Publications

Amicus Curiae: Google, Inc. v. Oracle America

December 8, 2014

Download PDF | Download ePub | Download Postscript

No. 14-410
_____________________________________________
_________________________________________________________
In The
Supreme Court of the United States
_______________________
Google, Inc.
Petitioner,
v. 
Oracle America, Inc.
Respondent.
_______________________
On Petition for Writ of Certiorari to the
United States Court of Appeals
for the Federal Circuit
_______________________
BRIEF OF SOFTWARE FREEDOM LAW CENTER
AND FREE SOFTWARE FOUNDATION,
AMICI CURIAE IN SUPPORT OF RESPONDENT
_______________________

Eben Moglen
   Counsel of record
Mishi Choudhary
Jonathan D. Bean
Software Freedom Law Center
1995 Broadway, 17th Floor
New York, New York 10023
moglen@softwarefreedom.org
212-461-1900
_________________________________________________________
_________________________________________________________

Question Presented

Should this Court grant Petitioner’s request for review by writ of certiorari of the Federal Circuit’s erroneous prediction that the Ninth Circuit would have reversed the District Court’s conclusion that the application program interface declarations of Respondent’s Java programming language are uncopyrightable?

Interest of Amici Curiae

Software Freedom Law Center

Much of the world’s most important and most commercially significant software is distributed under copyright licensing terms that give recipients freedom to copy, modify and redistribute the software (“free software”).1 One could not send or receive e-mail, surf the World Wide Web, perform a web search or take advantage of many of the other benefits offered by the Internet without free software. Indeed, this brief was written entirely with free software word processors, namely GNU Emacs and LATEX, each of which are not just competitive with or superior to non-free software programs, but which also provide their users with the freedom to improve the program to fit their needs and reflect their desires.

The Software Freedom Law Center (“SFLC”) is a not-for-profit legal services organization that provides legal representation and other law-related services to protect and advance free software. SFLC provides pro bono legal services to non-profit free software developers and also helps the general public better understand the legal aspects of free software. SFLC has an interest in this matter because the decision of this Court will have a significant effect on the rights of the free software developers and users SFLC represents. More specifically, SFLC has an interest in preventing misapplications of copyright law that prevent innovative development of interoperable free software, and in educating parties on the benefits of free software copyright licensing.

Free Software Foundation

This brief is filed on behalf of the Free Software Foundation, a charitable corporation with its main offices in Boston, Massachusetts. The Foundation believes that people should be free to study, share and improve all the software they use and that this right is an essential freedom for users of computing. The Foundation has been working to achieve this goal since 1985 by directly developing and distributing, and by helping others to develop and distribute, software that is licensed on terms that permit all users to copy, modify and redistribute the works, so long as they give others the same freedoms to use, modify and redistribute in turn. The Foundation is the largest single contributor to the GNU operating system (used widely today in its GNU/Linux variant for computers from PCs to supercomputer clusters). The Foundation’s GNU General Public License is the most widely used free software license, covering major components of the GNU operating system and tens of thousands of other computer programs used on hundreds of millions of computers around the world. The Foundation strongly rejects the use of copyright law to prevent implementation of interoperable free software by inappropriately applying copyright principles to ideas instantiated in the rules of inter-program communication called “application program interfaces.”

Summary of Argument

The Court of Appeals for the Federal Circuit failed to apply, as required, federal copyright law as interpreted by the Ninth Circuit to the review of the District Court’s judgment. Instead, the court below implausibly predicted, based on no evidence in prior decisions, that the Ninth Circuit would have departed conclusively from the settled law established by Lotus Development Corp. v. Borland International, Inc., 49 F.3d 807 (CA1 1995), aff’d by equally divided court, 513 U.S. 233 (1996). Fashioning a new and conflicting standard for testing both the copyrightability and infringement of interface declarations in computer programs meant to interoperate with other programs through previously published application program interfaces (APIs), the court below departed from the regime under which software makers—both commercial firms and non-proprietary “free software” communities—have been successfully innovating for the last fifteen years. The court below made no effort to demonstrate that its new test has any precedent in Ninth Circuit copyright cases, nor in any cases decided in any of the regional Courts of Appeal, during the twenty-first century.

Despite the manifestly erroneous character of the decision below, the petition for writ of certiorari should not be granted. The precedential weight of a decision so evidently mispredicting the law of another Circuit is essentially nil. This Court does not grant certiorari for the general correction of error. The decision of the District Court on the copyrightability of the standard class declarations of Respondent’s Java programming language was made on the basis of careful consideration of the particulars involved: both the nature of the application program interfaces at issue, and the special factual considerations (scale and complexity of the interfaces involved) urged by Respondent Oracle. Judge Alsop made his decision on the basis of an educated understanding of the principles of computer programming not equally apparent in the opinion of the Federal Circuit below. The District Court’s decision was not only correct, but narrowly factbound, unsuitable as a vehicle for the clarification of broad principles by this Court that would justify a grant of certiorari.

In addition, there is insufficient case or controversy remaining in this matter to present a public interest in its continued adjudication. The court below found that Petitioner Google had made infringing use of 7,000 source code lines of interface declarations. Oracle America, Inc. v. Google, Inc., 750 F.3d 1339, 1353 (CAFC 2014). Even if that code contained copyrightable matter, it is undisputed that petitioner could have copied, modified and redistributed all that material, royalty free, under the terms of amicus Free Software Foundation’s GNU General Public License, version 2, the world’s most widely-used free software copyright license, under which Respondent Oracle offers the entirety of its Java programming language, Standard Edition product. Because, regardless of the disposition of this matter, Petitioner Google can make entire royalty-free use of the material whose copyright, if any, was allegedly infringed (and was at all relevant times validly so licensed), the dispute between the parties is merely theoretical, without practical consequences in the resolution of any actual case or controversy, and a grant of certiorari would be inappropriate on prudential grounds.

Argument

I. The Decision Below is Wrong

The court below correctly stated that its role was to apply federal copyright law to the instant litigation as it would be applied by the Ninth Circuit. 750 F.3d, at 1353. The court below apparently would have us believe that the Ninth Circuit has been awaiting an opportunity to unsettle a legal consensus crucial to the software industry in place since this Court’s affirmance by equal division of Lotus Development Corp. v. Borland International, supra. But the opinion below discloses no evidence whatever that the Ninth Circuit would hold that application program interface declarations are copyrightable. This conclusion was reached by the Federal Circuit on a frolic and detour of its own, in the course of which it misunderstood the factual record found by the District Court, used the law of the Tenth Circuit to predict the law of the Ninth Circuit, and in other ways created legal confusion and uncertainty from whole cloth.

At issue in this case is the copyrightability of 7,000 lines of interface declarations in the application program interfaces (APIs) of the computer programming language called Java. APIs are the rules by which computer programs interact with one another, analogous to the rules of games played by humans, like the rules of checkers or chess, which are agreed by all to be uncopyrightable ideas.2 The particular API declarations at stake in this case are the lists of verbs in a language: the Java programming language. When the Java language was first developed at Sun Microsystems, Inc.—Respondent’s predecessor in interest—the language’s creators defined a series of “standard functions” that every program written in the Java language could use without having to write them from scratch. Because of the linguistic structure of Java, those functions are divided into “classes,” which define the “methods” applicable to operations on particular data “objects.”

As described by the court below, 750 F.3d, at 1350-51, respondent Google literally copied these declarations into the source code of its Android mobile operating system, so that programs written in the Java language would run in that environment. The declarations associate uncopyrightable names with the functions available to programs written in Java. It is uncontested that Petitioner implemented all those functions using code of its own, with a single exception (the “rangeCheck” method) which Google argued represented de minimis infringement. The District Court held that the Java standard language declarations were an uncopyrightable “method of operation,” without which it would be impossible to operate a Java language processor running users’ programs written for execution by a Java system. The court below reversed, holding the declarations copyrightable and their literal copying to be infringement.3 On rangeCheck, and some other ancillary files not otherwise at issue here, the court below remanded to the District Court for a decision on the defense of fair use.

The court below, in other words, predicted that the Ninth Circuit would have reversed the District Court, on the grounds that the declarations are in themselves copyrightable. It predicted that the Ninth Circuit would not follow the decision of the First Circuit in Lotus, supra, holding that the literal copying of a command menu hierarchy and macro language for a spreadsheet program was not infringing. In Lotus, the Court of Appeals held that because the command hierarchy was a “method of operation,” it was an idea, like the rules of a game, outside the scope of copyright under 17 U.S.C. §102(b).

The Ninth Circuit has never expressed any criticism of Lotus, supra, or indicated that it would not follow the decision if similar facts arose within its jurisdiction. The court below reached its tortured contrary conclusion by a series of wrong steps:

First, the Federal Circuit held that the Ninth Circuit in Sega Enterprises, Ltd. v. Accolade, Inc., 977 F.2d 1510, 1525 (CA9 1992) expressed approval for the Second Circuit’s approach to computer program copyright infringement that culminated in Computer Associates International v. Altai, 982 F.2d 693 (CA2 1992), which the court below asserted differs from the holding in Lotus, supra, because it “eschews bright line approaches and requires a more nuanced assessment of the particular program at issue.” 750 F.3d, at 1357. The Ninth Circuit, in other words, would predictably not apply the rule in Lotus, supra, because three years before Lotus was decided it expressed approval of the Second Circuit’s supposedly conflicting doctrine. Second, in support of this prediction of Ninth Circuit law, the court below cites Mitel, Inc. v. Iqtel, Inc., 124 F.3d 1366 (CA10 1997), in which the Tenth Circuit expressed doubt about the decision in Lotus, supra, shortly after its affirmance by equal division in this Court.

Thin as this basis is for the prediction that the Ninth Circuit would have rejected Lotus, supra, the argument is made still weaker by the fact that the test set forth by the Second Circuit in Altai, supra, and mentioned approvingly in Sega, supra, concerned the method for determining infringement in the event of non-literal copying from computer program source code (what in more familiar literary copyright examples would involve “adaptation” of the copyrighted work), while Lotus, supra, and the instant case are cases of literal copying of application program interface declarations or equivalent menu hierarchies. Why the court below confidently predicted that the Ninth Circuit would fail to follow the holding of the First Circuit—affirmed by equal division in this Court—on facts squarely apposite, in favor of a novel approach based on cases inapposite because about adaptation rather than literal copying, it does not say.

In order to interoperate with other programs over a common API, the cooperating programs use the same names for data and function entities that they will separately implement. The names themselves are, all agree, uncopyrightable entities. Their relationship and organization is the method of interoperation. That methods of operation expressed as application program interface declarations are not copyrightable, which is the central holding of Lotus, supra, has been a productive and valuable legal certainty for the U.S. software industry.

Every firm in the information technology industry assumes that it is free to implement programs interoperating with other programs through published, defined APIs. The communities of programmers who make and share “free software,” like amicus Free Software Foundation and other clients of amicus Software Freedom Law Center, also assume that they are free to implement software of their own creation that interoperates with other software over publicly-defined APIs. Oracle itself benefits from the rule in Lotus, supra, in its own widespread and highly profitable software support business. Overturning the rule in Lotus subjects every effort at software interoperation throughout the information technology economy to 110-year state-granted monopolies of uncertain breadth. To predict that the Ninth Circuit would casually overturn the stability of existing doctrine, after fifteen years, on the basis of dicta in its cases from more than twenty years ago is utterly implausible. To summon support for that prediction from a Tenth Circuit case decided more than fifteen years ago is even more unconvincing.

II. Nevertheless, the Petition for Writ of Certiorari Should be Denied

The court below predicted, on the basis of no compelling evidence, that the Ninth Circuit would depart from settled existing law in a novel direction which, as amici supporting the petition have said at length, would be destructive alike of commercial certainty and freedom to implement, thus inhibiting the progress of science and the useful arts. Neither the Ninth Circuit nor any other regional Court of Appeals is likely to defer to this improbable supposition, so the error is largely self-limiting. Such erroneous predictions of other courts’ holdings are not a suitable employment of this Court’s scarce resources in review by certiorari. In addition, the actual holding in this case is factbound, revolving entirely around the question of how computer programming language APIs work. Moreover, the parties do not dispute that Google has a right to use all the putatively copyrighted materials at issue royalty-free, under the terms of amicus Free Software Foundation’s GNU General Public License. No public interest is served by the continued adjudication of a dispute in which there is nothing in fact at stake.

A. Respondents Argument for Copyrightability Hinges on Particular Factual Assertions; The Holding is Therefore Narrowly Factbound

Respondent Oracle did not argue in the courts below the broad proposition stated by the Federal Circuit, that application program interface declarations are copyrightable. The rule in Lotus, supra, is in fact too widely embedded in the practices of the industry for even a party widely known for its indomitable pertinacity in litigation to have attacked so broadly. Instead, Oracle has argued throughout that the scale of the Java APIs and their overall complexity raises them from the status of mere “methods of operation” to copyrightable computer programs.

The District Court rejected this argument as a matter of fact. The implementations of the standard Java classes are indeed complex and copyrightable computer programs. But (with the exception of “rangeCheck”), the implementations are no part of the litigation, the District Court found, because there is no allegation that Petitioner copied them. Petitioner’s relevant conduct, so far as the record disclosed to the District Court, was limited to the literal copying of the declarations. And these, the District Court found, are, as a matter of fact, limited to the names of parts of the Java language, organized in the one way that the grammar of the language requires them to be organized. Declaring those names, the District Court found, can only be done in this fashion: If Java programs are to be written in the language those declarations specify, those declarations can only be written in that one way to make interoperation possible. Since no other way of writing the declarations will be functional, the organization as well as the names themselves is entirely functional in character. It is, as the Lotus court held on precisely similar facts, a method of operation, an idea, properly placed outside the scope of copyright by §102(b). See Lotus, supra, 49 F.3d, at 818-19.

This is a factual finding about how computer programming languages like Java define the method functions—which are analogous to verbs—that make up the language. The court below ignored this factual finding. Instead, it said, the issue was one of originality only: from the perspective of the original creators of the Java language at Sun Microsystems, the language could have used different names for basic functions, and organized its grammar differently. 750 F.3d, at 1363. But this entirely misunderstands the real technical context of the District Court’s decision. Can one make up a entire language and copyright, say, the lexicon of Greek, because one could instead have used different words and invented Arabic? In the extremely specialized and confined context of the APIs for a computer programming language, this is the claim which the court below substituted for the actual highly-limited factual finding of the District Court. Granting certiorari on this record commits this Court to a revision of the erroneous holding below on the basis of a review of the very narrow technical issue on which the District Court’s copyrightability holding actually rests.

B. There Is No Public Interest In Continued Adjudication of this Dispute

There is no dispute between the parties that Petitioner can use, copy, modify and redistribute all of the putatively copyrighted material at issue, royalty free, under the terms of the most widely-used free software copyright license, the GNU General Public License, version 2, published by amicus Free Software Foundation. The Free Software Foundation also publishes a Java programming language system, including standard class libraries declaring the same method names, under the GNU GPL license, version 2 or any later version.

It appears that Petitioner wrongly supposed at some time in the past that use of Respondent’s Java source code under the terms of GNU GPLv2 would limit the licensing of the Java programs that third parties could install and run on Android systems. Andy Rubin, then an important Google executive with responsibility for Android, wrote an email to that effect which was part of the evidence at trial:

[A]s far as GPL-ing the VM, everything that is linked with the VM would get infected. The problem with GPL in embedded systems is that it’s viral, and there is no way (for example) OEMs or Carriers to differentiate by adding proprietary works. We are building a platform where the entire purpose is to let people differentiate on top of it.

Trial Exhibit 230, at 1.

This was simply wrong. As users of the Java programming language as supplied historically by Sun Microsystems and now by Respondent Oracle well know, though the Java language and standard classes are available for all to copy, modify and redistribute freely, without payment of royalties, under the GNU GPL, these parties develop, use, and distribute Java programs for execution by Oracle Java under a great variety of proprietary and free software copyright licenses.

Petitioner Google is now and has been entitled to use all the material at issue in this case, royalty free, under GNU GPL. Although Petitioner unaccountably failed to assert and preserve this defense of license at trial, it was in fact at all times licensed under the GNU GPL, version 2, to take all the steps it took in relation to the Java standard class declarations by Respondent Oracle or by its predecessor in interest, Sun Microsystems.

Given that the parties are agreed that Petitioner has the right to royalty-free use of all the material at issue under GNU GPL, and it is in addition entitled to claim that its use was licensed at all relevant times, there is no public interest in the adjudication of a controversy which remains merely theoretical if not factually moot.

Conclusion

For the foregoing reasons, the petition for writ of certioriari should be denied.

Respectfully submitted,

Eben Moglen
Mishi Choudhary
Jonathan Bean
Software Freedom Law Center
1995 Broadway, 17th Floor
212-461-1900

December 8, 2014

1Pursuant to Sup. Ct. R. 37.6, amici note that no counsel for a party authored this brief in whole or in part, and no counsel or party made a monetary contribution intended to fund the preparation or submission of this brief. No persons other than amici curiae and their counsel made a monetary contribution to its preparation or submission. Petitioner has consented to the filing of this brief through a blanket consent letter filed with the Clerk’s Office, and waived timely notification. Respondent has been notified of and consented to the filing of this brief pursuant to Sup. Ct. R. 37.2(a).

2See United States Copyright Office, FL-108, Copyright Registration of Games (2011) (“Copyright does not protect the idea for a game, its name or title, or the method or methods for playing it.”); 1-2 Melville B. Nimmer & David Nimmer, Nimmer on Copyright §2.18[H][3][A] (2014) (“[N]o copyright may be obtained in the system or manner of playing a game.”).

3It is uncontested that these 7,000 lines of declarations contained in the source code files for Oracle Java are available for everyone to copy, modify, and redistribute freely under the terms of the GNU General Public License, version 2, published by amicus Free Software Foundation. No one has ever satisfactorily explained how Petitioner Google can be liable for copyright infringement for making literal use of them as interface declarations for other programs, when such use is freely allowed to all. Google unaccountably failed to raise this particular defense of license in the courts below.