Re: Patent clauses in licenses
Andrew Suffield writes:
> On Sun, Sep 19, 2004 at 02:41:03PM +0100, Martin Michlmayr wrote:
> > * Matthew Garrett <mgarrett@chiark.greenend.org.uk> [2004-09-17 10:05]:
> > > The GPL does much the same. If someone distributes GPLed software
> > > without complying with section 3 (which gives you various ways in
> > > which you have to make source code available to the recipient), then
> > > they lose the right to use that GPLed software. We have various
> > > licenses that terminate if you do something "wrong" - we've just
> > > come to the conclusion that it's acceptable that people not be
> > > allowed to do that thing.
> > >
> > > In the past, we've accepted various compromises on freedom because
> > > they help free software.
> >
> > I agree with this reasoning and think that we should treat at least
> > "Any patent action against the licensor connected to the licensed
> > work" as free. I'd like to hear more possible scenarios what "Any
> > patent action against the licensor" might mean in reality, such as
> > Nathanael's IBM example. I think such possible scenarios/examples
> > are a good way to think about the implications of these clauses.
>
> Here's a scenario for you:
>
> Company A releases a piece of software that includes this clause in
> its license.
>
> Company B releases a modified version of this software, that includes
> an extra feature.
>
> Company A has no interest or use in the piece of software created by
> company B; furthermore it desires to eliminate this version.
>
> Company A sues company B alleging that the extra feature in the
> modified version infringes some of its patents. Company A no longer
> has a license to the modified version, which it didn't want anyway, so
> it is not concerned about this.
>
> Company B cannot make counterclaims from its defensive patent
> portfolio, because that would invoke the termination clause and kill
> its modified version. Company B has no practical defence against this
> lawsuit, so the modified version is killed. They have been effectively
> trapped in a double-bind.
>
>
> I just pulled that one out of the air. There are countless more like
> it. All you are accomplishing is to permit copyright holders more
> control over their software; this cannot be a good thing. Trying to
> game the legal system *doesn't work*.
Company B's "defensive" claims also affect all other users of the
original software -- now that they attempt to enforce their patent
rights, no other users can assume themselves to be safe. Why is
Company B's self-defense more important to free software than
unrelated users?
> This is inevitable from first principles; significant arbitrary
> restrictions are non-free. You will always be able to find ways to
> abuse them to gain arbitrary degrees of control over the software.
One could claim that allowing others to access a program over the
network includes sufficient transfer of copyrighted material to
trigger section 3 of the GPL. Debian has rejected explicit "external
deployment" clauses in the past, but accepts the GPL despite this
possibility.
I just pulled that one out of the air. There are countless more like
it. Making up corner cases is not particularly useful.
Michael Poole
Reply to: