Sorry your browser is not supported!

i placed a simple chess game (2-player) on my website with source. There is no AI code, its multiplayer only. You can play offline "hot seat" or use my included matchmaker to locate other players. I haven't done any instructions yet, i'm tired now.

the source can demostrate one method of 3D object selection.

also simple chat code

board update techniques

If you are not sure about how to use the matchmaker, check out the notes from DUEL, it works the same way.

(for those wondering about how to acquire the game, simply click the 'web' icon underneath his post)

I'm looking into it now, it sounds good .

good A.I for the chess game would be a pretty impressive achievement!

Chess AI is easy to understand, a total bitch to impliment though because the memory management is crucial.

Simple rules for chess AI - give each piece a score, basically the number of spaces it can move to from any location denotes the pieces value. Then you use this to calculate a status score, like simply calculate the value of all your pieces not under threat - you can get clever with this to improve the AI - for example adding bonus points for pieces that have an opponents piece under threat. Now, for every possible move, you have to store the move and the resulting status score. Then you do this for every possible move after every possible move after every possible move until you die (wheeze) - that's where the challenge lies - organising all that data. If you decide that the most moves it thinks ahead are say, 10 (which is far more than most chess players anyway) - you can work that way, but you have to consider the opponent too. I actually started a DB chess game ages ago - I calculated all the moves, then calculated all the opponents moves and had them in a list, I simply chose the most obvious opponent move and branched off like that. I used text strings for the moves in the predicatable style A2-A4, this is good because you can build the board using a string if you keep it going, so by the time the AI has decided, you could have a move string of ''A2-A4:a7-a6:A4-A5:b7-b6:C2-C4:d7-d6:C4-C5:d7-d6:........'' All the way upto 255 characters (or about 25x2 moves) just to get 1 AI move.

Email me if you'd like to discuss this more or take a peek at the source I did get done.


Van-B

lol, you got that right! nice solution though.

Look under chess at www.howthingswork.com to find some nice resources on chess ai's.

thanks for all the interest. I have done chess AI before, but that was (cough) before windows (cough). Here, the intention was to test out multiplayer concepts in a simple game. A complex game would be RTS or something with hundreds of moving units. Duel has only 2 moving units (per player) and chess has 32 so i send the entire board upon turn complete. If chess were done as a RTS, I would need a better way to handle movement. AI is not really on my todo list. Next up is play via realtime FTP which really sucks and has many connection issues, but does not suffer from firewall issues. Its almost ready for release. If interest in this continues, I will update chess to permit 4 player chess with multiple rule options.

Or, maybe permit rule options with the 2 player version.

Otherwise, upon version 2's release, I'll stop working on chess and move on to Archon, a "card-based" FPS deathmatch.

Think No-Yu-Go where when you play a creature card, you become that creature and fight other players as in Duel, but each player is a different creature with different attacks and bonuses.

But, as others and myself have said, before jumping into a huge project, test out concepts in a simpler game first.

Duel tests out the FPS methods and simple effects
Chess is testing out FTP and game board updates
Both serve to test the matchmaking kit. Important if i actual want
people to play my games and find each other.

Sure, there are conceptual flaws with each game. Duel would have been better done if each critter in the game had its own memblock instead of all sharing just the one. But, these flaws are easier to fix in the next game instead of in the current game.

Also, I wish to experiment with Proxy support starting in Jan or Feb. Have done that before, but wish to see first hand what the lag time is like to see if I can come up with work arounds. To do proxy support, I'll upgrade my desktop (get a new one) remove the OS from the old one and replace it with Linex (maybe using RedHat). Then place my server app onto that PC (runs at 1G, should be fast enough) and walla.... my own server. A challange would be enuring all users know how to properly access the proxy server. This will be tested via the matchmaker....

See, all these little projects lead to my next project.....
Love having all you guys testing my stuff out too.

Great!!! Love the game

Keep it up