Simon Gardling
204ba85202
This made the code so fast, that the `board` benchmark became completely useless. So that benchmark was removed. Overall, we're looking at a futher 20% performance increase in `future_moves`
23 KiB
23 KiB