🚀 go-pugleaf

Simon Waters wrote:
> Looks to me like perhaps it tried to play 1. ... c6 and got
> misunderstood

I don't see how xboard could misunderstand GNU Chess 5's attempt to play
c6 (or c7c6) as Nc6 (or b8c6).  What seems a bit more likely is that
GNU Chess 5 played c6 internally but sent something different to xboard.
That seems pretty surprising too, but it think it's a lot more likely
than xboard having a bug in parsing chess notation.  xboard has a lot of
miles on it, and there has been no other sign of such a bug.

If it's possible to reproduce this bug at all, it would help to have the
-debug flag of xboard set when doing so.  That shows the complete interaction
between xboard and the chess engine and will thus let you know for sure
which side the bug is on.

One thing to be aware of is that GNU Chess 5.00 doesn't support the "edit"
command of the xboard protocol, so if you try to start a game in a non-
standard starting position, GNU Chess will still think that the pieces are
in the standard starting position.  I don't think this happened here, though.

Dan Oetting wrote:
> In the rules of chess, you must challange the illegal move before your 
> play otherwise the move stands. An exception is if an inconcistancy is 
> discovered that prevents further play like a king is missing. It does 
> sound like there is a bug in xboard that doesn't handle the situation.

I think Dan was mostly joking in his message, but I thought I'd explain
how xboard handles illegal moves from engines.

xboard is aware when an engine plays a move that appears to be illegal, but
it allows those moves just in case the engine is playing a chess variant
that xboard doesn't understand.  xboard only displays an error message if
the notation from the engine is so bad that it can't even be parsed. 
However, now that I've added more explicit support for variants, maybe I
should change that and always display an error message if the engine makes
an illegal move and xboard's Test Legality option is on.

Henrik Gram wrote:
> I've seen a similar bug with winboard + olithink.  I didn't care
> enough to investigate further, but I have not had a single problem
> with xboard + gnuchess4.0.77 or xboard + my own engine when I've
> matched them together (couple of thousand games).

I've heard from others that olithink sometimes makes illegal moves.
I'm not certain whether GNU Chess 4 ever makes illegal moves.  I know
it sometimes gets confused and makes false mate or draw claims, so I 
wouldn't be too surprised to see an illegal move.  I guess I would know
more if xboard always gave an error message for this!

	--Tim

-- 
Tim Mann  tim.mann@compaq.com  http://www.tim-mann.org
Compaq Computer Corporation, Systems Research Center, Palo Alto, CA