Also the AZDO presentation has a problem which tends to show up in NVIDIA presentations, which is that it's only applicable to NVIDIA hardware and follows a lot of custom built fast paths. Nevermind that the features are unavailable, inconsistent, or just plain implode on competing vendors' drivers. (And sometimes this is explicitly stated IN the NV presentations O_O) And then they act like everyone is crazy to be building these other APIs because their custom fast path and extensions supposedly do the same thing.

None of which even begins to cover the total failure to support even basic advances in graphics on the ES side.

Honestly I don't get the OpenGL is smart vibe in this thread at all.

To map a buffer you still have to bind it (p. 65 core spec 4.4). You bind it to target and then you map the target. You must be joking.

The best thing that could happen to OpenGL is to die instantly and be replaced by a set of built-in kernels in OpenCL. That's terrible.