Letters A fortnight ago we rued the fact that our most valuable social contract, the GPL was heading to the courtroom for the first time. FSF counsel Eben Moglen assured us not to worry, and that the GPL would be upheld under United States Law at its first major test.
But what about the rest of us? A win in the US does not guarantee a free passage for the rest of us, as Gary Lea points out. Gary is a law lecturer in London and highlights an important international dynamic that has been absent from the SCO vs. IBM coverage so far. We reproduce it in full.
In general terms, I both think and hope that Eben Moglen is right over the GPL - his interpretation makes sense, SCO's does not and it would be a total disaster if SCO's view is upheld. However, there were a couple of things in your report (at paras. 12-13) that did not add up and I would like to see clarified (possibly by back-reference to EM).
First of all, Moglen is reported as saying that "You don't need permission to use a copyrighted work" as there is no specific exclusive "right to use" in copyright law (unlike patent law); strictly speaking, this is correct as no right of that type/name exists in copyright legislation anywhere that I know of. However, the conventional view is that, because we have to copy in order to use, the exclusive right of reproduction effectively serves to control use when it comes to computer programs: let me expand on this a bit:-
a) Apart from the basic matter of copying during installation, "conventional" theory has it that when a program is loaded from hard disk (or flash or other storage) to RAM (and, remember folks, the run-time object code version is protected as a copyright literary work: Apple v Franklin 714 F.2d 1240 (3rd Cir., 1983)), it is reproduced by that copy being created in RAM. This, in fact, is the mainspring that makes software licensing work - it is the authorisation to make *this* copy that approximates to authorisation to "use" under patent law and, hence, allows control over use via license.
b) There is, in the US, a big "BUT" that runs up against the theory - the US Copyright Act demands that copies be "fixed" (recorded in some form) in order to be deemed infringing (if unlicensed). OK, sixty-four thousand dollar question: can a temporary copy held in RAM really be said to be "fixed"? There are arguments about this *even now* but most US courts faced with this issue have held that even the temporary copies held in RAM *are* sufficiently "fixed" to count (see, for example, MAI v Peak Systems 991 F.2d 551 (9th Cir., 1993)).
c) In the EU, however, there is absolutely no doubt at all - Art. 4(a) of the Software Directive (91/250/EEC) clearly states that the rightholder has the exclusive right to do or authorise "the permanent or temporary reproduction of a computer program by any means and in any form, in part or in whole". End of story.
In light of all this, it is (fairly) clear that, contrary to what Moglen is reported as saying, users (whether end- or other) *do* need licenses. But why is that a concern here? Surely, the whole point about GPL is that, thanks to Section 6, it really lives up to its name by being, effectively, "viral" as each person in the chain of copying, modification and distribution gets a license on GPL terms. The only quibble then might be that the permissions granted to those users who *only* copy (as opposed to "copy, modify and distribute") are inadequately covered by the license; but, again, on any reasonable reading, I think and hope not.
But despite the international Conventions, all the national copyright Laws and Acts around the world still vary enormously (e.g. for reasons I won't bore you with here, it is not possible to assign [i.e. voluntarily convey/transfer] rights in literary, scientific and artistic works under Germany's Copyright Law 1965 (as amended) - the most a third party can get is an extended form of license called a "right to exploit"). Yes, even inside the EU, harmonisation is only partial (Copyright in the Information Society Directive notwithstanding).
All this has a real impact when we think (i) about the exclusive rights a rightholder has, (ii) what constitutes infringement, (iii) what permitted acts/free uses/fair uses there are and, consequently, (iv) how licenses are/should be structured. This is just the copyright side, though: what about national/regional variations in allied fields like contract law?
One interesting problem that arises from this comes about because of the "one size fits all" mentality that many US corporate lawyers have i.e. they draft licenses for the US market and then assume that they will work elsewhere. Big mistake: one of Microsoft's key OEM licence provisions on preventing diversion of media into retail channels was declared unenforceable in Germany by the Supreme Court not too long ago.
Gary R. Lea
IPRI Senior Research Fellow & Lecturer in Industrial Property Law
Queen Mary Intellectual Property Research Institute