Not to divolve this into a GPL discussion, but I promised a followup:
On Wed, 18 Oct 2000, Daniel Stenberg wrote:
> On Wed, 18 Oct 2000, Brian Behlendorf wrote:
> > Hmm, ghostscript had some problems along these lines, though, as Stallman
> > claimed that implementing the readline API and telling the user to link
> > it themselves (optionally, not even mandatorily) caused GPL taint.
> As someone has said, sometimes RMS simply must be making all that up as he
> goes. :*)
> The license simply cannot limit how people deal with the program locally. The
> license limits how they're distributed and re-distributed.
I asked L. Peter Deutsch for a link to the Alladin/GNU
ghostscript/readline issue; it's not online anywhere, but in the
documentation, so it's appended below. Whether or not RMS is making it
up, without an independent interpretation of the GPL by a court of law,
what he says is no more or less valid than anyone else.
> After all, that's how the Galeon people and loads of other groups have done
> when they release GPL stuff that depends on "non-GPL-compatible" licensed
> software pieces.
Galeon is technically in violation of the GPL, or MPL, depending on how
you look at it. If they had the permission from every author of every
GNU-written component to allow for a variance, then that'd be OK - it
would be the same situation as Linus explicitly allowing for binary kernel
modules to be loaded, even though strictly speaking it'd be a GPL
violation. However, Galeon's GPL components come from Gnome, so all of
Gnome would have to agree to weaken its use of the GPL by allowing it to
be linked to MPL-licensed code. I don't know how feasible that is, and
the MPL/GPL dual-licensing discussions taking place on the
mozilla-licensing mailing list suggest quite a few legal complexities with
such a model.
Anyways, here's the writeup.
Received on Sat Oct 21 14:36:12 2006