Mesa 10.4 Has Been Branched - No OpenGL 4 This Year

[phoronix]

Phoronix: Mesa 10.4 Has Been Branched - No OpenGL 4 This Year

Mesa 10.4 was branched from Git master this weekend and that means the next Mesa release features only OpenGL 3.3 compliance and not OpenGL 4.0~4.2 as many had hoped...

http://www.phoronix.com/vr.php?view=MTg0MTE

[My8th]

Them being able to do what they have done in a short amount of time makes up for not reaching OpenGL 4 compliance. Everybody involved keep up the excellent work.

[89c51]

Is this race for GL4 compliance so important after all?? I mean things that the average user uses everyday need some attention also.

[nanonyme]

I suspect the cooler stuff (like DSA) will be done without waiting for previous version compatibility so there probably will be support for interesting parts of OpenGL4 up to 4.5 soon enough. After that it's just a matter of someone bothering to fill the gaps so we jump up quite a lot of versions up instead of going one by one. Iirc that's what happened with OpenGL3 too

[Mat2]

How much work is needed to complete the missing features?

Items left:

For GL4.0:
  GL_ARB_gpu_shader_fp64                               started (Dave)
  GL_ARB_shader_subroutine                             not started
  GL_ARB_tessellation_shader                           started (Chris, Ilia)

GL4.1:
  GL_ARB_shader_precision                              started (Micah)
  GL_ARB_vertex_attrib_64bit                           started (Dave)

GL4.2:
  GL_ARB_shader_image_load_store                       in progress (curro)

GLES3.1:
  GL_ARB_arrays_of_arrays                              started (Timothy)
  GL_ARB_compute_shader                                in progress (jljusten)
  GL_ARB_framebuffer_no_attachments                    not started
  GL_ARB_program_interface_query                       not started
  GL_ARB_shader_image_load_store                       in progress (curro)
  GL_ARB_shader_storage_buffer_object                  not started

Data from: http://cgit.freedesktop.org/mesa/mesa/tree/docs/GL3.txt

[nanonyme]

How much work is needed to complete the missing features?

Items left:

For GL4.0:
  GL_ARB_gpu_shader_fp64                               started (Dave)
  GL_ARB_shader_subroutine                             not started
  GL_ARB_tessellation_shader                           started (Chris, Ilia)

GL4.1:
  GL_ARB_shader_precision                              started (Micah)
  GL_ARB_vertex_attrib_64bit                           started (Dave)

GL4.2:
  GL_ARB_shader_image_load_store                       in progress (curro)

GLES3.1:
  GL_ARB_arrays_of_arrays                              started (Timothy)
  GL_ARB_compute_shader                                in progress (jljusten)
  GL_ARB_framebuffer_no_attachments                    not started
  GL_ARB_program_interface_query                       not started
  GL_ARB_shader_image_load_store                       in progress (curro)
  GL_ARB_shader_storage_buffer_object                  not started

Data from: http://cgit.freedesktop.org/mesa/mesa/tree/docs/GL3.txt

Maybe you would have better luck asking the people responsible rather than on Phoronix forum

[OneTimeShot]

Is this race for GL4 compliance so important after all??

Yes and no. Having excellent OpenGL 3.3 support is important, and a lot of applications and games target that kind of card. It is also a separate metric to the more important (for gamers) "how fast is my card?" numbers.

On the other hand, there is a new release of OpenGL each year, so basically Mesa is about 5 years behind. Which is roughly Windows Vista timeframe. This will have a knock on impact for developers who are already (maybe rightly) annoyed with OpenGL's driver support.

Fingers crossed that we'll see OpenGL 4.3 this time next year and OpenGL 4.5 in 2016. I doubt that Valve would bother supporting pre-OpenGL 4.3 in the Steam Machine if they intend to take on XBOne and PS4 as a game development hardware platform.

That's obviously before the fact that Mesa should currently be working on the "magic-gonna-fix-everything-and-gonna-make-game-development-somehow-trivial-honest" next version of OpenGL being developed by Khronos.

[scottishduck]

Is this race for GL4 compliance so important after all?? I mean things that the average user uses everyday need some attention also.

Tessellation and compute shaders are certainly a rather big deal from a developer standpoint and ultimately attracting developers is key to desktop linux becoming even remotely relevant.

[89c51]

Tessellation and compute shaders are certainly a rather big deal from a developer standpoint and ultimately attracting developers is key to desktop linux becoming even remotely relevant.

Linux will be relevant once it does what is expected to do for the average user. Developers will come only if they make money.

[przemoli]

Is this race for GL4 compliance so important after all?? I mean things that the average user uses everyday need some attention also.

Yes.

OpenGL 3.3 is good for porting DX9/10 software.

To port DX11 You need OpenGL 4.2. (And also all 4.x releases contain some improvements for porting process).

And also for fast & modern OpenGL one need AZDO which relay on few new extensions, those should be available before respective whole numbers, but game devs may just look for OpenGL 4.4/4.5.