OS X and the future of Open GL

[iBug2]

Do you guys think that Apple will keep adding new GL extensions with 10.11 or just stop porting of Open GL and focus on Metal for Mac OS X?

It seems like Open GL development is going the same direction as Metal/Mantle/DX12. A complete rewrite. So would it make sense to spend time on porting the remaining 4.X GL versions to OS X?

It should take some time before Open GL 5 (NG) is ready since it's a rewrite. Until then, Apple may be able to get Metal as sophisticated as Open GL for OS X instead.

 

[leman]

Do you guys think that Apple will keep adding new GL extensions with 10.11 or just stop porting of Open GL and focus on Metal for Mac OS X?

I think there are much better ways to utilise the resources. Bumping the GL version on OS X and adding new extensions is a waste of time IMO. What would the benefit be? This will not solve the deep-seated performance issues of OS X GL drivers and it certainly not attract more game developers to the platform. I'd prefer them to freeze the GL version on where its now and instead dedicate the resources to fix the plethora of bugs in the current OS software.

It seems like Open GL development is going the same direction as Metal/Mantle/DX12. A complete rewrite. So would it make sense to spend time on porting the remaining 4.X GL versions to OS X?

It should take some time before Open GL 5 (NG) is ready since it's a rewrite. Until then, Apple may be able to get Metal as sophisticated as Open GL for OS X instead.

OpenGL needs a complete reboot and I hope that the OpenGL Next will deliver what OpenGL 2.0 and Longs Peaks were supposed to deliver initially. I hope they will not mess it up this time.

I would certainly welcome something like Metal API for OS X, but it is unclear whether Apple has any plans for that. Metal would solve a lot of things and its mostly feature-equivalent to OpenGL anyway, except some minor details. If Apple decides to go this way, they could always leverage the OpenGL implementation on top of the native low-level API to support legacy implementations.