Re: Automake: use of modified Perl modules & GPL

[Russ Allbery]

Richard Stallman <address@hidden> writes:

> In certain cases, where you know the licenses of certain modules, you
> can put in explicit statements yourselves, following statements from the
> author.  Otherwise, I suggest moving the problem modules outside of the
> public tree, and inviting their authors to resubmit them.  If you make a
> well-publicized announcement, you won't need to send mail to all the
> authors, because most of them will see the announcement.  This should be
> easy enough to be feasible.

I have a feeling that this suggestion may prompt a bit of a storm of
protest, so let me try to head that off at least to some degree by
pointing out that CTAN, the archive CPAN was based on originally, did in
fact do precisely that a year or so ago, as well as moving contributions
that weren't distributable via the TeX Live CD into a separate tree.  It
caused a bunch of protests at first, but it apparently made life way
easier for the folks burning CDs from the archive, and it seems to have
settled down now (and prompted a bunch of folks to clarify their
licenses).

This is not to say that this is necessarily the best approach for CPAN to
take.  Having the packager look for a LICENSE file at the top level and
just complain to the author if it isn't present may actually take care of
most the problem much more simply, if it's decided that it's a problem
that wants solving.

I'm very sympathetic towards people who would really much rather forget
about this whole licensing business, but I think folks who want to do that
should really at least slap a public domain notice or a "same terms as
Perl" notice on their work.  And I think most modules on CPAN *do* already
have a license; there just isn't any automated way at present that I know
of to extract that information.  It would be a nice bit of information to
offer on the search.cpan.org interface as well, since it can affect one's
choice of module for a project (however unfortunate that may be, it's
reality).

-- 
Russ Allbery (address@hidden)             <http://www.eyrie.org/~eagle/>