Mesa 13.0 Release Candidate 1 Arrives With Intel OpenGL 4.5 Complete

[phoronix]

Phoronix: Mesa 13.0 Release Candidate 1 Arrives With Intel OpenGL 4.5 Complete

Now that release manager Emil Velikov has worked through his LLVM issue, Mesa 13.0 Release Candidate 1 is now officially available as the first development release that has core OpenGL 4.4/4.5 support in Mesa...

http://www.phoronix.com/scan.php?page=news_item&px=Mesa-13.0-RC1-Released

[plonoma]

Regarding the OpenGL conformance test suite.
Wouldn't continuous development of the mesa drivers present a problem for the conformance process?
Wouldn't having to submit the mesa drivers and then wait for conformance cause large delays?
What if bugs are uncovered and bug fixes are developed during the submission process?
(don't tell me there won't be bugs in the near future, software is sadly never really finished.)
Wouldn't that put the mesa drivers conformance in an state of limbo?
Pushing distributions to ship the mesa version with the most bug fixes instead of the version with conformance?
Leading to the mesa versions used being most stable and no conformance but the conformant version has known bugs?

Wouldn't an elaborate, tedious, slow conformance process be a hindrance to get new code development and conformance?

[AdamOne]

Does this mean OpenGL 4 on Haswell?

[frosth]

AdamOne, unfortunately no. Maybe they decide that it's too risky to put it into stable release.

[dungeon]

Wouldn't having to submit the mesa drivers and then wait for conformance cause large delays?

Delays are for everybody, so who cares about delays

[lowenz]

mesa 13 rc1 package is missing a .h, wglext.h

[crymsonpheonix]

Regarding the OpenGL conformance test suite.
Wouldn't continuous development of the mesa drivers present a problem for the conformance process?
Wouldn't having to submit the mesa drivers and then wait for conformance cause large delays?
What if bugs are uncovered and bug fixes are developed during the submission process?
(don't tell me there won't be bugs in the near future, software is sadly never really finished.)
Wouldn't that put the mesa drivers conformance in an state of limbo?
Pushing distributions to ship the mesa version with the most bug fixes instead of the version with conformance?
Leading to the mesa versions used being most stable and no conformance but the conformant version has known bugs?

Wouldn't an elaborate, tedious, slow conformance process be a hindrance to get new code development and conformance?

Khronos conformance is one per driver per hardware. So for example, once i965 was conformant on Skylake, it's always considered conformant on Skylake. Otherwise no one would ever update their drivers once they passed conformance, since it would be too much of a pain to go back through conformance for every change (for both the devs and for Khronos).

I think that Intel also runs the CTS in their CI system, since being conformant is useful beyond getting a gold star.

[Marc Driftmeyer]

mesa 13 rc1 package is missing a .h, wglext.h

It's here:

https://github.com/mesa3d/mesa/tree/13.0/include/GL

[dungeon]

It's here:

https://github.com/mesa3d/mesa/tree/13.0/include/GL

Yeah git fixes something, but he said it is missing from package and indeed it is missing from package

https://mesa.freedesktop.org/archive/13.0.0/

[pal666]

Wouldn't continuous development of the mesa drivers present a problem for the conformance process?

what makes you think non-mesa drivers do not have continuous development?