New release process for 1.8 (READ THIS)

[Ping]

2009/10/2 Michel Dänzer <michel at daenzer.net>

> On Thu, 2009-10-01 at 08:26 -0700, Daniel Stone wrote:
>
> > There were a couple of different motivations for this, one was to make
> > building things a great deal easier
> Would it really? IME complications have mostly been due to things like
> protos, not directly between xserver and drivers.
>
I agree protos could sometime bring complications to the existing X server
architecture or API/ABI support, especially when we try to introduce new
features.  However, driver out-of-sync with server is one of the major
issues for on-going xserver releases.  Without testing the driver on the
xserver release candidate, there is no way for us, device driver developers,
to say for sure if our drivers will be compatible with the new release or
not.


> > What're your concerns?
>
> First of all, surely the onus is on those who want to move the drivers
> into the xserver tree to present convincing arguments in support of it.
>
I think moving drivers into xserver tree benefits both driver and xserver
developers as well as distro maintainers.  IMO (I don't know what those at
XDC's BoF were agreed on), moving a specific driver into xserver tree
doesn't mean it will automatically be included into xorg releases.  A
confirmation of compatiblity will have to be made for each release.  If the
driver's maintainer could not provide the confirmation on a timely manner,
the driver could be excluded from the xorg release. Eventually, if they
failed to do so for certain times, it could also be totally removed from the
xserver tree (hopefully the xserver maintainers will give the driver
maintainers a reasonable time to react, just like kernel.org gave enough
warnings to Microsoft for updating their code).


> That said, e.g., what will happen with drivers which can't or don't want
> to be in the xserver tree? Like GPL input drivers, or the Gallium Xorg
> state tracker, which really wants to live in the Gallium tree at least
> for the time being. It seems like they would become second class
> citizens in the best case.
>
I don't worry about being a second citizen (my driver is GPL'd).  I don't
think anyone in the "second citizen" group has the right to complain about
it since it is their choice to stay outside of the tree.  I think we are
still under the same roof (xorg) but a different room (their own
repository). My concern was I am outside of the building.  No one at
xorg cared about if my driver was in sync with the xserver or not. I've
added "tons of" (exaggerated :) #ifdefs to catch up with so many xserver
releases.  Getting rid of that overwhelming feeling is more important than
becoming a "citizen" for me.

Ping
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.x.org/archives/xorg-devel/attachments/20091002/4a02fab4/attachment.html