🚀 go-pugleaf

This is an OpenPGP/MIME signed message (RFC 2440 and 3156)
--------------enigA8CCE4C6E32BFB47CA464E8C
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Alex Holland wrote:
> 
> I've been playing around with FEN positions in 5.07, and discovered a
> segfault problem - not exactly a bug, but it could be friendlier.

All segfaults are bugs - unless you messed with something deep
underneath GNU Chess that it couldn't possibly be expected to cope with.

> If you load in a FEN position where white has black in check, and the turn
> flag indicates that it's white's turn, GNU Chess segfaults. On the one
> hand, this is quite reasonable, as it's an impossible position to get into,
> but on the other hand the error could be handled more gracefully. I've just
> spent ages trying to automate GNU chess for a series of FEN described PGN
> books, only to suffer crashing on every other file and having no idea why -
> the solution dawned on me this morning and left me feeling like an idiot.

Can you supply a test case please.

$ cat test.epd
r1bqkbnr/pp1ppQpp/2n5/2p5/4P3/8/PPPP1PPP/RNB1KBNR w KQkq - bm 1; id 1;

$ gnuchess
GNU Chess 5.07
Adjusting HashSize to 1024 slots
Transposition table:  Entries=1K Size=40K
Pawn hash table: Entries=0K Size=28K
White (1) : epdload test.epd

 : Best move = 1

white  KQkq
r . b q k b n r
p p . p p Q p p
. . n . . . . .
. . p . . . . .
. . . . P . . .
. . . . . . . .
P P P P . P P P
R N B . K B N R

Board is wrong!
White (1) : go
Thinking...
Looking for opening book in book.dat...
Read opening book (book.dat)...
Loading book from book.dat.
105 hash collisions...  Opening database: 625 book positions.
 In this position, there is no book move.


Time = 0.0 Rate=494 Nodes=[11/1/12] GenCnt=96
Eval=[3/6] RptCnt=0 NullCut=0 FutlCut=0
Ext: Chk=6 Recap=0 Pawn=0 OneRep=5 Horz=0 Mate=0 KThrt=0
Material=[3600/3600 : 4400/4300] Lazy=[150/150] MaxPosnScore=[150/150]
Hash: Success=85% Collision=50% Pawn=0%

black  KQkq
r . b q k b n r
p p . p p Q p p
. . n . . . . .
. . p . . . . .
. . B . P . . .
. . . . . . . .
P P P P . P P P
R N B . K . N R


My move is : Bc4#
1-0 {computer wins as white}
Black (1) : quit

Okay not exactly legal chess - but no segfault - I think it should just
reset the pieces on "Board is wrong!".

> BTW, is 5.07 officially released? Gentoo seemed somewhat suprised when I
> pointed out to them that it's been out since august, and the webpage at
> gnu.org still isn't updated.

Yes  - but the release got messed up by various issues at GNU, which
delayed the update of the book to version 1.01. And I haven't had time
or inclination to revisit it since.

Book 1.00 had unmatched quotes which meant a strict interpretation of
PGN vast chunks of it were tags and not games - despite appearances to
the contary.

--------------enigA8CCE4C6E32BFB47CA464E8C
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"

-----BEGIN PGP SIGNATURE-----
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFATghUGFXfHI9FVgYRAvfUAJ0Van2hUnXck8bl/IrZoAJZtKHkYgCcC4uX
OKUSS1fAYGASZh1JOWDUuCI=
=4Dcf
-----END PGP SIGNATURE-----

--------------enigA8CCE4C6E32BFB47CA464E8C--