GPL compliance

[Gary W. Swearingen]

Mario Vukelic <mario.vukelic at dantian.org> writes:

> On Sun, 2006-07-02 at 10:50 -0700, Gary W. Swearingen wrote:
>> gain, very low risk,
>> but it confirms my long-standing distaste for "or later" clauses.
>
> [I have cc'ed sounder, please let's continue there ]

Sorry, it doesn't work that way.  You send your request and comment to
me and maybe sounder or a request-only to this list; you don't get
your 2 cents in _and_ ask the other guy to take his reply elsewhere.

But I got better things to do this week, so don't expect me to go on
much past today.

> When it says "you" it is talking not to the copyright holder, but to the
> licensee.

There are TWO copyright holders involved here (deriver and derived
from), but I agree that "you" is (are? :) the "licensee".

> Therefore when a GPLv3 comes out, the licensee can always fork
> off the last GPLv2 version, even if the maintainer accepts new code only
> in GPLv3.

Sure (except maybe that last phrase, which I don't understand), but it
misses my point.  I probably didn't make it well.  If the original guy
said "v2 or later" (before v3 exists, when it's X code is compatible)
and the second guy (your "licensee") later says "v3bad only" (after it
exists and when X code is incompatible, all hypothetical) he has
violated the terms of X (if not the first guy) when he publishes it
(even though he hasn't added any X code).  This is no sweat off the
"or later" user's brow, just the licensee's, so it's no big problem,
but it makes the point I said it was trying to make, that the original
X code is still under its original non-GPL license despite being part
of a derivative under GPL v2.