X Server 1.9 Is Scheduled To Be Here In August

[phoronix]

Phoronix: X Server 1.9 Is Scheduled To Be Here In August

X Server 1.8 was only released a few days ago, but Keith Packard is quickly focusing his attention to X Server 1.9, where it looks like he is interested in being the release manager once again...

http://www.phoronix.com/vr.php?view=ODEyOA

[curaga]

The X efforts are really improving. Props to Keith.

[Sadako]

So they're really going ahead with packaging the drivers with xorg-server?

Ugh.

I can see how it might make sense for the various input drivers, but the gfx drivers?

Makes me glad that so much is moving out of the ddx drivers and into the kernel and mesa parts via KMS and gallium3d...

[DeepDayze]

So they're really going ahead with packaging the drivers with xorg-server?

Ugh.

I can see how it might make sense for the various input drivers, but the gfx drivers?

Makes me glad that so much is moving out of the ddx drivers and into the kernel and mesa parts via KMS and gallium3d...

of course only the FOSS drivers would be included in the xorg-server package

This is reminiscent how Xfree86 packaged the drivers in the old days.

[disi]

The X efforts are really improving. Props to Keith.

Jupp, the upgrade from hal+1.7 to udev+1.8 were really smooth Works like a charm on my desktop and laptop. Can't wait to see what comes with 1.9...

[rvdboom]

I remember it was thought that with chip-specific code being increasingly moved to KMS and Gallium, in the end, a generic DDX driver could be used.
I guess that at this point, it does not make any sense keeping it as a separate package.
Is this inclusion in the X server the first step toward this kind of goal?

[FireBurn]

I remember it was thought that with chip-specific code being increasingly moved to KMS and Gallium, in the end, a generic DDX driver could be used.
I guess that at this point, it does not make any sense keeping it as a separate package.
Is this inclusion in the X server the first step toward this kind of goal?

They've changed their mind about that now unfortunately :-(

[bridgman]

I *think* the issue is that you still need chip-specific interfaces between ddx and drm for modesetting, so the ability to standardize acceleration via Gallium3D isn't enough on its own.

[murraytony]

Why can't they just find a happy medium (as regards to the modular/monolithic debate)?