reza: It may not be useless... the position I posted when I started this whole thread 4 months ago has different outcomes if this is considered a valid move or not. Just to bring it back:
-------------
White: Ra5, ph6, Kc3
Black: Kb1, pa6, ph7
White to play, mate in 2
-------------
The proposed solution is
1.Ra5-a5 Kb1-c1
2.Ra5-a1++

And in halma, as rabbitoid pointed out, you can have a move that doesn't change the board at all.

I'm not stating that this should be a valid move or not. I just pointed out what I read in a book, that's all and was wondering for your oppinions. But then again, I remeber a famous game in chess history where someone promoted a pawn to an opponent's piece. And it was considered a valid move at the time since the rules didn't say anything against it (that's why they've been changed *hint**hint*)

There are games where an acceptable "move" is a pass, where the board doesn't change at all... and it is still called a move.
This case is not so radical: the piece does move. Philleas Fog did "move" around the world, even if it ended up where he started ;)

Pafl: I don't see why pgn-programs would have a difficult time with that (but then again I've never used one and am not quite sure even what is is for). I mean, if they're flexible enough to allow moves over the side of the board why not a move to the same position? :) (que ! was not part of the notation, mind you - just my way of showing surprise). Do these programs do any type of validation on the moves made?

In a book about games I found this cylinder chess problem:
-------------
White: Ra5, ph6, Kc3
Black: Kb1, pa6, ph7
White to play, mate in 2
-------------
The proposed solution is
1.Ra5-a5! Kb1-c1
2.Ra5-a1++

I haven't tested if this move is possible here at BK yet. My feeling is that it isn't... Truth be told the rules don't exclude the possibility that a Rook (or a Queen for that matter) loops around the board to finish the move in the same position it started in.
Nowhere on the web where I found rules for cylinder chess is this option discarded (nor noted) either.

What do you all think? And if it isn't possible in BK, should it be?

a draw can be agreed upon...
doesn't necessarily have to be a stalemate

"black King on A1
white Majarajah on B3

black to move."

This isn't stalemate... White just checkmated black by placing the maharajah at that position. So it's not black's turn: a win as already been accomplished.

when you mean they: the black or the white?
If the white can't move I think it's stalemate. So a draw.
If the black can't move... can this even happen?

... somehow, on the my diagram I had placed the queen at d1 and not d2 as BFG stated. I must agree with grenv then. Sorry for the confusion.

Since the black king can get close to the white king, and just chase him around from then on (preventing the queen from blowing him up because the white king would go in the process)... I think it's a draw.

in this game it wouldn't be a protecting move. But it may be a tactical move to expose a piece to the right spot...
I don't know wether it should be allowed or not (I don't know any 'official rules' to guide me by other than the ones on this site). But if it says the rules are 'the same as in regular chess with some exceptions' and castling is not in the list of exceptions then it should be allowed... or the rules corrected :)

I agree. But if castling is allowed then why couldn't I do it? There was no piece I could take, the king wasn't in check nor crossing check or anything...
I just got surprised. I think that either there is an error in the implementation of the game or the rules should be updated then...

The rules should say:
"-There is no check or checkmate and no castling

(you know... that movement where the rook and the king move at the same time... at least I think that's what it's called :} )

Either that or I'm failing to see the reason why I can't castle in this game...