IFOSS L. Rev. Vol 1 Issue 1: "Bad facts make good law: The Jacobsen case and Open Source" by Lawrence Rosen

Back to Volume 1, Issue 1 table of contents

Bad facts make good law: The Jacobsen case and Open Source

Lawrence Rosen1



For the first time, a major U.S. appeals court has held that an open source license is enforceable through preliminary injunctions. The court also found that exacting conditions in the form of compliance with open source requirements for disclosure and explanation of changes is entitled to recognition as consideration for a contractual license.



Open source licenses; contracts; enforcement; infringement;





For many years the open source community has been eagerly awaiting a U.S. lawsuit that would bless the open source licensing model. In what in retrospect may seem like a leap of faith, millions of software programmers around the world published their works expecting that their open source licenses, including the GPL, would be honored and enforced in court. That fundamental assumption was never effectively tested in court, until the Jacobsen v. Katzer case [Page references in this article are to 535 F.3d 1373; 208 U.S.App. LEXIS 17161, Aug. 13, 2008.]


This decision finally explains how U.S. courts should analyze open source and open content licenses. The bottom line for us is that copyright law provides the remedies but contract law provides the analytical tools.


This is a strange case, based upon facts that did not seem favorable to Jacobsen ("J" in this article) when I first heard of him several years ago. For one, the license actually used by J was the first version of the Artistic License, a document of which one could generously say that the non-lawyers who drafted it years ago took artistic license with legal style and method. It is not a very good license. Anyone looking for the formalities and magic words of modern proprietary or open source software licenses would be disappointed with the old Artistic License. It is not the license of choice on which we would want to hang the entire open source copyright licensing model in federal court.


The Jacobsen case found its way to the Court of Appeals for the Federal Circuit (CAFC), arguably the most important court short of the U.S. Supreme Court for intellectual property matters. That itself is a story that may interest lawyers who enjoy civil litigation, and may serve as an object lesson for plaintiffs and defendants alike. This case ended up in the CAFC because of statutory and factual reasons completely unrelated to the license itself. The court in which one litigates can present opportunities—and risks—that the parties to that lawsuit may not intend. Lawyers who enjoy civil litigation practice may appreciate the story that unfolds in this partly historical paper.

The most interesting part of the story took place during the recent few months in which, under the leadership of attorneys at Creative Commons, various open source organizations cooperated to write an amicus brief that focused the CAFC on the critical issues that we wanted the court to answer firmly for us.


We got just that. The Jacobsen decision finally settled that open source licensors can enforce their licenses effectively in U.S. courts. Defendants can't avoid copyright law by relying on contract law principles that are inapplicable in the open source and open content world.


As is true for many decisions that reach the highest appellate courts, this case has important implications far beyond open source. Creative Commons lawyers were involved in this case because Jacobsen would also settle the enforceability of their licenses, under which hundreds of millions of musical, literary, and artistic works are made available to the world on generous copyright terms. The authors of those works now have a valuable remedy—the preliminary injunction—to enforce their copyrights in U.S. Courts.

Jacobsen's Complicated Civil Litigation Path

The plaintiff in this case, J, is a physicist who, as a hobby, developed software for controlling model trains. He licensed it to the world under the first version of the Artistic License. This license authorizes anyone to copy, modify and distribute the software on condition that they place certain notices in those works as attribution to its original author. Licensees are also required to identify that they have changed the work, to preserve the original author's reputation.


A small open source community developed around that model train software. Among those who took that software was the defendant, Katzer ("K" in this paper). According to the complaint, K modified the model train software, placed his own name on it, and distributed it to his customers. K didn't place the required attribution notices in his copies.


As this case progressed, K never actually denied that those copies were made and distributed, but as described below, this case never reached the factual resolution stage. To the date this article is written, K has never formally answered the complaint, so we must assume that the facts as portrayed in J's complaint are true.


Before a lawsuit was filed, during the early stages of this open source dispute about attribution requirements of the Artistic License, K took the unusual step of retaliating by sending a letter to J accusing him of infringing a patent. The legal basis for that accusation was dubious at best; companies around the world throw such letters into the virtual trash can every day. But J (and his wife) didn't know how to react; their life savings were at risk.


J subsequently reacted by filing a lawsuit for declaratory relief that he had not infringed K's patent. Filing a declaratory relief action is sometimes an appropriate reaction to a false accusation of patent infringement.


A litigation strategy involving patent claims can sometimes be dangerous. In the U.S., patent infringement claims are always litigated in District Court, and all appeals are automatically routed to the CAFC. The Court of Appeals for the Federal Circuit is a careful and sophisticated appeals court that often deals with complex technical issues. Patent lawsuits require a high level of diligence before they are filed so they don't waste the court's time.


Both a potential defendant in a claim for declaratory relief for patent infringement (before he writes the accusatory letter) and the plaintiff seeking that relief (before he files the declaratory relief lawsuit because of that accusatory letter) ought first to analyze the patent claims carefully. It soon turned out that K's patent infringement allegations were bogus. The plaintiff, J, and his wife need not have been so afraid. But by then the litigation path was fixed: Even without those mooted patent claims being relevant any longer and until a final resolution of the case, federal law requires that appeals be heard by the CAFC (or ultimately the U.S. Supreme Court). This automatically became an "important case."


A second characteristic of this litigation was the large catalog of additional tort claims asserted in J's complaint for declaratory relief. Some of those additional causes of action, of course, were the fundamental copyright claims that later became the focus of the CAFC decision. But the litigation was also burdened by other tort causes of action relating to trademark, unfair business practices, and breach of contract. Some claims involved state law, some federal law. Many of these non-patent claims involved complicated choices of law, with different issues of available remedies and different penalties, including attorney's fees. This case was for months mired in preliminary motions, all involving the non-patent claims. Dispute resolution attempts failed.


In such situations, it is not unexpected that the District Court would try to resolve this case as if it were a contractual dispute: If the Artistic License demands that licensees place attribution notices in their copies but they don't, determine the dollar value of those notices and assess damages. The District Court's decision to treat this as a simple contractual complaint would, in many situations, make it more likely that the parties would settle the dispute themselves for money. Such negotiated resolutions are the predilections of most courts because they encourage mediation or arbitration rather than expensive motion practice and trials. But these parties didn't settle, for various reasons unrelated to the software itself or its license.


All those matters were destined to be resolved in federal court because the declaratory relief patent claim carried those other claims along by supplemental jurisdiction. Federal court, and in particular the CAFC, is not usually the correct venue for such disputes, and so the CAFC used the law of the parties' local jurisdiction, California and the Ninth Circuit, in deciding this case.2

Open Source Frames the Important Issue

This case deals with hobby software, given away by J under an open source license, subsequently used by K for minor commercial purposes, albeit without honoring an attribution provision in the license. The dollar value of a settlement on contract law terms would be small. At the scale of the vast software world the value of notices in J's and K's model train software is de minimis; such a claim cannot justify the cost of federal litigation.


More practically, for contractual disputes in the U.S., there is no automatic provision in federal court for attorney's fees. Clients take on such cases sometimes at great personal financial risk that they will end up with nothing but some small penalty assessed against a bad faith licensee, and a large bill for attorney's fees. (Of course, a contract can expressly provide for attorney's fees to the victor, but the Artistic License contains no such provision.)


Copyright law, on the other hand, permits a preliminary injunction as a remedy for infringement, and the law provides for attorney's fees in certain circumstances. Among the many remedies sought in J's complaint, injunctive relief for copyright infringement was perhaps the most important. Such an order from the court could stop K's business cold, no matter how large or small. And then K might also be liable to pay attorney's fees.


It was understandable that K fought against that preliminary injunction in court. It was understandable that J demanded it.

A preliminary injunction usually requires at least some evidence of likely harm if the court doesn't take action and, in this particular case, there was no such ready factual evidence before the court. No matter, the District Court said, that a copyright infringement claim is appropriately and often resolved by a preliminary injunction, because that is irrelevant for contract claims. The District Court took a specific path toward resolving this dispute by treating the Artistic License as a contract, and thereupon applied contract law to refuse a preliminary injunction.


This became the fundamental issue of the case on appeal: Was it proper for the District Court to view this as a contract dispute rather than a copyright infringement dispute when considering the remedy of preliminary injunction?


That's also what caught the attention of the open source and open content organizations when J appealed the District Court's refusal to order a preliminary injunction. That became the single issue upon which these groups had a shared interest: Unless a breach of a copyright license can be stopped outright by a preliminary injunction, there is often no value in a contract lawsuit, especially for works published for free public use and with small commercial value.


Only under copyright law—which generally permits preliminary injunctions as way of dealing with copyright infringement, and where attorney's fees can be assessed against bad faith defendants—is there some hope of enforcing the conditions that many authors demand for their works.

The CAFC Analyzes the Important Issue

The CAFC boiled the Jacobsen case down to its fundamental issue: "The heart of the argument on appeal concerns whether the terms of the Artistic License are conditions of, or merely covenants to, the copyright license.3


Whether a provision is a condition or a covenant affects the way courts have previously dealt with copyright licenses. A "mere covenant" (particularly an "independent covenant"), when breached, is to be remedied under contract law. A "condition," on the other hand, is a limitation on the scope of the copyright license itself, and courts should treat its breach as copyright infringement.4


Then specifically analyzing the Artistic License, the CAFC said that J's requirement to publish attribution notices was a "condition" of the license. K's failure to do so was a use outside the scope of the license and therefore infringing.5  The remedy for such infringement is found in copyright law. For copyright infringement, the court said, "where a copyright holder has shown likelihood of success on the merits of a copyright infringement claim," the Ninth Circuit has held that irreparable harm is presumed.6


The Jacobsen case was remanded "for further proceedings consistent with this opinion."7

The people who wrote the Artistic License, and those who wrote the GPL, and those who wrote many other open source licenses, lucked out on the Jacobsen case. Many of us license authors didn't know the legal difference between a "covenant" and a "condition" when our licenses were written (and many attorneys still don't).  Fortunately, the Artistic License states on its face that the document creates conditions, and so the CAFC found conditions in the Artistic License when we needed them most. If the court had found covenants in that license, open source might have lost this case.


This presumably does not mean that the new magic word for open source and open content licenses is "condition." The CAFC relied on a long-ago California Supreme Court decision to the effect that a condition can be found by "attributing the usual and ordinary signification to the language of the parties."8 Thus the CAFC determined that the Artistic License, when it also uses the phrase "provided that," "denotes a condition" under California contract law.9


This does, however, raise an interesting question: Of the current approved open source and Creative Commons licenses, which of them clearly distinguish their conditions from their covenants, and under which state's contract law do we analyze that question?

Other Important Lessons from the Jacobsen Case

The CAFC also resolved for us certain other concerns that arise when licensors try to enforce their open source copyright licenses. These strengthen our ability to set conditions for the use of our copyrighted works and to enforce those terms against those who don't meet those conditions.


At oral argument in the CAFC, the court asked the attorneys for both parties whether it is important that this software, like much other open source software, is licensed more for its reputational and cooperative value in the "community" rather than for financial reward. This was an important enough issue that it warranted a supplemental brief by the Creative Commons attorneys. The CAFC decision reflected the arguments in that brief.


The Artistic License was "clear" that it intended to "accomplish the objectives of the open source licensing collaboration, including economic benefit."10 But it is not only an economic benefit that is at stake in the Jacobsen case.


"The choice to exact consideration in the form of compliance with the open source requirements of disclosure and explanation of changes, rather than as a dollar-denominated fee, is entitled to no less legal recognition."11


The court's decision is very helpful for reassuring millions of copyright holders who engage in open source and Creative Commons licensing that they "have the right to control the modification and distribution of [their] copyrighted material."1112


The CAFC decision also requires district courts to enforce open source license restrictions (another term like "condition" that has ambiguous meaning here) through the remedy of injunctive relief. Otherwise, "those types of license restrictions might well be rendered meaningless."13  


Even though injunctive relief is available, the plaintiff (J) is still required to prove that the conditions of the Artistic License were actually violated. The Jacobsen case was remanded for further action in District Court, not finally determined by the CAFC.14


As of the date this article is written, litigation continues. But we can take great comfort from this case already.15

About the author

Lawrence Rosen is both an attorney and a computer specialist. He is founding partner of Rosenlaw & Einschlag, a technology law firm with offices in Los Altos Hills and Ukiah, California, that specializes in intellectual property protection, licensing and business transactions for technology companies.


In addition to this law practice, Larry also served for many years as general counsel and secretary of the non-profit Open Source Initiative (OSI). He currently advises many open source companies and non-profit open source projects including Apache Software Foundation and the Python Software Foundation.  In 2005-2006 he was a Lecturer in Law at Stanford Law School.


Larry currently serves on the board of directors of International Characters, and on the advisory boards of SpikeSource, Black Duck Software, and JasperSoft.


Licence and Attribution

This paper was published in the International Free and Open Source Software Law Review, Volume 1, Issue 1 (July 2009). It originally appeared online at http://www.ifosslr.org.

This article should be cited as follows:

Rosen, Lawrence (2009) 'Bad facts make good law: the Jacobsen case and Open Source', IFOSS L. Rev., 1(1), pp 27 – 32

Copyright © 2009 Lawrence Rosen

This article is licensed under a Creative Commons UK (England and Wales) 2.0 licence, no derivative works, attribution, CC-BY-ND.

As a special exception, the author expressly permits faithful translations of the entire document into any language, provided that the resulting translation (which may include an attribution to the translator) is shared alike. This paragraph is part of the paper, and must be included when copying or translating the paper.

1© Lawrence Rosen 2008. This paper was prepared for presentation at the PLI Open Source Software 2008 program in New York (11/18/2008) and San Francisco (12/10/2008, also webcast). It is also published at the author's website, http://www.rosenlaw.com.

2Jacobsen v. Katzer, 535 F.3d 1373, 1377-1378 (Aug. 13, 2008).

3Id. at 1380.

4Id., and cases cited therein including Sun v. Microsoft, 188 F.3d 1115 (90th Cir. 1999).

5Id. at 1381.

6Id. at 1378.

7Id. at 1383.

8Id. at 1381, citing Diepenbrock v. Luiz, 159 Cal. 716 (1911)



11Id. at 1382.

12Id. at 1381.

13Id. at 1382.

14Id. at 1383.

15I want to acknowledge the diligent efforts of Jacobsen's attorney, Victoria Hall (Law Offices of Victoria Hall), who  persevered in this case despite my skepticism and won important additional allies to her client's cause. The legal  heroes were two attorneys from Creative Commons, Anthony Falzone and Christopher Ridder, who under the  leadership of Prof. Lawrence Lessig (Stanford Law School) submitted the amicus brief to the CAFC. Also providing  legal support during litigation and appeal were Mark Radcliffe, David Gross and Steve Chiari (DLA Piper and  counsel to Open Source Initiative (OSI)); Karen Copenhaver (Choate Hall and counsel to the Linux Foundation);  Allison Randal (PERL Foundation, author of the newer and better version of the Artistic License); Roberta Cairney  (Law Offices of Roberta Cairney and counsel to the PERL Foundation); and Scott Peterson (HP and a  member of the OSI Legal Advisory Counsel). I was honored to work with them on this case. I apologize that this list  is incomplete.