Stephen J Baker wrote:
>
> On Wed, 15 Sep 1999, Keith Whitwell wrote:
>
> > I think that the long development cycle for 3.1 was justified (I have to say that
> > really...), but it'd be nice to cycle more rapidly for the next few iterations.
>
> Yes - the dev cycle for 3.1 was justified - but it highlights
> the need to go to an odd-even development cycle.
>
> > I've actually got a fair chunk of the Mesa core work done, so
> > it should be pretty much ready to drop into the 3.2 as soon as
> > that work starts.
>
> Could I plead for 3.1 to progress to 3.2 as soon as it's stable,
> then to maintain 3.2 as the "stable" release with new dev going
> into 3.3 (kinda like the Linux kernel dev advances). That somewhat
> removes the pressure to cycle major versions quite so quickly.
>
> There were quite a few simple fixes that could have gone into 3.0.n
> that have been delayed for too long while we wait for 3.1. to appear.
I like this too. The display-list stuff is pretty simple, but not simple enough
to push into Mesa just before a one-time-only stable release. With a stable
*branch*, this sort of stuff could be back-ported with ease to satisfy the Xracer
guys for 3.2.small.
People tell me there are problems with TuxAQFH. Has anyone distilled these
problems to a soundbite sized description?
Keith
_______________________________________________
Mesa-dev maillist - [EMAIL PROTECTED]
http://lists.mesa3d.org/mailman/listinfo/mesa-dev
