WinBoard 4.3.15m bugfix available

[hgm]

Unfortunately in fixing one bug I broke something else. The WinBoard 4.3.15 that was included in the Gold Pack does contain a bug that only manifests itself in human-ICS play, and is totally fatal there.

For those who already downloaded the Gold Pack, a bug-fixed winboard.exe (4.3.15m) is available here.

For those loading WinBoard 4.3.15 for the first time, it is recommended you load the complete Gold Pack, including pre-installed support programs from the WinBoard forum.

Linux users will have to recompile from the sources.

Sorry about this mess-up.
And thanks to Slobo for pointing out the bug.

[slobo]

Unfortunately in fixing one bug I broke something else. The WinBoard 4.3.15 that was included in the Gold Pack does contain a bug that only manifests itself in human-ICS play, and is totally fatal there.

For those who already downloaded the Gold Pack, a bug-fixed winboard.exe (4.3.15m) is available here.

For those loading WinBoard 4.3.15 for the first time, it is recommended you load the complete Gold Pack, including pre-installed support programs from the WinBoard forum.

Linux users will have to recompile from the sources.

Sorry about this mess-up.
And thanks to Slobo for pointing out the bug.

Now, with bmp images for the board squares and a true-type font for chess pieces, WB is one of the nicest software to play on line.

Great!

What I need now is to make a Poliglot book visible, in order to navegate through it and chose position to played by chess engines.

But even more importante would be, in my opinion, to create a fairy-max versions with possibility to fix the engine´s search deep in all chess variants - in two, 3, 4, 5, 6 plies. With this new features, the popularity of new WB + FM would increase significantly.

Slobo

[hgm]

Fairy-Max does support the WinBoard sd N command to limit the search depth. Just include /sd=N on the WB command line.

As for the book, I cannot help you, as I know nothing about book building. I suppose tools exist to tune and maintain Polyglot books.

[Michel]

What I need now is to make a Poliglot book visible, in order to navegate through it and chose position to played by chess engines.

Have a look at this beta version of SCID.

http://sourceforge.net/project/showfile ... _id=651357

It can show the moves in the book and also allows you to tune the book.

I have also some (not yet released) utilities which can print the lines in a book. As an example see here

http://alpha.uhasselt.be/Research/Algeb ... _white.txt
http://alpha.uhasselt.be/Research/Algeb ... _black.txt

for the famous "performance.bin".

(by definition a "line" is a sequence of moves in which one of the players makes only book
moves and the other player makes arbitrary moves).

DISCLAIMER: since most PG books contain many isolated positions these lines do not tell the whole story.

EDIT: I forgot to say: creating books you do of course with the PG book building utilities.

[slobo]

Fairy-Max does support the WinBoard sd N command to limit the search depth. Just include /sd=N on the WB command line.

As for the book, I cannot help you, as I know nothing about book building. I suppose tools exist to tune and maintain Polyglot books.

I´ll try it later. Now I have another bug report.

In View and Edit mode, WB does not recognize FRC castling move. It claim an illegal move.

[Event "ChessCube Game"]
[Site "chat.chesscube.com"]
[Round "?"]
[White "stevandjuko@chesscube.com"]
[Black "daniel8@chesscube.com"]
[Date "2009.01.14"]
[Result "1-0"]
[Variant "chess 960"]
[SetUp "1"]

[FEN "rnkbbnqr/pppppppp/8/8/8/8/PPPPPPPP/RNKBBNQR w KQkq -"]


1. f4 d5 2. g4 h5 3. gxh5 Rxh5 4. Qg4+ f5 5. Qf3 e6 6. Ng3 Rh6 7. Bf2 Bc6 8. Bd4 Bf6 9. e3 g5 10. fxg5 Qxg5 11. Nc3 e5 12. Qxf5+ Bd7 13. Qxg5 Bxg5 14. Bxe5 Nc6 15. Bg7 Re6 16. Bxf8 O-O-O 17. Bc5 b6 18. Ba3 d4 19. exd4 Nxd4 20. Bg4 Rg6 21. Bxd7+ Rxd7 22. O-O-O Bf4 23. Nge4 Re6 24. h4 c5 25. Bxc5 Rxe4 26. Nxe4 bxc5 27. Nxc5 Ne2+ 28. Kb1 Rc7 29. Ne6 Rd7 30. Nxf4 Nxf4 31. h5 Rh7 32. h6 Ng6 33. Rdg1 Nf8 34. Rg8 Rh8 35. Rxh8 1-0

By the way, ChessCube is the best, if not the only site to play FRC on line.
WB stopted reproducing the game at move 16... O_O_O

Regards.

[slobo]

Thanks, Michel. I´ll try it.

[slobo]

Unfortunately in fixing one bug I broke something else. The WinBoard 4.3.15 that was included in the Gold Pack does contain a bug that only manifests itself in human-ICS play, and is totally fatal there.

For those who already downloaded the Gold Pack, a bug-fixed winboard.exe (4.3.15m) is available here.

For those loading WinBoard 4.3.15 for the first time, it is recommended you load the complete Gold Pack, including pre-installed support programs from the WinBoard forum.

Linux users will have to recompile from the sources.

Sorry about this mess-up.
And thanks to Slobo for pointing out the bug.

Thanks for your work, in the first place.
.
I have another observation. When I play Glaurunf with winboard option sd=3 (reduced search depth), frequently the game is stoped. I don´t know what is the cause of this problem, It seems that "xboard exit".

Look at the debug file:

437922 >first : new
random
437922 >first : variant fischerandom
437922 >first : force
write FEN 50-move: 0 0 0
e0. p=-4
437922 >first : setboard rkqnbrnb/pppppppp/8/8/8/8/PPPPPPPP/RKQNBRNB w FAfa - 0 1
437922 >first : ics -
437922 >first : level 0 10 3
437922 >first : sd 3
437922 >first : post
437922 >first : hard
437922 >first : easy
437922 >first : ping 9
437953 <first : pong 8
437953 <first : pong 9
release
5 0 1 5 0 1 Legality test? g2g3
FRC castling, 5 0 1 5 0 1
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=0 (6,1)-(6,2) -
5 0 1 5 0 1 Legality test? g2g3
FRC castling, 5 0 1 5 0 1
movetype=33, promochar=0=-
move: g2g3
, parse: g3 (
)
nps: w=-1, b=-1
441516 >first : name slobodan
nps: w=-1, b=-1
time odds: 1 1
441516 >first : time 60000
441516 >first : otim 60300
441516 >first : usermove 441516 >first : g2g3
441516 >first : go
(extra)
441547 <first : 2 -54 0 57 Nf6 d4
441547 <first : 2 -43 0 124 g6 Ne3
441547 <first : 3 -11 0 300 g6 Ne3 Ne6
441547 <first : 3 -3 0 460 Nf6 Ne3 d6
441547 <first : 3 -3 0 583 Nf6 Ne3 d6
441547 <first : move g8f6
machine move 1, castling = 5 0 1 5 0 1
move to parse: g8f6
5 0 1 5 0 1 Legality test? g8f6
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=23 (6,7)-(5,5) -
FRC castling, 5 0 1 5 0 1
move: g8f6
, parse: Nf6 (
)
release
5 0 1 5 0 1 Legality test? e2e4
FRC castling, 5 0 1 5 0 1
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=0 (4,1)-(4,3) -
5 0 1 5 0 1 Legality test? e2e4
FRC castling, 5 0 1 5 0 1
movetype=33, promochar=0=-
move: e2e4
, parse: e4 (
)
time odds: 1 1
449125 >first : time 60296
449125 >first : otim 59842
449125 >first : usermove 449125 >first : e2e4
449141 <first : 2 +7 0 50 Ne6 d3
449141 <first : 3 +21 0 285 Ne6 e5 Ng4
449141 <first : 3 +21 0 527 Ne6 e5 Ng4
449141 <first : move d8e6
machine move 3, castling = 5 0 1 5 0 1
move to parse: d8e6
5 0 1 5 0 1 Legality test? d8e6
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=23 (3,7)-(4,5) -
FRC castling, 5 0 1 5 0 1
move: d8e6
, parse: Ne6 (
)
release
5 0 1 5 0 1 Legality test? e4e5
FRC castling, 5 0 1 5 0 1
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=0 (4,3)-(4,4) -
5 0 1 5 0 1 Legality test? e4e5
FRC castling, 5 0 1 5 0 1
movetype=33, promochar=0=-
move: e4e5
, parse: e5 (
)
time odds: 1 1
450016 >first : time 60595
450031 >first : otim 60054
450031 >first : usermove 450031 >first : e4e5
450047 <first : 2 +23 0 161 Ng4 Nf3
450062 <first : 3 +27 0 388 Ng4 f4 Nxh2
450062 <first : 3 +27 0 495 Ng4 f4 Nxh2
450062 <first : move f6g4
machine move 5, castling = 5 0 1 5 0 1
move to parse: f6g4
5 0 1 5 0 1 Legality test? f6g4
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=23 (5,5)-(6,3) -
FRC castling, 5 0 1 5 0 1
move: f6g4
, parse: Ng4 (
)
release
5 0 1 5 0 1 Legality test? f2f4
FRC castling, 5 0 1 5 0 1
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=0 (5,1)-(5,3) -
5 0 1 5 0 1 Legality test? f2f4
FRC castling, 5 0 1 5 0 1
movetype=33, promochar=0=-
move: f2f4
, parse: f4 (
)
time odds: 1 1
458969 >first : time 60890
458969 >first : otim 59464
458969 >first : usermove 458969 >first : f2f4
459000 <first : 2 +54 0 131 Nxh2 Rf2
459000 <first : 3 +78 0 413 Nxh2 Rf2 Ng4
459000 <first : 3 +78 0 620 Nxh2 Rf2 Ng4
459000 <first : move g4h2
machine move 7, castling = 5 0 1 5 0 1
move to parse: g4h2
5 0 1 5 0 1 Legality test? g4h2
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=23 (6,3)-(7,1) -
FRC castling, 5 0 1 5 0 1
move: g4h2
, parse: Nxh2 (
)
release
5 0 1 5 0 1 Legality test? f1f2
FRC castling, 5 0 1 5 0 1
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=3 (5,0)-(5,1) -
FRC castling, 5 0 1 5 0 1
move: f1f2
, parse: Rf2 (
)
time odds: 1 1
462344 >first : time 61187
462344 >first : otim 59429
462344 >first : usermove 462344 >first : f1f2
462375 <first : 2 +92 0 160 Ng4 Re2
462375 <first : 3 +127 0 326 Ng4 Re2 f5
462375 <first : 3 +127 0 392 Ng4 Re2 f5
462375 <first : move h2g4
machine move 9, castling = -1 0 1 5 0 1
move to parse: h2g4
5 0 1 5 0 1 Legality test? h2g4
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=23 (7,1)-(6,3) -
FRC castling, -1 0 1 5 0 1
move: h2g4
, parse: Ng4 (
)
release
-1 0 1 5 0 1 Legality test? f2e2
FRC castling, -1 0 1 5 0 1
TC string = '+10+3'
mps=0 tc=600000 inc=3000
CoordsToAlgebraic, piece=3 (5,1)-(4,1) -
FRC castling, 5 0 1 5 0 1
move: f2e2
, parse: Re2 (
)
time odds: 1 1
464594 >first : time 61484
464594 >first : otim 59507
464594 >first : usermove 464594 >first : f2e2
GameEnds(40, xboard exit, 2)
540328 >first : result * {xboard exit}
540328 >first : force
540328 >first : ping 10
540328 >first : quit

[hgm]

"XBoard exit" is what WinBoard normally says as game result, when you close it during a game in progress. It does not seem that WinBoard stops automatically here: the time stamps show that the game is terminated (by sending result to the engine) after nothing happened for 76 sec after Glaurung was sent the move f2e2. Presumably because you concluded that no move will be forthcoming, and closed the WinBoard main window.

It seems Glaurung simply goes into paralysis after receiving the move f2e2. I cannot detect anything strange in the GUI <-> engine communication upto that point. So it must be a Glaurung problem. Perhaps it cannot handle such small depths. Send this debug file to Tord, and he can probably tell us what caused Glaurung to lapse into a coma.