For Approval: MLL (minimal library license)

[Chris Travers]

On Nov 13, 2007 12:45 PM, David Woolley <forums at david-woolley.me.uk> wrote:

> I basically made the assumption that there was something valid in the
> proposal, and, as the proposer seemed to be assuming that the referenced
> BSD version was GPL incompatible, and the only incompatibility I know of
> between GPL and a BSD variant is the advertising clause, I just assumed
> he was referring to a version with that clause.

That of course makes the assumption that there is no conflict between
the GPL v3 permissions removal provisions and the BSDL.  (There seems
to be a near concensus that one cannot remove permissions on a mere
literal copy of BSD code.)

Note that the MITL may not be safe in this matter either (as it
addresses all downstream users) so....

Note that these are not abaolute problems and only raise compatibility
questions where whole files are included verbatim, not when code is
mixed in a single file by my reading.
>
> > So what is all this about?  The 3-clause BSD is just as GPL-compatible
> > as the MIT/X license.
>
> Yes.  It does seem even more useless than it already appeared.

I think it gets back to misunderstanding about what sublicensing
really is.  Because the MITL variant on the OSI site lists
sublicensing (the MITL version from MIT distributed with Kerberos does
not, interestingly), people assume that one can remove permissions
from mere copies.  I am not sure that is the case either.

BTW, the SFLC's official recommendations are that licenses should
*not* be changed on permissive-licensed files included verbatim, which
would seem to raise questions about applicability or limitations on
the permission removal clauses in the GPL v3.

As for FUD, there is a lot of *reasonable* fear, uncertainty, and
doubt about the GPL v3 in large part because the license is so long
and complicated, and some sections seem to be asking for trouble (such
as the permissions removal section).

Best Wishes,
Chris Travers