Three interesting items for the discussion here, concerning the
Contributor Agreement:
- "Eric Raymond presented his reasons why contribution agreements are
harmful, and there was extensive discussion on the point."
- "Daniel Brookshier moved to drop the requirement for a CLA for foundation
projects. Seconded by Chris Holmes. With the understanding that individual
projects can still use the CLA, the motion was carried without any
objections."
- "Eric Raymond took an action item to draft a statement for OSGeo on CLAs
and policy with respect to third-party IP rights and remediation thereof
for review by Rich, and the board."
(... yes, *the* Eric Raymond, it is good to see really experienced people
interested in this discussions.)
I assume that the second point will help with regard to the possibility
of relicensing of GRASS to MIT/X11 as discussed earlier.
Markus
On Mon, Mar 20, 2006 at 07:29:19PM +0100, Markus Neteler wrote:
Hi,
for those being interested to participage in
the #osgeo IRC chat about project incubation,
please participate. The laywer will be online
as well.
> > I assume that the second point will help with regard to the possibility
> > of relicensing of GRASS to MIT/X11 as discussed earlier.
> >
>
> What about GPL dependencies used in grass (e.g.FFTW) ?
Since FFTW is not "integrated" in the grass code, I think there is no
licencing problems.
If you link against a GPL'd library, the resulting binary is a
derivative work of that library and is covered by the GPL.
You can always make the source available under other licences in
addition to the GPL, but you can't withhold the source or only provide
it under terms which preclude redistribution under the GPL.
> > I assume that the second point will help with regard to the
> > possibility of relicensing of GRASS to MIT/X11 as discussed
> > earlier.
>
> What about GPL dependencies used in grass (e.g.FFTW) ?
The concern, I believe, was that by giving OSGeo control of the
copyright to the GRASS source code, others outside the GRASS project
could (if they decided to, which is doubtful) relicense GRASS as
something other than GPL. AFAICT this would be against the wishes of the
majority of the developers. Of course any code incorporating non-GRASS
GPL code would have to be removed if GRASS ever went non-GPL.
Since FFTW is not "integrated" in the grass code, I think there is no
licencing problems.
If a program is linked against a GPL library (such as FFTW), the program
must be licensed as GPL as well. In this case the module must be GPL but
libgis et al. could be LGPL. (although I'm not advocating that)