Re: GCC 3.0 released

[Florian Weimer <fw@deneb.enyo.de>]

Ted Dennison<dennison@telepath.com> writes:

> In article <871yohih9w.fsf@deneb.enyo.de>, Florian Weimer says...
>>
>>Yes, that's true.  Shall we guess what is causing problems?  Legal
>>matters (i.e. the copyright assignment to the FSF)?
> 
> That can take a while.

Robert Dewar told me via email that this is not the problem.  The
copyright was executed years ago, and apparently it is not necessary
to assign copyright for the components which are not directly part of
the compiler.  

> I'm trying to go through the process with the SETI@Home Service, and
> its been incredibly slow going.

Well, it happens quite fast if you've contributed some code to GNU
Emacs through some backdoor without signing any document. ;-)

OTOH, the copyright assignment stuff really slows down development in
some areas.  For example, at the moment, our university cannot donate
code to the FSF because of legal matters (the assignment contract is
completely incompatible with German law).  In fact, someone already
partly reimplemented a security fix for GNU Emacs because of this
problem.  As a result, we do not work on improving software (beyond
our immediate needs) any more unless contributing something does not
require any paperwork.  (Forking is not an option of course, because
of the additional work required.)

> So it does look like most of the stuff in there is already FSF, and
> I'd guess (just my guess) the rest isn't FSF *on purpose*, and will
> probably remain that way.

Yes, I think so.

> Which begs the question (and this is just genuine curiosity), what
> exactly is going into the official baseline when it gets merged?
> Will it just be the FSF code? (probably not) Just the GPL'ed code?

The latter is not an option, I think.  After all, you need quite a bit
of source code which is covered by the ARM copyright in order to be
able to compile typical Ada programs.