Re: what is the current status of GPL v3

[David Kastrup]

Alexander Terekhov <terekhov@web.de> writes:

> David Kastrup wrote:
> [...]
>> It sure is part of any GNU system, in the form of an aggregation (in
>> the case of GCC).  The C library, however, is linked with the
>> executables, and that exceeds mere aggregation.  The C library,
>> however, is licensed under the LGPL.
>
> Function calls have no impact on copyright status of combined work.

The law begs to differ.

> It's still a compilation ("mere aggregation" in GNU speak), not a
> derivative work ("modified work as a whole" in GPL 2b speak).

We are not talking about merely "function calls" in Glibc, but mostly
a standardized interface (POSIX).  Where you can drop-in replace the
library with an existing independent one, aggregation is a plausible
interpretation (there has been a readline clone mostly for legal
purposes IIRC, and the stance of the FSF towards Motif interfaces has
changed with the availability/feasability of Lesstif, even before
Motif got opensourced).  Standard compliant libraries from the GNU
project tend to be LGPL, and I guess that is because the alternatives
could force the FSF into undesirable lawsuits with uncertain outcome.

However, for nonstandard libraries and code bodies, the FSF feels
comfortable enough about its interpretation of the legal situation
that they are usually licensed under the GPL (as long as the FSF is
copyright holder).  Up to now, nobody was willing to have this tested
in court, so it seems they are not alone with that interpretation.

Companies like Trolltech live from the proceeds on the dual-licensed
GPL/propriatary Qt toolkit, so it appears that nobody is willing to
bet against Trolltech here, either (and Trolltech is not exactly a
large shop).

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum