Re: RFC: proposed GPLv3 license exception draft

[Ralf Wildenhues]

Hello,

* Robert Collins wrote on Wed, Apr 22, 2009 at 02:41:31AM CEST:
> > 2. No Weakening of Autoconf Copyleft.
> > 
> > The availability of this Exception does not imply any general presumption
> > that third-party software is unaffected by the copyleft requirements of
> > the license of Autoconf.
> 
> I have two comments:
> 
> Clause 2 seems like something that *should* be provided by the GPLv3
> itself, or else all exceptions will need it, won't they?

I think it may be a simple legal requirement that a statement of the
form "this exception E to some requirement R does not imply that other
reasons that R may hold are void" accompanies E not R.  IANAL though,
this is pure speculation on my part.

> Should we have a 'drafting an exception' guidebook somewhere.

I'd say "have a lawyer do it for you" is a good guidebook.
Really, I don't think anyone can seriously recommend otherwise.

> Secondly, I wonder if the definition for EOM could be a little more
> precise. Something like 'EOM consists of the helper scripts [x, y, z],
> and the minimum configure script that can be output by autoconf to
> configure a project.

Well, it would not be good if the license would need to be changed for,
say, every other Autoconf version, due to some technical details that
changed.

> Secondly, I wonder if the definition for EOM could be a little more
> precise. Something like 'EOM consists of the helper scripts [x, y, z],
> and the minimum configure script that can be output by autoconf to
> configure a project. I guess I'm saying its not clear to me that saying
> 'minimally verbose non-debugging non-tracing' is sufficient - if someone
> adds a non-debugging, non-tracing non-verbose mode that sucks in
> autoconf evalution code to the output, it would be outside the intention

I guess you're saying here that the definition is not strict enough to
prevent abuses, right?

Cheers,
Ralf