This reminds me of when Richard Stallman cracked down on CLISP (before it went open-source) for being distributed with, and linking to, a GPL-protected library

This reminds me of when Richard Stallman cracked down on CLISP (before it went open-source) for being distributed with, and linking to, a GPL-protected library of some sort. The legal argument of the distributor was that he was providing the library, and it was the user that was linking it. Stallman checked with a lawyer, and apparently the law treats that as a "subterfuge" and that you're "really" distributing only one program, which has only been "disguised" as two (quotes from [1]). IANAL, but I would assume that judges would follow similar lines of reasoning in this implementation, that for practical effects you are "really" distributing a copyrighted file, even if the distribution process at one point happens to involve only a pair of statistically random data[2]. In the limiting case, they can just say that the two files comibined constitute a distribution method, and you'd be hard-pressed to argue that. [1] http://clisp.cvs.sourceforge.net/*checkout*/clisp/clisp/doc/Why-CLISP-is-under-GPL http://clisp.cvs.sourceforge.net/*checkout*/clisp/clisp/doc/... [2] Note that while either file on its own is technically random in a statistical sense, they are not independently random. Even if you XOR in Firefox, the data is only pairwise-independently random, not completely-independently random. So the random argument actually falls on its face. [N] It also made me think of this, even if it's not directly related: http://xkcd.com/538/ http://xkcd.com/538/