The language of efficiency is interesting in our field. We call something is "high performance" if some conceptual things are twiddling inside at high speed. I

The language of efficiency is interesting in our field. We call something is "high performance" if some conceptual things are twiddling inside at high speed. I guess this comes from the "engine performance" meaning of the word. Then there's the "employee performance" sense of the word, ie getting the right things done in a satisfactory way, and delivered on time. The latter mostly depends on product development acumen and high programmer productivity, etc. -- Regarding the specifc case of WebAssembly, it'll be interesting to see how WebGL compute and WASM will coexist. I guess in many applications it would be a good solution to have a compute kernel in both, with fallback to WebAssembly in case a device lacks WebGL 2 support or something. Is there compiler tech in the works that can target both?