Stiffling interoperability and competition

Mike Madison has an interesting post on the lawsuit between the proprietary software package EndNote and the open source rival Zotero, from George Madison U, funded by the Mellon Foundation. Zotero is a Netscape plugin for organizing academic papers that I found less useful than a commercial alternative.

In its lawsuit, the owners of EndNote (Thomson Scientific, now Thomson Reuters) argues that its contract prohibits reverse engineering. Thus, it contends, Zotero’s developers violated contract law by using reverse engineering to figure out how to get interoperability with EndNote.

The lawsuit seems prompted by a recent Zotero feature that enabled importing EndNote libraries journal styles and thus eliminating reducing the lock-in Thomson thought it owned. Since EndNote compatibility is mainly for one-time imports, it seems like Zotero could easily strip this functionality and have it distributed as a plug-in by an open source coalition located in .ru or some other scofflaw domain).

I’m no lawyer and haven’t kept up with the law, so I largely have to defer to Prof. Madison’s analysis of the case. Madison notes that under federal law, reverse engineering for compatibility is probably legal fair use, and thus Thomson filed in state court.

My first exposure to reverse engineering started in practice. Like many other software engineers making Apple-compatible software, I did a fair amount of reverse engineering of Apple’s code for just this purpose. Apple knew it was going on (or at least the tech support people did) but never sued anyone except the folks who tried to make unlicensed clone computers.

I also immersed myself in reverse engineering law for a night school term paper in an Asian studies class I took with Chalmers Johnson. This led in 1995 to my first published journal paper (and second† academic paper) about a Japanese policy controversy over how copyright law handled reverse engineering of software for compatibility. At the time, Sun (and some Japanese firms) wanted to reverse engineer IBM systems to provide interoperability and the IBM people wanted to discourage such reverse engineering. The identical issues played out with Linux and Microsoft in more recent years.

As I found in 1995, two major US appeals verdicts in 1992 — Sega v Accolade and Atari v. Nintendo — set a precedent under US law for allowing reverse engineering. After the 1992 precedents, two other important articles were published on reverse engineering and interoperability: one by Julie Cohen in the California Law Journal and another by (my recent friend) Charles McManis in the High Technology Law Journal.

Then as now, I think the precedent and the policy outcomes were reasonable, since copyright law provides production only for the expression of an idea, and not against revealing trade secrets (which are compromised by public distribution, even of the object code). The cost of reverse engineering and re-implementation usually provides adequate value and protection for the author’s original expression, while adding allowing it to be used for compatibility enables competition.

So while I have not inspected the Thomson lawsuit, I am hoping that their competitor (who happens to be a university developing open source) wins the case on its merits. No software rents are perpetual: software gets commoditized over time, and if you can’t deal with it, you’re in the wrong business.

† The first was designing a programming language, predating by nearly a decade my decision to become an academic.

Updated 3:30pm: Oops. As anonymous Bruce points out, the lawsuit is about importing the EndNote styles, which are a set of templates defining the reference format of a wide range of publications. (I have never used EndNote, so I didn’t understand the distinction.) Since this is not conversion of user content but of EndNote’s content, it seems to me that the Zotero is in a much more precarious legal position.