I just read “Google Continues Working On “Magma” For Mesa Cross-Platform System Call Interface” on Phoronix and didn’t get it. That made me realise my knowledge and understanding of these things is barely existent. I did write an MS paint clone on linux in C++ a really long time ago and the entire thing was with opengl (it looked like crap), but since then… nothing.
So my understanding is that the graphics card (or CPU if there’s no graphics card), writes to a component which is connected to a screen and every cycle (every 1/60 seconds if 60Hz) the contents are sent or read by the screen. OpenGL provided a common interface to do so, but has been outdated since… a while and replaced by Vulkan. Then there are libraries either built on top of are parallel to OpenGL. Vulkan can be parallel or use OpenGL if that’s the only one supported IIRC.
However, I’m not sure if OpenGL is implemented at the hardware level (on the graphics card), software level, or both.
Furthermore, I don’t understand where Magma, Meta, and MESA come in.
Maybe my core understanding is wrong or just outdated. I can’t tell. Can anybody eplain?
I think it’s more like Vulkan was a Metal knockoff. Metal released (for mobile) June 2014 and Vulkan research kicked off July 2014. Vulkan was only announced March 2015, and I would think it took more than three months of work for Apple to release Metal for macOS June 2015. And then Vulkan’s specs and SDK were released in February 2016. Though I doubt Apple was pushing for Metal to become an open cross-platform standard either.
AMD donated the Mantle API to the Khronos group (which Apple has been a part of since 2008 focused on OpenCL) and that group developed Vulkan. Of course Apple has a proprietary version. They have their own silicon, own OS, why wouldn’t they have their own graphics layer? Neither of them are knockoffs. Vulkan is open source and widespread for many applications, Metal is proprietary, and applies purely to Apple OS’s.