Closed GoogleCodeExporter closed 9 years ago
Just to clarify: in order to castle you have to move the king to g1?
Original comment by bvk...@gmail.com
on 9 Oct 2011 at 7:13
This is the position:
r1bk1r2/ppp4p/3p2p1/n2q1p2/4nP2/2P1B1P1/PPP2N1P/R2K1RQB w KQkq - 4 11
Castling is not legal in the above position as white is in check. Otherwise,
A-side castling would be Kc1/Rd1 (the king and rook always end up the same as
in FIDE chess).
In FR, castling by moving the king should not be allowed as there is ambiguity
in some positions e.g. could either be a simple king move or a castle. This is
why CDay added the castle buttons (I think). Some interfaces also allow 'K
takes rook' as a means to indicate castling intentions.
Original comment by mdrmike...@gmail.com
on 9 Oct 2011 at 8:19
So the bug here is the following as I understand: Raptor sends long castle when
you make move to c1, but other interfaces castle after king move to b1. So the
fix would be to reassign castling trigger from king move to c1 to king move to
b1?
Original comment by bvk...@gmail.com
on 14 Oct 2011 at 6:18
[deleted comment]
The bug is raptor won't make the move, instead it "illegal move" beeps and
moves the king back to d1. It never sends the move to FICS, obviously it thinks
it's illegal.
You can duplicate the bug by following the steps in comment 1 above. Remember
to move the king with a mouse instead of just hitting the arrow button. For
instance, using the mouse, you can move the king to e1 just fine but you CAN'T
move the king to c1. However, both moves are legal and in examine mode you
should be able to make either move just fine.
Debugging hint:
Can you paste a FEN into Raptor? Try pasting two FEN's one with castling
perms set and won with castling perms unset. In each case try to make the move
Kcl.
Original comment by mdrmike...@gmail.com
on 15 Oct 2011 at 8:50
PS Possibly, I'm wrong in assuming "Raptor never sends the move to FICS". If it
does send a move, it's sending the WRONG one. It's either not sending a move
or sending a wrong move to FICS. Pehaps it's sending "o-o-o".
Original comment by mdrmike...@gmail.com
on 15 Oct 2011 at 9:26
Don't reassign any castling triggers. Simplest would be to only allow castling
via the castling buttons. I assumed Raptor was already coded this way.
The only other reasonable castling trigger would be "white king takes white A
rook" for A side castling and "white king takes B side rook for B side
castling. By rule the king is always between the rooks on a FR setup. In some
positions the king does not even move at all on castling, only the appropriate
rook! In other positions the king only slides one square presenting an
ambiguity problem. Don't change any code unless you thoroughly understand all
this.
Finally, on the comment(s) I made above. I wasn't sure whether or not Raptor
did it's own legality checking before sending a move to FICS. Never looked a
the code. I'm a c++ guy.
Original comment by mdrmike...@gmail.com
on 15 Oct 2011 at 10:43
[deleted comment]
Sorry to be so wordy on this but I think there's been some confusion. We're
talking about a simple king move here which is not the same as a castle.
Hence "Kc1" or king to any square is a completely different move than
"o-o-o". In Fischer Random (FR), just like FIDE chess, you can either move
the king or castle. In both variations, you still must have castling perms
and it must be legal in the given position. When castling, you can't move
the king through check, etc.
I could not make a *simple king move *in the alluded to game. Raptor, for
what ever reason, does not send "Kc1" to FICS, at least not from mouse input
on the graphical chess board. .
Original comment by mdrmike...@gmail.com
on 20 Oct 2011 at 1:54
I've overridden the board castling triggers, so the only way to castle in FR
now is to press castling button. You may check if it works properly using
newest package found here http://dl.dropbox.com/u/46373738/raptor-0.98u2.deb
Original comment by bvk...@gmail.com
on 21 Oct 2011 at 3:38
Original issue reported on code.google.com by
mdrmike...@gmail.com
on 9 Oct 2011 at 12:26