New week, new Rust! What are you folks up to?
I'm still trying to bring rendering time, but also e.g. temporary memory allocations down. Here an example from last week with the ambient occlusions algorithm:
Watch the first entry of Most Temporary Allocations and see below:
Basically the lowest part of the graph (in orange, left side) represents about 1 million temporary allocations which were prevented by just pointing into already allocated memory instead of copying parts of it ... unfortunately that can't always be done, but in this case it can ...
Trying a principled approach to fix potential busy-looping and poll starvation problems in