No subject

Under the hood, those drawing operations will render to a client-side surface instead of communicating to an X server to render everything.  This can be done any way the client wants, be it software, OpenGL, or any other way.  The plan is to enhance the drawing toolkits to use OpenGL paths where possible so they can be accelerated in hardware.

Wayland then uses OpenGL ES2 calls to composite those surfaces onto the actual framebuffer.  The reason why OpenGL ES2 was chosen is that it is a subset of OpenGL that is less expensive to implement in low-end graphics accelerators, like those that mobile devices have.

The demo applications use OpenGL directly only because the toolkit work isn't ready yet.

The major benefit of Wayland is simplicity and removing the X server as the drawing bottleneck.