Possibility of using zink for Gles?

[foxtro...@gmail.com]

Hello, I was wondering the viability of using zink as the Gl renderer, While zink is not very performant at the moment, I believe it to be an important step for two primary reasons.

1. not in the near future, but eventually OpenGL support on hardware will be depreciated, or if not depreciated, severely underdeveloped and neglected. there are multiple pieces of consumer hardware already showing symptoms of this, albiet not in the x86 world. it is a matter of time, even if said time isn't soon. (Raspberry pi for instance, where vulkan 1.1 conformance is planned is a potential example), this will allow us to preemptively ensure backwards compatibility in the future.

2. As I am sure many people already know Virglrenderer is working on a vulkan backend and supposedly achieves very impressive performance nearly native. this is in stark difference to the OGL backend, which at best can achieve around 30% from my experience. Even with the overhead of zink in it's current state, and the small overhead of Virglvulkan, Preformance can be assumed to be higher than current implementation in the near future.

I think the combination of these is incentive to at least explore the viability of making sure zink works within android. While I do not know much, i'd at least do what I can to help debug, however I unfortunately do not have the ability to compile android atm due to hardware constaints