Thread: chess bitrepresentation - OO question

Bitboard

It is well known amoung implementors of chess AI that in order to get it running at a decent speed, having the most efficient representation of boards possible, is critical.
As such it is necessary to forego the usual happy world of OOP, objects, heirarchies, and highly-readable-code, at least in terms of the bitboard level. That doesn't mean it's effectively obfuscated, it just means that you do your OOP at a higher level.

You can still do OO with bitboard representation. The board class uses an internal bitfield to represent the board. But when you place a piece on the board, the place piece function takes a Piece class. The Piece class could, for example, contain information about what numerical value would represent that piece, and the board could use that and set and appropriate bits.
When using the get piece function, the board could simply create a Piece object, fill it out and return it.
Yes, it may not ideal and it might not be the best solution, but it works. The only thing you have to do is find a clever way to "hide" or encapsulate the bitboard implementation into some usable interface.