Am Thu, Feb 17, 2022 at 12:38:55PM +0100 schrieb Mattia Rizzolo: > > I suggest he should name the people that contributed most to PoDoFo > > and that should take part in this decision. > > I'll defer to who better knows Copyright laws and stuff, but from what I > concerned, the people that needs contacted are those who actually wrote > copyright claims, that for the LGPL parts of PoDoFo (src/* and some > tests/*, the rest is GPL-2+) it means: > > 2000-2016 Dominik Seichter > 2008 Craig Ringer >
This is wrong. Since the USA joined the Berne convention, explicit copyright claims are practically irrevelant and the actual changes matter. As such, you have to ask anyone who has contributed a non-trivial patch. > > I think in many of these cases we can take for granted that posting of > > small "trivial" patches into PoDoFo mailing list automatically makes > > these patches "public domain". > > Depending how small, such patches may very well not even be > copyrightable, so you can easily get away from those. Posting a patch doesn't make it public domain, but they can be trivial enough to not be copyrightable. Part of the problem is that a single patch can be trivial, but they can accumulate. IANAL, but the general guide lines given are ~7 lines of non-mechanical changes. But that can vary depending on the nature of the change as well, i.e. how complicated it is. > > Currently the license of pdfmm is LGPL 2.1 fixed[5], which it would make > > it incompatible to merge new files from pdfmm into PoDoFo > > I don't think it would. PoDoFo is LGPL-2.0+, if you merge LGPL-2.1 > stuff into it, the whole final work becomes LGPL-2.1, and that is not > considered a relicensing due to the way the "+" in the LGPL-2.0+ works, > afaik. Yeah, that's the excuse the FSF used when dropping the GPL 3. > > With one exception (libidn), all the dependencies have permissive (non > > coercitive) licenses: > > - libidn: LGPL -> FAIL > > > > I believe at some point we can just take a more permissive stringprep > > implementation written in another language/framework and port it to C++. > > Why is this not a blocker for going forward? > Wouldn't linking against libidn make the resulting library effectively > be under LGPL too? Only if you statically link. That's exactly why I consider this whole relicensing exercise overkill, even if I would certainly approve of the result. Joerg _______________________________________________ Podofo-users mailing list Podofo-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/podofo-users
