OpenGL 4.6 Appears To Be On The Way

[nanonyme]

In addition to this they made claims about contents of ARB stuff which is still subject to change

[dungeon]

Well who cares, it won't be part of mesa until someone reviews it anyway... everything under development is subject of change, patches, communication, etc... nothing devs do communicate during development is official anyway In my book stuff from git is unofficial software, only releases are official

But even that isn't official OpenGL, since only conformance passed OpenGL 4.6 drivers could be official isn't it, etc... so until that happen they can talk whatever they want there

[schmidtbag]

Yeah. I'm betting intel will be first to support 4.6 under linux.

Honestly, AMD could be just as likely. If 4.6 really is just a collection of existing extensions (plus maybe a few new ones) both Intel and AMD are roughly on par with each other.

[Wilfred]

[...]
But even that isn't official OpenGL, since only conformance passed OpenGL 4.6 drivers could be official isn't it, etc... so until that happen they can talk whatever they want there

Aren't they opening up the tests? That is, anyone can run them?

[smitty3268]

Honestly, AMD could be just as likely. If 4.6 really is just a collection of existing extensions (plus maybe a few new ones) both Intel and AMD are roughly on par with each other.

I'd actually say AMD is currently ahead, particularly if they try to include bindless textures into 4.6. That's a big extension for Intel to try to add in, and they're missing several other AZDO extensions as well.

From features.txt, Intel is missing:

• ARB_compute_variable_group_size
• ARB_indirect_parameters
• ARB_shader_ballot
• ARB_shader_group_vote
• ARB_sparse_buffer
• ARB_bindless_texture

In comparison, the ones AMD is missing are:

• ASTC & GLES3.2 - probably unlikely to get added as a requirement to desktop GL since 99% of the hardware doesn't support it.
• ARB_post_depth_coverage
• ARB_transform_feedback_overflow_query
• KHR_blend_equation_advanced and _coherent