xboard-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [XBoard-devel] Re: [Bug-XBoard] Winboard bugs


From: Tim Mann
Subject: Re: [XBoard-devel] Re: [Bug-XBoard] Winboard bugs
Date: Sat, 21 Feb 2004 19:37:52 -0800

Well, moves.c already has a move generator in it.  It's just that its
callers don't take the game history into account right now.  I think it
also doesn't take a flag to say whether we're in the wildcastle variant,
so you can castle from the d-file even when we aren't.  I'm pretty sure
it does look for castling in/through/out of check.  What's mostly needed
is to keep track of whether the kings and rooks have moved and which if
any pawn was pushed two squares on the previous move.  You'd have to add
in that information to make a different move generator handle these
cases just as much as to make the current one handle them.

On Sun, 22 Feb 2004 02:02:51 +0100, Daniel Mehrmann <address@hidden> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Tim Mann wrote:
> 
> That bring me on another old idea from me, discussed with Dieter
> (Yace-Author), that we are put a real move generator into
> XBoard/Winboard to check the moves. My old plan was to backport a
> "light version" of my own generator from my engine Homer to X/Winboard.
> It's a 0x88 and have less code. It would be not so difficult do that :)
> 
> 
> | It's not WinBoard's job to catch violations of the rules by the engines.
> | They are assumed to know the rules.  WinBoard just makes a preliminary
> | (and as you've noticed, incomplete) test that you haven't made an
> | illegal move before sending it to the engine.  It also happens to catch
> | most illegal moves from the engine too, but not all.  The test is
> | especially loose for castling and en passant.
> |
> | We could modify WinBoard to test exactingly for legal moves from
> | engines, and we may do that in the future.  However, there's nothing
> | that WinBoard could really do about that other than declare the game a
> | forfeit in your favor.  An engine can hardly be smart enough to deal
> | with some other program telling it "no, that move is illegal -- think
> | again."  :-)
> |
> [...]
> 
> - --
> Daniel Mehrmann
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.2.4 (MingW32)
> 
> iD8DBQFAN/+7Pt1V5Pj1nvYRAjwfAJsHmcECckZLFNV6+59RCb+nREfg0QCbBPWB
> 0Na0EyLznZQJHm5/JtxX9es=
> =x12p
> -----END PGP SIGNATURE-----
> 


-- 
Tim Mann  address@hidden  http://tim-mann.org/




reply via email to

[Prev in Thread] Current Thread [Next in Thread]