Chess Program Development

Question

Hello,
I'm in the process of developing a chess program to be played over a distributed network. Due to the complexity of the problem I'm going to be starting with the very basics. I've read some of the previous posts on this forum about chess games and found them very helpful. I've already made a text-based Tic-Tac-Toe game, which was suggested as a good starting point on this forum, and I would like to do the same with the chess game to start with. I figured that this would enable me to concentrate on the game mechanics. Below is a brief outline of how I would like to go about it:
1)Game class - used to set up the game and alternate the turns between the players
2)Board class - maintains the current game state and where the pieces are. Checks for mates, check mates, stalemates and draws (after 50 moves) as well as illegal moves.
3)ChessPiece - this will be an abstract class which will contain information about the colour of the piece, whether it has been captured or not and its position on the board
4)The specific pieces themselves will be sub-classes of ChessPiece. This would contain information about what type the piece is, type of move it is capable of and how much it is worth.
Is there anything else that would be useful to include in these classes that would make development of the game easier?In a previous post, someone suggested representing the Board as ChessPiece[8][8]. Would this be a good idea? Would it be worth keeping track of which squares on the board are black and which are white?What would be a good way of checking for illegal moves efficently if I'm going to use an array?
Thanks in advance

Answer 1

representing the Board as ChessPiece[8][8]. Would this
be a good idea?Choosing how to represent the board is a trade off:
1) array[8][8]: looks like a human thinks of a chess board but can have a performance impact if you're writing an AI. Keeping x & y coords is simple but it complicates checking for when a piece can move off the board, particularly knights.
2) array[10][10]: same as [8][8] but this eases the range checking because you keep a ring of "out of bounds" squares around the board. Knights can still move off the board in one jump though, so you could use array[12][12] to solve this if you're not too worried about memory.
3) array[64]: a little faster than [8][8] but suddenly keeping (x,y) coords is trickier. Checking for out of bounds moves is a little easier though.
4) or you could use an array[100] (or [144]) to include the ring of out-of-bounds squares as in (2) above.
Whatever you do you should develop the API first before deciding which setup you want to use. IIRC most big chess progs use a straight arrays like 3 or 4, but that's because they're written in assembler ;-)