcast was a "substantial portion of the work." WGN Cont’l Broad. Co. v. United Video, Inc., 693 F.2d 622, 624-28 (7th Cir. 1982).

Nor can the modifications at issue be dismissed as inconsequential. E.g., Kempin ¶ 37. As plaintiffs acknowledge, "[b]etween 1995 and 1999, Microsoft spent more than $100 million each year" developing the IE technologies in Windows. Pls. Br. at 15. Having made that substantial investment, Microsoft had a legitimate business interest in preventing OEMs from modifying Windows to deny "user access" to the functionality provided by IE. In addition, "Microsoft advertises Windows as including Internet Explorer, and many product reviews of Windows 98 discuss that feature of the operating system." Kempin ¶ 69. Plaintiffs do not deny that "considerable customer confusion would result and Microsoft’s brand image could suffer if customers purchasing a new computer preinstalled with Windows could not find and utilize the Internet Explorer functionality that Microsoft advertises as part of the operating system." Kempin ¶ 69.

b. Wholly apart from Microsoft’s rights under federal copyright law, the challenged provisions of Microsoft’s license agreements do not limit Netscape’s access to the OEM channel. See DX 10 at 48339. Plaintiffs do not dispute that "OEMs were permitted to pre-install Navigator on PCs and that many OEMs did so." Pls. Br. at 74. In fact, plaintiffs’ own exhibits show that Compaq, Fujitsu, Gateway, Hewlett-Packard, IBM and Sony—six of the world’s largest OEMs—distributed Navigator on some of their personal computers. GX 421 at 680; GX 2116 at 201070-71; see also Kempin ¶ 21. Rather, plaintiffs contend that, as the quality of IE improved, many OEMs decided either not to preinstall Navigator or to feature it less prominently on their personal computers because Windows already included IE. See Pls. Br. at 20, 74-75. Even if true, that contention does not establish an antitrust violation.

To support their claim that Navigator was "exiled" from the OEM channel, plaintiffs challenge the authenticity of the document prepared during AOL’s due diligence investigation of Netscape that reports that Navigator was present on "22% of OEM shipments." DX 2440 at 341778. Plaintiffs assert that "[n]othing in the record explains the source or method of the ‘[e]stimate’ in this anonymous document, which was not subject to cross-examination." Pls. Br. at 75; see also id. at 78. To the contrary, Barry Schuler, a senior AOL executive, testified that this so-called "anonymous document" represents a "summary of the findings and proposed financials and operating plans" resulting from AOL’s November 1998 due diligence investigation of Netscape. DX 2810 at 74-75. Schuler also stated that AOL used this document to "make our final determination of whether to proceed ahead with" the $10 billion acquisition of Netscape. DX 2810 at 75. Plaintiffs themselves identified the document in their proposed findings of fact as "the due diligence report by Goldman Sachs on the Netscape/AOL transaction," Pls. Joint Proposed Findings of Fact ¶ 380.5.i, and the district court credited the statement in the document that Netscape distributed 160 million copies of Navigator in 1998, 87 F. Supp. 2d at 53.

No Restriction on Distribution of Navigator. Plaintiffs suggest that Microsoft’s OEM license agreements "restricted Navigator’s distribution." Pls. Br. at 73. That suggestion is contradicted by the facts, as the district court found. 84 F. Supp. 2d at 63 (FF 217). Plaintiffs also suggest that Microsoft dissuaded Compaq, Gateway and IBM from distributing Navigator. Pls. Br. at 18. In fact, plaintiffs go so far as to state that Microsoft had an exclusive contract with Compaq requiring it to ship only IE. Id. at 77. These statements are likewise false. The undisputed evidence establishes that Compaq, Gateway and IBM all distribute Navigator. Kempin ¶ 21; 2/19/99 am Tr. at 49-50; 6/9/99 pm Tr. at 63. Furthermore, Compaq’s John Rose and Stephen Decker both testified that Microsoft never limited in any way Compaq’s ability to preinstall Navigator on its new computers. Rose ¶¶ 32-33; DX 2564 at 44, 45, 50.

Plaintiffs contend that Microsoft’s agreements with ISPs and OLSs (collectively referred to by the district court as "IAPs") improperly excluded Navigator from an important distribution channel in violation of Section 2. They concede, however, that the district court rejected their claim that those same agreements violated Section 1. Pls. Br. at 5. They also do not dispute that courts apply exactly the same standard to alleged exclusive dealing agreements under Section 2 as they do under Section 1. MS Br. at 109-10 (citing cases).

In an attempt to salvage their monopoly maintenance claim, plaintiffs argue that the district court’s "Section 1 ruling addressed the effects of Microsoft’s conduct in the Web browser market, but the Section 2 monopoly maintenance violation concerns the operating system market." Pls. Br. at 77-78 (emphasis in original). This is pure sophistry. Under plaintiffs’ theory of the case—adopted by the district court—the challenged conduct had an anticompetitive effect in the alleged operating system market because it allegedly prevented Navigator from becoming the "standard" Web browsing software in a separate "Web browser" market. Id. at 14. Thus, accepting plaintiffs’ theory, if the challenged agreements did not unreasonably retrain trade in the alleged "Web browser" market—as the district court concluded—they could not possibly have had a derivative anticompetitive effect in the alleged operating system market.

Plaintiffs also take issue with the fact that, notwithstanding Microsoft’s agreements with IAPs, Netscape was able to distribute large quantities of Navigator through the IAP channel. Once again, they seek to dismiss as an "anonymous document" (Pls. Br. at 78) the same report summarizing AOL’s due diligence investigation of Netscape, which estimated that Navigator accounted for a "24% share of top 20 ISP’s distribution." DX 2440 at 341778. Plaintiffs argue that this document "establish[es] only that, after Microsoft dropped some of its restrictions in the spring of 1998, Navigator was not totally excluded from the IAP channel." Pls. Br. at 78. Contrary to plaintiffs’ assertion, Navigator was never excluded from the IAP channel. At a time when Microsoft was struggling to obtain "measurable browser share," DX 1545 at 2006380, Microsoft entered into the challenged agreements with only ten of the more than 4,500 ISPs in North America and with four leading OLSs, 84 F. Supp. 2d at 72-73, 85 (FF 256, 305); Myhrvold ¶ 8. The agreements expressly permitted the IAPs to distribute Navigator to their subscribers upon request, MS Br. at 34, 35, and Navigator’s share of the subscriber base of IAPs did not vary significantly depending on whether the IAP had entered into an agreement with Microsoft, Schmalensee ¶¶ 445-62 & App. D ¶¶ 100-14. To the extent Microsoft’s ISP agreements could be said to have had any adverse effect on Navigator’s distribution, that effect disappeared when Microsoft waived the distribution provisions of those agreements in April 1998, approximately 18 months after they went into effect. Myhrvold ¶ 91.

Plaintiffs also premise their monopoly maintenance claim on (i) Microsoft’s agreements with certain ICPs and ISVs, (ii) aspects of Microsoft’s Java implementation, and (iii) Microsoft’s 1997 technology agreement with Apple. See Pls. Br. at 42, 65, 79-83. They concede, however, that the district court concluded that these matters "failed to have a ‘significant exclusionary impact,’" and thus cannot "be independent bases for liability" under Section 2. Id. at 81 (emphasis in original). Plaintiffs nevertheless argue that "a series of actions, which standing alone would not be unlawful, can . . ., in combination, result in a violation of the Sherman Act." Id. at 82. The Federal Circuit recently rejected a similar "monopoly broth" theory in Intergraph Corp. v. Intel Corp., 195 F.3d 1346, 1366-67 (Fed. Cir. 1999), a case plaintiffs ignore. Plaintiffs not only are wrong on the law, see MS Br. at 112-14; they also have not shown that any of the challenged acts were anticompetitive.

Microsoft’s Agreements with Certain ICPs and ISVs. The district court expressly found that plaintiffs failed to prove that Microsoft’s agreements with the 24 ICPs included in the Channel Bar "actually had a substantial, deleterious impact on Navigator’s usage share." 84 F. Supp. 2d at 92 (FF 332). That finding defeats any claim that those agreements had antitrust significance, under Section 1 or Section 2.

Similarly, plaintiffs’ assertion that Microsoft "contractually required ISVs to use Internet Explorer-specific technologies in return for timely and commercially necessary technical information about Windows" and "precluded important ISVs from distributing Navigator with their products" is false and not supported by any testimony. Pls. Br. at 30 (citing GX 2071 & GX 2400). Microsoft entered into the agreements in question—the so-called "First Wave" agreements—with only a fraction of the thousands of ISVs that write applications for Windows. 84 F. Supp. 2d at 93 (FF 339). Pursuant to those agreements, the ISVs received certain technical and marketing support from Microsoft in exchange for their commitment to, among other things, use IE 4 as "the default browser" if their application’s user interface was "HTML based." GX 2071 at 3. The "First Wave" agreements did not preclude ISVs from distributing Navigator with their products, and there is no finding or evidence that those agreements had any adverse effect on Navigator’s distribution, much less a substantial one.

Microsoft’s Java Implementation. Plaintiffs do not dispute that Microsoft’s Java Virtual Machine ("JVM") permits cross-platform Java applications to run on Windows. Pls. Br. at 80. Nor do they deny that Microsoft’s Java development tools, Visual J++ 6.0, enable ISVs to write cross-platform Java applications. Id. Plaintiffs nevertheless insist that Microsoft improperly obstructed the development of cross-platform Java applications. See id. at 35-38, 79-80. In particular, they claim that Microsoft designed Visual J++ 6.0 so that ISVs would "unwittingly" write Java applications that run only on Windows, id. at 80, and that Microsoft’s mechanism for allowing Java applications to make native calls to Windows APIs, called J/Direct, created an "intentional incompatibility" with Sun’s mechanism, called JNI, id. at 67.

As to the first claim, plaintiffs cite no evidence—or finding by the district court—that any ISV that wanted to write a cross-platform Java application ever unintentionally wrote one that ran only on Windows by mistakenly using the optional keywords or compiler directives in Visual J++ 6.0. See MS Br. at 54-55. As to the second claim, a Java application that makes native calls to Windows (using J/Direct or Sun’s JNI) by definition will not run on any other operating system, so the differences between the two have no bearing on the creation of cross-platform Java applications, which is what Microsoft allegedly sought to obstruct. For ISVs that chose to write Java applications that rely on Windows system services, Microsoft’s creation of J/Direct was plainly beneficial because, as plaintiffs admit, Java applications utilizing J/Direct "ran faster than applications written to use the Sun-compliant methods of access [i.e., JNI]." Pls. Br. at 37. Building a better mousetrap cannot be condemned because it is "incompatible" with an alternative, technically inferior product supplied by a competitor.

Plaintiffs also wrongly assert that Microsoft "threatened" or "pressured" Intel into not supporting cross-platform Java. Id. at 36-37, 80. Microsoft simply advised Intel that, in Microsoft’s view, supporting Sun’s Java strategy was contrary to Intel’s own interests. See MS Br. at 55. In the past, Intel has offered Microsoft similar unsolicited advice on a range of subjects. E.g., DX 985 at 173567. There is no evidence that either Intel or Microsoft has ever felt compelled to follow such advice.

Microsoft’s Agreements with Apple. Attempting to rely on alleged conduct outside the market defined by the district court to support their monopoly maintenance claim, plaintiffs contend that Microsoft induced Apple to "remove Navigator from the default installation of the Mac OS 8.5" by threatening to cancel Mac Office 97. Pls. Br. at 29. For one thing, plaintiffs ignore the important role that Apple’s threat of a $1.255 billion patent infringement action played in discussions leading up to the companies’ three August 1997 agreements, see MS Br. at 59-61, as well as the testimony of Apple’s Avie Tevanian that the patent cross-license, Microsoft’s $150 million investment and the technology agreement were really one "overall agreement," 11/4/98 pm Tr. at 54-55; see also Tevanian ¶¶ 28-43; GX 583; GX 584; GX 1167. For another, the technology agreement does not require Apple to exclude Navigator from the default installation of the Mac OS, but rather expressly provides that "Apple may bundle browsers other than Internet Explorer with . . . Mac OS system software releases." GX 1167 at 44. Apple was free under the agreement to include Navigator "inside a folder" on the Mac OS desktop like the OLS folder on the Windows desktop, Tevanian ¶ 38, and Apple continued to bundle Navigator with the Mac OS after the agreement was executed, Tevanian ¶ 24; 11/4/98 pm Tr. at 56-57.

Plaintiffs concede that they were required to show that the challenged conduct had the "requisite impact" on the alleged market for Intel-compatible PC operating systems to make out a monopoly maintenance claim. See Pls. Br. at 82-83. They have not done so. The district court expressly found that "[t]here is insufficient evidence to find that, absent Microsoft’s actions, Navigator and Java already would have ignited genuine competition in the market for Intel-compatible PC operating systems." 84 F. Supp. 2d at 112 (FF 411). Plaintiffs are silent on this finding, which is fatal to their claim.

Plaintiffs’ monopoly maintenance claim is unusual in that they do not claim that Microsoft’s conduct adversely affected "existing competitors" in the so-called "operating system market." Pls. Br. at 83. As plaintiffs acknowledge, "That was not Microsoft’s goal." Id. They instead contend that Microsoft improperly maintained a monopoly in operating systems by foreclosing distribution of Navigator and Java, which are not part of the product market defined by the district court. In fact, plaintiffs themselves assert that "[s]uch middleware is . . . a complement to, rather than a substitute for, an operating system." Id. at 54. Plaintiffs failed to prove that Microsoft’s allegedly unlawful conduct directed at these "complements" contributed significantly to the maintenance of Microsoft’s purported operating system monopoly. See MS Br. at 115-18. That failure of proof is particularly acute if Microsoft’s inclusion of IE in Windows was lawful, as plaintiffs now concede. Pls. Br. at 66, 104 (challenging only Microsoft’s refusal to permit OEMs to remove "user access" to some IE functionality).

Plaintiffs seek to derive evidence of causation from a May 1995 Microsoft document stating that Netscape’s strategy is "to commoditize the underlying operating system" by moving "the key APIs" into Navigator. Id. at 13, 85 (citing GX 20 at 112876.3). That document does not predict, however, that such a strategy by Netscape, even if successfully executed, would spur entry into the alleged market for "Intel-compatible PC operating systems" (as opposed to depriving those operating systems of their principal value as platforms for applications). See 1/13/99 Tr. at 503. Plaintiffs presented no testimony from other operating system vendors that Navigator had the potential to create greater competition among operating systems. Nor did plaintiffs ever explain why any rational company would invest the resources required to develop a new PC operating system in order to participate in a "commodity" business. As Microsoft’s economist, Richard Schmalensee, explained, there is normally "less entry" in a "commodity" business "because there is less to be won from entry into such a business." 6/21/99 am Tr. at 55.

In responding to Microsoft’s description of the speculative chain of causation on which their monopoly maintenance claim is premised, MS Br. at 117, plaintiffs cast further doubt on whether Microsoft did anything to "hobble" middleware threats to Windows. See Pls. Br. at 85-87. Plaintiffs seek to dismiss as "immaterial" Barksdale’s concession that Netscape never saw Navigator as a complete substitute for Windows. Id. at 86. But Barksdale went much farther than that, declaring that Netscape "never maintained in any serious way" that Navigator was a substitute even for the "platform characteristics of Windows." 10/20/98 pm Tr. at 73. Navigator thus falls out of the equation as anything other than a distribution vehicle for Java. But see Muglia ¶ 82 (JVM included in Navigator for Windows did not support JNI until early 1998).

With regard to Java, plaintiffs argue that if ISVs wrote hybrid applications that call on APIs exposed by the Java class libraries as well as APIs exposed by Windows, it might be less costly to port those applications to other operating systems. Pls. Br. at 86. Even if that were true, there is no evidence—much less a finding by the district court—that Microsoft prevented ISVs from writing such hybrid Java applications. In fact, by creating a well-regarded Java implementation for Windows and distributing a high performance JVM with every copy of Windows, Schmalensee ¶ 148; DX 53 at 2, Microsoft actually increased the stock of applications that rely in whole or in part "on APIs exposed by the Java class libraries," 84 F. Supp. 2d at 29 (FF 74). On plaintiffs’ view that such hybrid Java applications are easier to port to other operating systems than applications written in programming languages such as C++ that rely exclusively on Windows APIs, Pls. Br. at 86, Microsoft’s actions actually lowered the so-called applications barrier to entry, which runs directly counter to plaintiffs’ causation argument.

Although plaintiffs assert that "new firms would have greater incentives to enter the market" for Intel-compatible PC operating systems if more applications were written to cross-platform middleware like Java, id. at 86-87, they provide no support for that supposition. In short, plaintiffs’ theory of causation remains in the realm of speculation, with no findings by the district court to support it. Plaintiffs insist that Navigator and Java were not competitors in the PC operating system market, and they have not shown that Navigator or Java would have created any additional competition in that putative market.

Plaintiffs defend the district court’s attempted monopolization ruling on two grounds. First, plaintiffs argue that Microsoft’s alleged proposal to Netscape in June 1995 to "divide up the market" alone constituted attempted monopolization. Pls. Br. at 87. Second, they contend that the "anticompetitive acts engaged in by Microsoft . . . after Netscape refused that offer [i.e., the very same acts that underlie plaintiffs’ monopoly maintenance claim] also constituted attempted monopolization of the Web browser market." Id. Plaintiffs are wrong on both counts.

Plaintiff’s contention that Microsoft’s June 1995 discussions with Netscape constituted attempted monopolization is based entirely on American Airlines, a decision the Fifth Circuit carefully limited to its unique facts. 743 F.2d at 1119 ("We see [the] alleged conduct as uniquely unequivocal and its potential, given the alleged market conditions, as being uniquely consequential."). Indeed, American Airlines is the only case in which a court has found that making a proposal that is then rejected by the other party can satisfy the anticompetitive conduct, specific intent and dangerous probability elements of attempted monopolization. See Pls. Br. at 88-93. Even accepting plaintiffs’ version of events, Microsoft’s efforts in June 1995 to persuade Netscape to utilize Internet-related functionality in Windows 95 are a far cry from the brazen price-fixing proposal ("Raise your goddamn fares twenty percent. I’ll raise mine the next morning.") in American Airlines. See MS Br. at 122-23. To proscribe the sorts of discussions Microsoft had with Netscape would cast a shadow on the legality of a wide range of necessary and important discussions between companies that develop complementary products.

The discussions between Microsoft and Netscape were not "uniquely unequivocal," and they did not relate to control of price, the hallmark of monopoly power. Although plaintiffs assert that Microsoft proposed to "divide up the market," Pls. Br. at 87, that simplistic assertion cannot be squared with the district court’s findings, see 84 F. Supp. 2d at 31-32 (FF 83-86). In addition, plaintiffs’ description of Microsoft’s alleged proposal has been a moving target. Netscape’s Barksdale originally testified that Microsoft "offered to allow us to continue to develop browsers for other operating systems, as long as we did not try to compete with them in developing a browser for the Windows 95 platform." Barksdale ¶ 110 (emphasis added). Plaintiffs now assert, however, that Microsoft "proposed that Netscape leave platform-level browsing technology for Windows 95 to Microsoft, in exchange for Microsoft’s leaving the browser business for other operating systems to Netscape." Pls. Br. at 63 (emphasis added). And the documents on which plaintiffs rely (id. at 13, 93) describe a completely different proposal. GX 22 at 167106 ("[T]he concept is that for 24 months they agree to do certain things in the client and we agree to help make their server business successful."); GX 540 at 10340 ("They may be willing to give us a client advantage, if we give them the server advantage.").

Nor can it be said that the potential impact of the June 1995 discussions was "uniquely consequential." The Fifth Circuit stressed in American Airlines that the participants in the discussion were the CEOs of the two airlines, which "jointly had a high market share [76%] in a market with high barriers to entry [FAA limitations on the number of arrivals at the Dallas-Fort Worth airport]." 743 F.2d at 1118. Here, by contrast, the Microsoft participants in the discussions with Netscape were not senior executives. Rosen ¶¶ 68-75. More importantly, the alleged market—Web browsing software—was a nascent business with many different competitors, including major companies like Apple, IBM and Sun. See Chase ¶¶ 13, 15. At the time, Microsoft had a zero share of the putative market, and the district court did not find that there are "high barriers to entry" into the "Web browser" market. (Netscape was incorporated in April 1994 and shipped its first commercial product in December 1994. Allchin ¶ 205; DX 68 at 69214.) Thus, unlike American Airlines, "at the moment of acceptance," Microsoft would not have "acquired monopoly power" over a relevant product market, i.e., the power to control prices or exclude competition. 743 F.2d at 1118.

Plaintiffs also ignore Barksdale’s admission that Microsoft "didn’t ask us to stop that moment from building our product, since we already announced it and had it in beta." 10/27/98 pm Tr. at 70. On cross-examination, Barksdale retreated from his earlier assertion that Microsoft insisted that Netscape "not develop our own browser for Windows 95," Barksdale ¶ 110, contending instead that Microsoft proposed that Netscape "gradually move back out of the Windows 95 business" over "the next year or two," 10/27/98 pm Tr. at 70. In fact, Barksdale admitted that Microsoft was eager for Netscape to release a version of Navigator for Windows 95:

[T]hey wouldn’t have asked us to get out of it immediately. We were going to release product in about a month. It was going to be great for Windows 95. It was going to help them sell a lot more because the Navigator was a hot application that ran on top of Windows 95. It was the killer app of 1995 . . . . They liked it because it helps sell operating systems.

10/27/98 pm Tr. at 71. This testimony shows that that the June 1995 discussions were neither "uniquely unequivocal" nor "uniquely consequential."

Lastly, Microsoft’s request that Netscape utilize Microsoft’s Internet-related platform technologies in subsequent versions of Navigator for Windows 95 would have had no immediate consequence. Most of the Internet-related APIs that Microsoft wanted Netscape to utilize did not ultimately become available until the August 1996 release of IE 3. See MS Br. at 31. As a result, even if Netscape had accepted on the spot Microsoft’s suggestion that it utilize some or all of those APIs, any effect on competition would have been more than a year into the future. Of course, one can only speculate about whether competition actually would have been affected and, if so, what that effect might have been.

Plaintiffs contend that Microsoft’s "post-June 1995" conduct also constituted an attempt to monopolize the "Web browser" market. Pls. Br. at 43. Because it is based on the same alleged conduct as their monopoly maintenance claim, see id. at 93-94, plaintiffs’ "post-June 1995" attempted monopolization claim fails for the same reason: the conduct was not exclusionary. In addition, plaintiffs have not shown that Microsoft acted with a "specific intent" to obtain monopoly power in Web browsing software or that there was a dangerous probability that Microsoft would achieve such power. Both of these failures are likewise fatal to plaintiffs’ "post-June 1995" attempted monopolization claim.

"[S]pecific intent in this context refers to a purpose to acquire monopoly power by driving one’s rival from the market by exclusionary or predatory means." Ass’n for Intercollegiate Athletics for Women v. NCAA, 735 F.2d 577, 585 (D.C. Cir. 1984). Plaintiffs have never argued that Microsoft intended to drive Netscape from the market. Under their theory of the case, Microsoft’s goal was much more modest: to prevent Navigator from becoming the "standard" Web browsing software so that ISVs "would continue to focus their efforts on the Windows platform." Pls. Br. at 14; see also 84 F. Supp. 2d at 43 (FF 133); GX 710 at 203013. As plaintiffs explained in their proposed findings of fact:

Preventing the browser threat from materializing . . . did not require Microsoft to eliminate other browsers entirely—or even to monopolize the browser market. Rather, Microsoft needed only to prevent any one browser rival from obtaining a large market share.

Pls. Joint Proposed Findings of Fact ¶ 359.3; see also GX 1372 at 111. Seeking to prevent a rival from obtaining a large market share is not a specific intent to acquire monopoly power.

Plaintiffs point out that the district court "expressly found that Microsoft ‘set out to maximize Internet Explorer’s share of browser usage at Navigator’s expense.’" Pls. Br. at 94 (quoting 84 F. Supp. 2d at 43 (FF 133)). Of course, every competitor seeks to win market share from its rivals—that is competition. Plaintiffs cite no case in which a court held that the goal of winning as much business as possible at the expense of a rival satisfies the specific intent element of attempted monopolization. To the contrary, the intent to "expand one’s market share is presumptively lawful." MCI v. AT&T, 708 F.2d 1081, 1113 (7th Cir.), cert. denied, 464 U.S. 891 (1983). Courts thus have held that "[m]ore than an intent to win every sale, even if that would result in the demise of a competitor, is required before it can be concluded a defendant has the type of exclusionary intent condemned by the antitrust law." Transamerica, 481 F. Supp. at 1010; accord Great Escape, Inc. v. Union City Body Co., 791 F.2d 532, 541 (7th Cir. 1986) ("[M]ere intention to exclude competition and to expand one’s business is not sufficient to show a specific intent to monopolize."). Indeed, if an intent to maximize market share at the expense of a rival were sufficient, all competitive companies would harbor an illicit intent to monopolize.

Plaintiffs argue that the rapid increase in IE’s usage share from an insignificant share in 1995 and 1996 to over 50% compels a finding of dangerous probability of monopolization. Pls. Br. at 95 (citing Multistate Legal Studies, Inc. v. Harcourt Brace Jovanovich Legal & Prof’l Publ’ns, Inc., 63 F.3d 1540, 1554-55 (10th Cir. 1995)). Leaving aside the district court’s failure to perform the analysis required to define a legally cognizable "market" for Web browsing software—which alone is fatal to plaintiffs’ attempted monopolization claim—plaintiffs have not shown the dangerous probability required by Section 2.

The fact that IE’s usage share increased rapidly between 1996 and 1998 does not establish a dangerous probability of monopolization. As Areeda explains:

[I]f the defendant can experience rapid growth in market share, others can as well. Market shares that go from 0 to 60 percent in two years . . . suggest an unstable market in which it is unlikely that any firm could maintain a monopoly output reduction for very long. In such cases especially close attention should be paid to entry barriers, for it would seem that any new entrant could quickly undermine a monopolist’s position.

IIIA Areeda, supra ¶ 807e2, at 359-60 (emphasis added) (footnote omitted). The district court did not find that there are any barriers to entry into the "Web browser" market. Nor did it find that Netscape—now owned by AOL Time Warner—is even close to being driven from the putative market; to the contrary, usage of Navigator is growing. 84 F. Supp. 2d at 103 (FF 378). In the absence of such findings, the attempted monopolization ruling cannot stand. See Dial A Car, Inc. v. Transp., Inc., 82 F.3d 484, 487 n.2 (D.C. Cir. 1996) ("[N]o facts have been alleged to indicate that any company . . . is even close to being driven from the . . . market."); id. at 488 ("Equally important, appellant’s complaint does not allege any barriers that would prevent entry into the market by any number of additional providers.").

Faced with AOL’s announcement that it intends to replace IE with Navigator in AOL’s proprietary client software—which would reduce IE’s usage share to between 24% and 32%—plaintiffs argue that AOL’s purchase of Netscape is irrelevant because dangerous probability of monopolization is determined "as of the time of the conduct." Pls. Br. at 96. Between 1995 and 1998—the time of the challenged conduct—IE’s usage share rose from 0% to approximately 45%. See 84 F. Supp. 2d at 98-99, 100 (FF 360, 364). Under the very case law cited by plaintiffs, Pls. Br. at 95, such a usage share—even if it reflected control over productive capacity, which it does not—is insufficient to establish a dangerous probability of monopolization, see M & M Med. Supplies & Serv., Inc. v. Pleasant Valley Hosp., Inc., 981 F.2d 160, 168 (4th Cir. 1992) ("claims involving between 30% and 50% shares should usually be rejected"); see also MS Br. at 123-24. Plaintiffs thus lose either way: either dangerous probability is determined at the time of trial, in which case AOL’s purchase of Netscape defeats such a probability, or it is determined at the time of the conduct, in which case IE’s usage share was insufficient to establish a dangerous probability of monopolization.

It is now apparent from the district judge’s public comments that the relief awarded in this case was not a carefully tailored response to the antitrust violations found, but rather an attempt to punish Microsoft for what the district judge termed "‘arrogance which derives from power and unalloyed success.’" Auletta, supra at 397. In February 2000, the district judge reportedly told Ken Auletta that "he saw a structural remedy as too draconian and too risky" and "feared that the courts would not be able to divide Microsoft without damaging the company, and perhaps the economy." Ken Auletta, Final Offer, New Yorker, Jan. 15, 2001, at 43. But those well-founded concerns were discarded when, based on extrajudicial sources of information, the district judge became angry that Microsoft’s senior officers "‘to this day . . . continue to deny that they did anything wrong.’" Auletta, supra at 371. It is not an appropriate exercise of judicial power to break up Microsoft because it maintains that its conduct was lawful. Nor is the breakup justified because the district judge believes Microsoft’s chairman has "‘a Napoleonic concept of himself and his company,’" id. at 397, or becomes convinced, based on viewing a 25-year-old photograph of Microsoft’s founders (not in the record), that Bill Gates is "‘a smart-mouthed young kid’" who "‘needs a little discipline,’" id. at 168-69.

Plaintiffs insist that the entire decree should be affirmed even if this Court concludes that part of the district court’s liability determination is incorrect, Pls. Br. at 128, but their argument is flawed. The absence of any factual findings or legal reasoning on the subject of relief renders it impossible for this Court to determine which provisions of the decree were intended to redress which antitrust violations. In Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 459-60 (1993), the Supreme Court reversed a judgment because "the jury’s verdict did not negate the possibility that the § 2 verdict rested on