On Thu, Mar 04, 2004 at 05:12:34PM -0500, Evan Prodromou scribbled: > >>>>> "MH" == Marek Habersack <grendel@debian.org> writes: > > Me> Sorry I'm talking down to you [...] > > MH> Oh, I'm used to dick wars here. And I will say it right away - > MH> your dick is bigger. > > OK, you got me. Zippers up. B-) :) > MH> It was a reflection based on the quoted gnu.org site. It was > MH> also induced by reading the Debian's mozilla copyright file > MH> which is vague in stating that some files are (solely) > MH> licensed under NPL or MPL (which, if true, would create a > MH> problem if GPL code was used from within those files). > > So, if I follow you, your question _isn't_ about emacs and python > being shipped on the same Debian CD. That's fine. Yep, no doubts here. > The question is that the NPL isn't compatible with the GPL, and you > think parts of Mozilla improperly link to or otherwise incorporate > stuff that's under GPL. Or that they _could_ do that, and you're > wondering what the maintainer should do in that situation. Is that > correct? Precisely > AFAICT, that's an upstream license problem. That doesn't make it not > Debian's problem, but still. If, on the other hand, the _DD_ is the Debian is the party that'd violate the license in that case, though. If, let's assume, there is a conflict between libfreetype6 and Mozilla, then debian should not link mozilla with freetype. > one doing the incorporating (say, by adding a GPL'd add-on to a piece > of software with an incompatible license) that's a bug that should be > fixed by the DD. > > A third, thornier possibility might be a package A that incorporates > code for feature B and code for feature C, but the licenses for B and > C are incompatible. I think that's a build issue: the DD has to choose > B or C but not both. That's what made me suspect Mozilla might be a problem (again, based on its debian copyright file). I glanced at the Mozilla sources and it seems that most of the files are triple-licensed, so no worries there (but the changelog file should be updated nevertheless) Although all that gave food for thought and created the theoretical (and unclearly formed) question I posted :) marek
Attachment:
signature.asc
Description: Digital signature