I don’t particularly like talking law on this blog; it’s boring and – for the most part – disinteresting. However, recent developments in SAS Institute Inc v World Programming Limited (as written up here – thanks to Cristian for bringing this up at FFII) deserve to be aired.
The basic story is that the Judge in this case is deeply unsure of the boundary of copyright. For those who don’t know, SAS is a statistical package which is both popular and influential, and to a large extent can be thought of as a programming development environment. WPL, the defendants, wrote software which could interpret SAS programs. There is no direct analogy in the free software world, but LibreOffice Calc interpreting Excel spreadsheets is close enough for the purposes of our discussion.
The Judge, unsure of the boundary, has sent a number of questions to the European Court of Justice (ECJ). The questions are hypothetical, but clearly designed to test the waters and figure out where this line falls. As an example of some of the questions in our Calc vs. Excel example, he’s asking:
- Does accessing the file format of Excel constitute copyright infringement?
- Does interpreting Excel-compatible formula constitute copyright infringement?
- Is it copyright infringement to copy the behaviour (intended or otherwise) of Excel when processing spreadsheets?
- Does it make a difference you copy functionality from Excel into Calc by reading Excel’s manual versus observing Excel’s behaviour?
- Does it make a difference if you have a license to Excel?
(Just to be clear, we’re not talking about Calc and Excel, I’m just interpreting J. Arnold’s questions in this context to make them more readily understood)
Now, the armchair lawyers amongst my readership have probably already thrown their arms up at all these questions and exclaimed, “Copyright doesn’t extend that far!”. And to an extent, they would be correct: originality has always been a defence to copyright infringement, and if any of the questions above were to be answered in the affirmative, we would see the start of that changing.
What makes this different, I think, is that we’re really seeing the weakness of copyright law treating software as a literary work. This has always been bunk, really: software is no more literary than a shopping list, and although the case of verbatim copying (with or without transformation) is open-and-shut copyright infringement there have always been “grey areas”. As one example, the FSF’s position on dynamic linking and the GPL: as a derivative work it does seem to fall under the purview of copyright, but it’s obviously a world away from literary copyright.
The WPL case is also one where the copying was explicit, deliberate and planned: they definitely did copy things. They just didn’t literally copy the software code, or decompile the software: they re-created it from the ground-up. So we’re definitely talking about a case of copying here, which it would seem could also be the purview of copyright.
This is going to be a really interesting case, and is going to have a fundamental effect on free software if we get some interesting answers to these questions. On one hand, it casts an immediate dark shadow over a number of projects: Samba being an obvious case in point, which has previously reached legal agreement in Europe about how it can copy Microsoft while still avoiding the patents that Microsoft hold on certain functionality. But while desktop apps which copy Microsoft make the most obvious cases, you could equally see problems for 3D graphics drivers, people implementing compilers, all sorts of areas – particularly where free software is still catching up to proprietary software.
But of course on the other hand, this would also strengthen the copyright position of free software applications. Companies that currently dance around the (L)GPL-style licenses will find themselves on thin ice indeed, and those proprietary implementations of leading free software will start having to be extremely careful.
It’s very unlikely that many of the questions will be answered in such a way that the copyright system becomes like the patent system: for one, it would be such a massive change that it would require primary legislation at a European level to become legally sound. And there are few cases exactly like this one, where the copying is so obvious and blatant.
The precedents being set here will be extremely important, though. Our understanding of copyright will almost certainly change from the outcome in this case, and will necessarily become more nuanced. The idea of “clean-room reverse engineering” may become more nebulous, and the “I wrote it from scratch” defence could become weaker.
If nothing else, this highlights that no law is truly ever settled, and possibly portents to more movement in this area in the future: I’ve described before how the UK Government is making noises about revisiting intellectual property laws, and in our current weak economic state it is extremely tempting for politicians to beef up some of these laws in order to “create wealth”. Cameron, our Prime Minister, is particularly in thrall to Google, as if they set any good example for our businesses. It’s sometimes very easy to just think about patents and lose sight of the bigger picture.