[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <200706141744.08408.bernd.paysan@gmx.de>
Date: Thu, 14 Jun 2007 17:44:07 +0200
From: Bernd Paysan <bernd.paysan@....de>
To: "Alan Milnes" <alan@...oundation.org.uk>
Cc: linux-kernel@...r.kernel.org
Subject: Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3
On Thursday 14 June 2007 16:08, Alan Milnes wrote:
> Agreed - if you want to take my work you are welcome as long as you
> contribute back your changes. That's the deal that GPL2 enforces and
> why it has been so successful.
That may be a side effect of the GPL, but it's actually not how the GPLv2
works (nor is it the intention). "Contribute back" means upstream. There's
no such provision in the GPLv2, you contribute only downstream. And there
are cases where you don't need to contribute at all.
E.g. the kernel hacking I'm doing at the moment: I have bought a uClinux
blackfin board, for testing my digital audio amplifier. For that, I took
one of the blackfin alsa audio drivers, and changed it so that it could
talk to my digital audio amplifier. I'm not distributing this software,
it's a complete in-house project, so I'm not obliged to contribute back. At
the moment, I'm the only person in the world who has both access to the
digital audio amplifier and the blackfin board, so releasing this driver in
that early stage is a rather pointless excercise.
I think this above explains fairly well the "misunderstandings" that are
appearing here. The GPL is not reflective (tit-for-tat), it's transient. If
there's a loop in the transient propagation, it becomes reflective through
the loop, but not by itself. This was the case in GPLv1, is the case in
GPLv2, and will be the case in GPLv3.
--
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/
Content of type "application/pgp-signature" skipped
Powered by blists - more mailing lists