Re: Was Re: KDE not in Debian?
Chris Lawrence wrote:
> On Jan 29, Andreas Pour wrote:
> > > (3) real permission to distribute from the authors.
> > I do not quite know what you mean by this, but if you mean that to
> > conform to your practice noted above of confirming from package
> > authors that packages can be distributed by Debian, I will see if I
> > can get the core KDE developers to send you their approval that you
> > distribute KDE code. Mail me privately please if you think it is
> > worth any effort and I will get started on it.
> It's a bit more complicated than that with some of the KDE software,
> because the GPL does not technically permit the linking of software
> against libraries that have licenses more restrictive than the GPL
> (like the QPL, which has restrictions on for-profit use on Win32). If
> all of the code in a particular KDE app is written by KDE members, all
> we need is something like:
> KFlarg is (C) 1999 Foo and Bar. You may use and distribute KFlarg
> under the terms of the GNU General Public License, version 2 (or a
> later version, at your discretion). As a special exception, you may
> also link KFlarg against the Qt widget library.
Oh, this is an old point. I thought for a moment we were breaking new ground. .
> (I forget the exact phrasing we decided was appropriate; but, some
> stuff that uses XForms in contrib uses it. I do know the "correct
> version" is longer).
> However, there are several instances of software in KDE being
> repackagings of existing software, with some work to integrate it into
> KDE (I am told KGV fits into this category). In these cases, because
> not all of the work was done by the KDE group, we also need permission
> from the author of the original software (GV in this case) to link
> against the GPL-incompatible software.
> [Personally, I think if you wrote the software yourself and link it
> against Qt, it's pretty obvious from a legal standpoint that you
> accept people linking it against Qt.
I think so too. So why not just exclude kgv and kfloppy and distribute the
> However, if you take someone
> else's software and do the same thing, I can't see how we (or anyone
> else) can interpret that as "acceptable". A lot of people have made a
> possibly erroneous assumption that authors like that of GV won't
> consider linking against Qt an abuse, and there are plenty of fat
> targets out there for a lawsuit (Corel, Red Hat, Caldera...).]