960: Now I CAN castle when the rules say I CAN'T

In this position, 0-0 is available to me, even though my king would be traversing d1, which is under attack by Blacks rook on d8.

I would actually like to make that move. If we're playing under local rules where that's allowed, I will. But more likely, this is a bug. I will wait for official word. If I'm informed by Richard that it's a bug he will fix, I'll make a different move even if he hasn't had time to fix it yet.

I'd rather like you not to be able to castle there. Just saying.

I'm going to assume we both thought I could not 0-0 in this position (I can and should be able to 0-0-0, but don't want to). So at least in this game, I'll forgo what I want to do. But the implementation here needs to either conform to rules published elsewhere of document the differences to avoid future disputes among various players.

Yes, it is a bug. I will fix it.

Thank you.

I meant "...or document the differences", not "of document..."

A strange castling bug in chess960 again. In this position I am allowed to O-O-O: http://www.littlegolem.net/jsp/game/game.jsp?gid=1285302&nmove=1

After the move I would have a rook at d8 and squares b8 and c8 would be empty - the other rook and the king disappear. :D

you should do that. Without a king, you cannot be mated. That is worth a rook, I think.

I have the same problem as Micco: http://www.littlegolem.net/jsp/game/game.jsp?gid=1279469&nmove=19 When I castle queenside, my king and rook disappear. Audience poll: Should I take Aganju's advice and castle anyway, or should I move my g-file knight and see what kingside castling does to me?

I suggest you only break teh game if your opponent goes along with the idea.

I was only kidding.

On the other hand, it might be interesting to see what happens if you are in a situation without a king. Depending on how the code is written, you might

- have lost immediately,

- or the server crashes,

- or really your opponent cannot win any more (maybe if he takes all your pieces it will be stalemate? - you can't move anymore...)

Yeah, I meant to put a smiley after my question. :) I forgot. I'll make a legitimate move. (probably with that knight -- It's strange how with normal chess, I'll stare at the position and analyze it, but with random chess, I just say "yay, this move looks fun" and play it. It must be different neurons firing.)

Castling is always a big pain to program. I programmed the rules of chess recently for a project of mine; I've done it a few times before, but this time I wanted to redo it "right". The first time I tested it, everything worked perfectly. En passant, castling, promotion, everything. That never happens; usually there's a bug somewhere. Did I miss something? So I tried it again. No, it all worked! Then my friend manages to castle out of check. And sure enough, in the code, I was checking to see whether the WRONG player was in check. Oh, my fragile ego.......

Has anyone of you (Micco, Tommah) sent message/mail to Richard about these situations too?

Good point, ypercube. :O I'll send him one now.

@Tommah: I think you should go ahead and castle into oblivion. This is a rare opportunity to perform an experiment. Of course, I'm saying this without knowing how important this game is to your chances in the tournament, nor how important the tournament is to you. If your king disappears, probably Richard can restore the game to move 19 after he fixes the bug.

It is OK now (I hope :))

It looks OK, I don't have the O-O-O option anymore in that position. Thanks, Richard!