Synaptics MIT license, again

William Lahti xfurious at
Wed May 23 07:20:36 PDT 2007

Just my 2c on the matter... I think allowing GPL drivers is something
that should not be done (and I'm a GPL developer) because the legal
ramifications of linking is still there, even with's modularity.
However, drivers are in fact separable pieces of the distribution, and
should probably not be considered "part of the core" (as used as
reasoning for required MIT/X11 licensing). There's no reason the LGPL
(lesser GPL) couldn't be used, which tosses the whole "viral" stuff,
but that leads into what licenses are allowed and what licenses are.
If this happens then developers need to be much more careful to check
the license of the work they might use code in new works

Although it looks like the MIT relicensing for Synaptics is already
going to happen, if this discussion results in more liberal usage of
other FOSS licenses, then the drivers should be separated from
MIT/X11-licensed drivers in the source package, perhaps under a
'restricted' or 'licensed' folder. They could also remain synchronized
with the release schedule and core, but be in a different
package, like Xorg-drivers-restricted or Xorg-drivers-licensed. But
that's just if it ends up happening (if it's even a good idea). Like I
said, just my 2c.

On 5/23/07, Graeme Gill <graeme2 at> wrote:
> Matthew Garrett wrote:
> > Of course - or someone could provide some sort of argument more
> > compelling than "more people might possibly work on the driver". The
> > Synaptics driver isn't terribly important, but it's likely that there
> > are going to be more (L)GPL drivers appearing over time. If that
> > situation doesn't seem desirable, then people will actually have to make
> > some sort of statement as to why.
> The general argument is that it's not nice to offer
> contributions with a more (technically) restrictive
> licence than the existing body of code being contributed to
> (at least, not without agreement from the existing contributors),
> since this potentially "poisons" the whole project with those
> restrictions. The existing contributors may not have contributed
> their code if they had known that such extra restrictions could
> subsequently be applied.
> The analogous situation would be someone contributing a driver
> to the Linux kernel with a copyright license that added extra
> restrictions (eg. "You can't run this on Sundays", or
> "Not for commercial use").
> Do you think such a contribution would be accepted ?
> Graeme Gill.
> _______________________________________________
> xorg mailing list
> xorg at


long name: William Lahti
handle :: fury
freenode :: xfury
blog ::

More information about the xorg mailing list