On Monday 15 June 2009, Nicolas Pitre wrote: > However, considering that the oldest version of OpenOCD as published by > Dominic Rath, the initial author, did contain support for ftd2xx > already, it would be hard to dispute the fact that this wasn't the > initial author's intention to allow this usage.
That doesn't address code from other sources (RedHat, FSF, etc) that merged ... and which was published under straight GNU GPLv2. Nor does it address the fact that the license explicitly DISallows that use *FOR DISTRIBUTORS* at this time. Recall that the initial code was for a diploma thesis, not for distribution. Or the fact that whatever the intent might once have seemed to be, the only legal agreement is saying "distributors may not do that". > And subsequent > contributors didn't complain about this either, so they could be > considered to have agreed implicitly to that exception as well. And > given the visibility and importance of the ftd2xx library usage (it was > listed in the documentation from the beginning, etc.) then those > contributors won't be able to claim in good faith that they didn't know > about the existence of that lib and its nature and retroactively revoke > their consent after all that time (in practice they actually might not > have noticed nor considered the issue and if they had known it then > might not have agreed to the situation but that's not easily > defendable). No, it's extremely easily defensible. They saw the license was GNU GPLv2, and they knew just what that means. It's hardly news to anyone ever looks at licenses. There would be no "retroactive revocation" going on, to just acknowledge reality by updating the docs so they don't suggest distributors may violate that license. (Any distributor with a brain already knew not to do that, but that doesn't mean anyone should encourage such violations.) This isn't even what I'd call "fine print" in the legal language. It's one of the most distinguishing features of the GPL. Don't forget that the issue isn't having a "you may optionally use <closed source library/> in personal builds" option. The issue is simply whether someone *DISTRIBUTING* binaries is allowed to rely on that library. And permission for that has never been granted, through the license, by any contributor. - Dave _______________________________________________ Openocd-development mailing list Openocd-development@lists.berlios.de https://lists.berlios.de/mailman/listinfo/openocd-development
