Open achien opened 4 years ago
Oh, it is unlikely we re-search if the halfmove clock is lower because whenever the halfmove clock resets that's due to an irreversible move. So there might be some weirdness if we search some deeper lines first then search some shallower lines in the same position but I think that's unlikely to happen due to iterative deepening?
Right now the transposition table is not aware of the halfmove clock so we're almost certainly doing something non-optimal here.