Re: Copyright lawyers analysis of Andreas Pour's Interpretation
Scripsit Raul Miller <moth@debian.org>
> > > However, from context, it looks as if you mean "binary" to mean "file"
> > > and not "program".
> > Excactly.
> The GPL doesn't require that you distribute all of the executable files
> for a program. However, it does require that you distribute (or make
> arrangements to distribute) all the sources for the program,
No. GPL.3 does not speak about "programs" at all. It talks about
"the complete source code" for the derived work. It then defines
source code as "the preferred form for making modifications to [the
derived work]".
That means that when I distribute a derived work in object code form,
the source code I have to distribute is the source code that is
necessary for making modifications to *the work I distribute*.
Not some superset work that you might imagine me to have published
instead.
> There would be little point to using the GPL if someone could just
> dynamically link it to a proprietary library which was available, free
> of charge, in source form, to anyone -- with the slight caveat that
> anyone who distributed modifications to that proprietary library must
> contribute everything owned by any author of the program to Bill Gates.
Perhaps. But I don't think the author of a program has the legal power
to prevent that. Well, of course he could offer copying permission on
contractual terms that said that the distributor must refrain from
distributing any proprietary work that somehow interacted with the
original work. But that would be damned close to discriminating against
a field of endeavour (creating proprietary software is a field of
endeavour), so if you want your license to be DFSG-free, you simply
can't have your cake and eat it, regardless of what the GPL says or
doesn't say.
--
Henning Makholm "In my opinion this child doesn't
need to have his head shrunk at all."
Reply to: