Subject: Re: brands, trademarks, and the GPL
From: "Andrew C. Greenberg" <werdna@gate.net>
Date: Sun, 3 Oct 1999 23:03:05 -0400

> >     I agree.  What you can do, however, by making the GPL more permissive,
> >     is to destroy businesses that rely on the propagative and restrictive
> >     nature of the GPL.  You can't cause a hardship for people who benefit
> >     from the GPL's *permissions*, but you can cause one for people who
> >     benefit from its *restrictions*.
> >
> > That would be possible.  However, the GPL says that new versions have
> > to be similar in spirit, and differ only in detail.  That was designed
> > to give you an out, if there were drastic changes
>If I adopt the "or any subsequent version" language, it seems to me that if
>FSF were to adopt a dissimilar-in-spirit GPL at some time in the future, the
>"similar in spirit" clause only gives me an out if (1) I take FSF to court
>for violating their promise to change the GPL only in similar-in-spirit
>ways, and get the revised GPL retracted, and then (2) I contact every person
>who may have distributed my work under the dissimilar-in-spirit GPL, and
>either persuade them or take *them* to court to get them to stop doing it.
>That isn't much of an out.  I agree it's a heartwarming statement of
>intention, but I think it has almost no useful force in practice.
>
>Do any of the attorneys on this list have opinions about this (other than
>"maybe", that is :-))?

But, of course.  The answer to this one is, "it depends."  It depends 
upon the particular language and the particular changes.  There are 
analogous issues, and a fair amount of case law, when adopting choice 
of law in international transactions, such as the UNCITRAL terms (the 
international analogy to UCC, sideways sort of).  In practice, the 
consensus on UNCITRAL versioning is that its a bad idea to rely on 
the enforceability of a malleable agreement -- the other side can 
always argue there is a failure of consideration since the entirety 
of the quid pro quo can be removed in theory if a version change is 
or could be unsatisfactory with respect to any material term.

In short, I haven't thought deeply enough about the issue.  My 
general advice would probably be to pick a license version and stick 
with it -- not so much to avoid FSF misbehavior as to ensure the 
enforceability of the earlier version provisions.