Re: SPEC 2004 submission and GCC project

[Joe Buck <jbuck at synopsys dot COM>]

Laurent Guerby writes:
> It might be interesting for the GCC project to submit an up to date
> GCC to SPEC and try to get a SPEC license and the money, is the license 
> offer new this time?

Sigh.  I'm saying the following for information only; anyone wanting to
argue or flame about it, please find some other forum than the gcc list,
where you'll just waste the developers' time.  Remember that the
decision-maker here (RMS) doesn't read this list, so if you don't like
his policies, please take it up with him, not here.

I doubt if the GCC project can officially accept a SPEC license.  The
reason is that to do so we would have to officially agree to its terms,
which violate FSF principles.

Individual GCC developers can do so (any SPEC licensee can use it to
benchmark GCC and can publish the results if they follow the rules).
In my official work and at home, I do all kinds of things the FSF
would not approve of.  But when I'm acting as an SC member I do not.

Now, it's possible that RMS could be persuaded that official performance
benchmarks are somehow different from things like VMware licenses and an
exception can be made, but I doubt it.

On the other hand, the FSF happily accepts monetary contributions from
proprietary software companies, and even prints thank you's in the GNUs
Bulletin for them.  So I don't see why we can't give SPEC an updated
gcc and suggest that they donate the monetary equivalent of the license fee
to the FSF.