Paul Krill of Infoworld recently reported on a recent US federal court ruling, which Mark Radcliffe, counsel for the Open Source Initiative,  fears “could derail enforcement of open-source licenses altogether” by foreclosing licensors’ ability to get an injunction against license violations.  (Quotation to article, not Mark Radcliffe).  In this ruling, Judge Jeffrey White analyzed the scope of an open source license (the Artistic License,  which wins the award for best name) and found that provisions requiring attribution were contractual terms rather than limitations on the scope of the copyright license.  As a result of this finding, Judge White denied the plaintiff’s motion for a preliminary injunction, raising the fear that courts could limit the remedy for violations of open-source licenses to uncertain money damages. 

Some people are working overtime to find the silver lining for the FSF.  While Krill quotes Eben Moglen as saying "

[i]t seems to me a not very important ruling," ZDNet bloggers Dana Blankenhorn and Paula Rooney believe that this case will allow the FSF to seek money damages for "interfering in a business relationship."  (Who is interfering with whose relationship?   See Richard Wilder’s white paper on Legal Risks.) 

Mark Radcliffe feels that “this decision is simply wrong.”   In particular, he believes that “[t]he use of the term ‘condition’ in the Artistic License should mean that the terms imposed are restrictions on the scope of the license.”  Radcliffe’s position is supported by Professor Raymond Nimmer (in the treatise Information Law § 11.153) and others.  I don’t agree.  By this reasoning, you and I can use the word “condition” to modify the scope of rights allowed under copyright law, regardless of the substance of the condition. 

I am going take this opportunity to try to simplify the position that Richard Wilder and I took in GPLv3 is a Contract and Why it Matters.   In essence, we agree with Professor Nimmer  that whether a particular term is contractual depends on whether the term is intended to create an enforceable obligation.  Rather than focusing on certain “magic words” like “condition,” I would contend that there are three basic propositions that control:

1)    The scope of copyright is specified in statute; for example, the exclusive rights listed in 17 U.S.C. § 106.   A license is essentially a waiver of the right to sue someone for exercise of these rights.

2)    For breach of a license term to be considered copyright infringement, the breach must be the exercise of one of the exclusive rights that wasn’t allowed in the license.  For example, you may divide your licensable copyrights by field of use.  Thus, if you limit your license to use of your software in non-commercial settings, it’s use in a commercial setting would be copyright infringement.  The use exceeded the scope of the license.

3)    Any other provision is contractual and is ruled by the law of contracts – including that a contract can be formed only if there is an offer, acceptance, and consideration.

In the Jacobsen v. Katzer case,  the plaintiff contended that the defendants redistributed JMRI code without properly attributing it to the original authors, as the Artistic License requires.  For literary works (which include computer software) attribution is not one of the exclusive rights under US copyright law (although it is a right under the Berne Convention for the Protection of Literary and Artistic Works  – to which the US is a party).  Judge White correctly determined that “[t]he condition that the user insert a prominent notice of attribution does not limit the scope of the license” because the U.S. copyright statute doesn’t speak to the issue of attribution.

So how would this affect GPLv3?  Let’s take one provision in particular, the so-called “anti-Tivo” provision in section 6,  which requires anyone that distributes GPL’d code as part of a consumer product to include installation information and the ability to install modified versions of the code on the consumer product.  One could argue that the FSF has limited the scope of the license to prohibit distribution in consumer devices that don’t allow modification of device software, meaning that a breach would be copyright infringement.  I think the more logical reading is that the FSF has imposed an additional obligation on producers of consumer devices.  Failure to fulfill that obligation would be a breach of contract, not a copyright infringement. 

That doesn’t preclude preliminary injunctions in license enforcement actions if the copyright holder can prove irreparable harm along with the other requirements for a preliminary injunction.  What it means is that if the FSF hopes to bind parties  to the contractual terms of GPLv3, they are going to have to show that all of the prerequisites of contract formation have been followed, including offer, acceptance, and consideration.