Re: Termination clauses, was: Choice of venue
Glenn Maynard <firstname.lastname@example.org> wrote:
>On Tue, Jul 13, 2004 at 09:17:51PM +0100, Matthew Garrett wrote:
>> I'm arguing that I don't believe them to be obviously non-DFSG-free,
>> which is not the same thing.
>It seems so obvious and self-evident to me that it's actually difficult
>for me to argue it. I'll try.
>It's a superset of every other restriction. The author may decide he doesn't
>really want people to modify the software, and terminate the license for
>everyone who does so; then he's saying "you can modify and distribute this
>software, but you can't actually do so". He might even wait a couple of
>years, letting people become strongly invested in the software before doing
At the point where the termination clause is used, the software is
obviously non-free. I'd argue that this is directly analagous to the way
we deal with patents. Almost all software we ship has the sword of
patent suits hanging over its head, and could become non-free at any
time as a result. We accept this because there's no way we can indemnify
our users. RMS has claimed that failing to comply with the GPL means
that your license is effectively terminated, even if you cease doing so.
Even an accidental breach of the GPL could result in the copyright
holder contending that you no longer have any license at all.
So, really, we can't prevent the cases you've described. Based on the
number of patent suits I've seen reported compared to the number of
license terminations I've seen reported, being worried about termination
clauses when it's far more likely something else will bite the user just
as badly seems unnecessary.
>While one could interpret the DFSG to say "freedom to modify and distribute,
>unless the author changes his mind", I don't think such an interpretation
>is useful. It's certainly not in the spirit of the DFSG.
There's already an implicit "or there's a nasty legal situation" in
there, along with a couple of other exceptions.
Matthew Garrett | email@example.com