From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chris Sloan To: Carlo Wood , Bernard Dautrevaux Cc: "'rms@gnu.org'" , gcc@gcc.gnu.org, license-discuss@opensource.org Subject: Re: GPL and NDA Date: Sat, 21 Jul 2001 00:48:00 -0000 Message-id: <20010721004750.E28423@ghs.com> References: <17B78BDF120BD411B70100500422FC6309E2F4@IIS000> <20010720153615.A1158@alinoe.com> X-SW-Source: 2001-07/msg01469.html You say that it can work if you can atomically GPL and distribute it but that doesn't make any sense to me. By creating a dervied work from a GPL work, the new work is GPL'd from the time of its creation. You can't (AIUI) un-GPL it distribute it and then re-GPL it and expect that everything is ok. If for some reason the NDA + the GPL prevent the distribution when both in force, then it can't be distributed because you can't avoid the GPL in the situation described. Chris On Fri, Jul 20, 2001 at 03:36:15PM +0200, Carlo Wood wrote: > Imho, it works as follows: > > anything + GPL can not be distributed unless it is first GPL-ed. > If 'anything' disallows changing the license, then it can not be > distributed period. However, if the NDA allows B to perform the > following atomic operation: distibute it back to A and GPL it at > the same time, then there is no incompatibility with the GPL. > However, and that is certainly an interesting legal question, > can something like that be atomic? If you first GPL it and THEN > distribute it to A, then there CAN NOT be a restriction to also > distribute it to C. While when you first distribute it to A and > then GPL it, you violated the GPL. And even when the operation > can be atomic, A can not distribute it anymore of course - if he > does that then the orginal authors of the GPL-ed code as well as > B can legally request a copy - which would be GPL-ed. > Imho, this is only possible when the transfer from B back to A > is not considered a distribution as per GPL. > > -- > Carlo Wood >