I approve/agree as well. On Feb 20, 2018 10:31, "Eric Blake" <ebl...@redhat.com> wrote:
> On 02/19/2018 05:20 PM, Bruno Haible wrote: > >> Hi, >> >> Often gnulib has relicensed modules for use in LGPLv2+ packages (such as >> libvirt). >> >> Some modules have also be relicensed for use in "dual LGPLv3+ or GPLv2" >> packages (such as GNU libunistring). [1] >> >> Today, I would like to ask for relicensing of specific modules for use in >> GPLv2+ programs, namely GNU clisp. >> >> GNU clisp is under GPLv2+, not GPLv3+, because it was designed, from the >> beginning, as a vehicle for running computer algebra systems. The most >> prominent computer algebra system written in Lisp is Maxima, and is under >> GPLv2. If clisp was only distributed under GPLv3+, one could no longer >> distribute maxima with/in clisp. >> >> So, regarding gnulib, I'd like >> >> 1) to introduce an option --gpl=v2+ that, like the --lgpl option, verifies >> license compatibility and updates the copyright header in the source >> files, >> >> 2) ask for relicensing LGPL -> 'LGPLv3+ or GPLv2' of the modules >> no-c++ >> mkfifo >> mknod >> strftime, nstrftime, time_rz, tzset >> >> 3) ask for relicensing GPL -> GPLv2+ of the modules >> c-strtod >> getloadavg >> link-follow >> libsigsegv >> vma-iter >> >> Thoughts? >> > > Makes sense to me. > > >> Can we go ahead, as usual, with per-module approval by the authors of said >> modules? >> > > Upstream ligsigsegv is still GPLv2+ (I'm actually a bit surprised that it > hasn't bumped to GPLv3+ in the meantime), so that one should be an easy > sell. None of the others are standing out to me as an obvious sore point, > so I hereby give approval for the proposed relicensing of my contributions > to those modules. > > -- > Eric Blake, Principal Software Engineer > Red Hat, Inc. +1-919-301-3266 > Virtualization: qemu.org | libvirt.org > >
