For the last couple of days I have been working on a rewrite of the AI system in my Javascript game
squares. Today I finished implementing the
Minimax algorithm with
alpha-beta pruning. Unfortunately, looking ahead more than 3 moves is beyond piratical. Right now looking ahead 3 moves (just from the start of the game) takes about 13 seconds. I gave the computer 30 minutes, but it never completed searching ahead 4 moves before I stopped it. There may be some optimization I can do, but the initial results do not look promising.
Took a short bicycle ride this evening and snapped a few shots to try some more HDR images. The challenge this time was the capital building, which is lite very brightly. Even at -4 f-stops it was washing out. But I did get a few proofs I like.
