License to kill
Yesterday, a US federal appeals court reversed a lower court ruling that might have invalidated open-source licenses. The case, Jacobsen v. Katzer, began more than two years ago with a patent claim.
Open-source software developer Robert Jacobsen manages the collective effort that produced Java Model Railroad Interface, which allows enthusiasts to reprogram the controller chips in their trains. JMRI is distributed under the Artistic License, an older and less-well known one of the free licenses (it isn't one of the Free Software Foundation's approved licenses, though its successor, Artistic License 2.0, is). Matthew Katzer and Kamind, aka KAM Industries sells a functionally similar commercial product that, crucially, Jacobsen claims is based on downloaded portions of JMRI. The Artistic License requires attribution, copyright notices, references to the file listing copyright terms, identification of the source of the downloaded files, and a description of the changes made by the new distributor. None of these conditions were met, and accordingly Jacobsen moved for a preliminary injunction on the basis of copyright infringement. The District Court denied the motion on the grounds that the license is "intentionally broad", and argued that violating the conditions "does not create liability for copyright infringement where it would not otherwise exist". It is this decision that has been reversed.
This win for Jacobsen doesn't get him anything much yet: the case is simply remanded back to the California District Court for further consideration. But it gets the rest of the open-source movement quite a lot. The judgement affirms Richard Stallman's original insight that created the General Public License in the first place, that copyright could be used to set works free as well as to close them down.
The decision hinges on the question of whether the licensing terms are conditions or covenants, a distinctions that's clear as glass to a copyright lawyer and clear as mud to everyone else. According to the Electronic Frontier Foundation's helpful explanation (and they have lots of copyright lawyers to explain this sort of thing), it's the difference between contract law and copyright law. Violating conditions means you don't have a copyright license; violating covenants means you've broken the contact but you still have a license. In the US, it's also the difference between federal and state law. When you violate the license's conditions, therefore, as Lawrence Lessig explains , what you have is a copyright infringement.
It's hard to understand how the district court could have taken the view it did. It is very clear from both the licenses themselves and from the copious documentation of the thinking that went into their creation that their very purpose was to ensure that work created collectively and intended to be free for use, modification, and redistribution could not be turned into a closed commercial product that benefited only the company or individual that sells it. To be sure, it's not what the creators of copyright - intended as a way to give authors control over publishers - originally had in mind.
But once you grant the idea of a limited monopoly and say that creators should have the right to control how their work is used, it makes no sense to honor that right only if it's used restrictively. Either creators have the legal right to determine licensing conditions or they have not. (The practical right is of course a different story; economics and the size of publishing businesses give them sufficient clout to impose terms on creators that those creators wouldn't choose.). Seems to me that a creator could specify as a licensing condition that the work could only be published on the side of a cow, and any publisher fool enough to agree to that would be bound by it or be guilty of infringement.
But therein lies the dark side of copyright licensing conditions. The Jacobsen decision might also give commercial software publishers Ideas about the breadth of conditions they can attach to their end-user license agreements. As if these weren't already filled with screeds of impenetable legalese, much of which could be charitably described as unreasonable. EFF points this out and provides a prime example: the licensing terms imposed by World of Warcraft owner Blizzard Entertainment have been upheld in court.
Blizzard's terms ban automated playing software such as Glider, whose developer, Michael Donnelly, was the target of the suit. EFF isn't arguing that Blizzard doesn't have the right to ban bots from its servers; EFF just doesn't think accusing Glider users of copyright infringement for doing is a good legal precedent. Public Knowledge has a fuller explanation of the implications of this case, which it filed as an amicus brief. Briefly, PK argues that upholding these terms as copyright conditions could open the way for software publishers to block software that interoperates with theirs. (Interestingly, Blizzard's argument seems to rely on the notion that software copied into RAM is a copyright infringement, an approach I recall Europe rejecting a few years ago).
You'd think no company would want to sue its own customers. But keeping the traditional balance copyright law was created to achieve between providing incentives for artists and creators and public access to ideas continues to require more than relying on common sense.
Wendy M. Grossman's Web site has an extensive archive of her books, articles, and music, and an archive of all the earlier columns in this series. Readers are welcome to post here, at net.wars home, at her personal blog, or by email to netwars@skeptic.demon.co.uk (but please turn off HTML).