the project is in early stage. there is only a very basic ai.

playing: moves are entered in standard algebraic notation. see the godoc for more information. use gochess -h for more help on command line parameters.

things implemented so far: * negamax search with alpha-beta-pruning * quiescence search * transposition table (4 buckets per entry) * iterative deepening * killer heuristics * a very basic position evaluator (using material score and piece-square-tables) * fen, san, epd, coordinate notation parser and printer * test suites for measuring engine strength and to ensure code correctness * several tools - epd suite solver, perft calculator * console and uci interface

missing things: * draw by threefold repetitions * draw by fifty move rule

bugs: * mate scores are not correctly stored into the transposition table

current activity: * improve uci support * refactor the interface to the search (text/uci) to reduce code duplication and to make writing tests easier * apply more rules for position evaluation (pawn structure, king safety...)

things to do: * improve move ordering by implementing new heuristics and tuning the move generator * store pv in a dedicated data structure * null move pruning * principal variation search/internal iterative deepening