gpgme license

[Marcus Brinkmann]

On Thu, Jul 25, 2002 at 07:33:43PM +0400, Max V. Zinal wrote:
> I think that this should be considered as a general and very common 
> mistake. The distance between `using' some work and `basing' your own 
> work on it is unsharp.

It's sharp enough to defend GPL'ed software when it is abused.

> One may say: GPL conditions _are_ absolutely 
> clear. But they are based on some special type of a software design, 
> which is very common, but not absolutely neccessary. There are no good 
> reasons for forcing GPL for programs that are `linked' with some 
> library, while allowing anyone to create some sort of GPL-covered 
> utility that acts as some intermediate interface between GPL-covered 
> code and proprietary software.

The GPL is not based on technical terms like linking.  Although we consider
linking to be one way to create a derived work, this is by far not the only
way to create a derived work.  Werner mentioned the process boundary, which
is often a good indication.  But a case can be made that if you write a
prorgam that communicates over pipes, command line or whatever with another
program (GPLed), and the first program was written to only and specifically
work with this other program (eg it won't do the job without it), and both
are distributed together, then we are already talking about a derived work
in the sense the GPL talks about it.

To repeat: If you write an intermediate interface, which is tightly bound to
a GPL'ed program on the one side and a proprietary program on the other
side, and distribute all of them together with the intention that the user
should use them in this way, then you are likely violating the GPL.  I think
there has been at least one case like this where the FSF has defended the GPL
successfully in the past.

There is a strong reason why the GPL does not talk about implementation
details like that.  The GPL only speaks about "derived work" without
specifying further what "derived" means to achieve _maximum_ protection under
any interpretation that the court is willing to follow.  Any further
specification of what "derived" means can only weaken the GPL, as a
"derived" work is already the broadest term available in copyright law.

Thanks,
Marcus

-- 
`Rhubarb is no Egyptian god.' GNU      http://www.gnu.org    marcus at gnu.org
Marcus Brinkmann              The Hurd http://www.gnu.org/software/hurd/
Marcus.Brinkmann at ruhr-uni-bochum.de
http://www.marcus-brinkmann.de/