Tinkering with Winboard

Discussions about Winboard/Xboard. News about engines or programs to use with these GUIs (e.g. tournament managers or adapters) belong in this sub forum.

Moderator: Andres Valverde

Re: PolyGlot and book

Postby Marc Lacrosse » 28 May 2005, 08:57

Fabien Letouzey wrote:
Marc Lacrosse wrote:At the opposite, most UCI engines rely on the GUI for book operations (I know that Yace is a brilliant exception !) and what I miss mostly is some kind of book management by polyglot.

I read somewhere that you intended to add book-features to Polyglot.
Did you progress in this way ?

PS bookthinker is not a satisfying answer for book management; not only its use is a pain, but it cruelly lacks user-book features like the ones of crafty, yace or rebel (to name a few)


There is an experimental version that supports only NAIVE book making, I am afraid not as good as BookThinker. I had in mind to improve it before a release but then switched on to working on something else (that was in February).

It is now clear: I will release that version "as is" (but stable) next week, and try to improve things only afterwards.

Basically I don't want to spend time researching opening-book heuristics.
However, if one (yes that's YOU) tells me EXACTLY what he wants, I can do it hopefully in little time. Please open a thread in that case, thanks.

As an example, Leo mentionned a heuristics regarding minimum win % (separate for white and black, e.g. 50 and 45) for moves with many games (say at least 100 or 1000).

I have nothing against user books, just make sure I can figure out EXACT move probability from your data (you tell me how).

However I fail to see in which way PolyGlot would be easier to use than BookThinker.

Fabien.


Dear Fabien,
I think a few features could make opening book management within polyglot very interesting.
I will try to give you more precise specifications.
I have extensively tested a few programs book management features but most of them are not completely satysfying in my humble opinion.
May be I should send a few comments to you in french (ouside the forum) on this topic because I am not to fluent in english for precise description ... :(

regards

Marc
Marc Lacrosse
 
Posts: 116
Joined: 29 Jan 2005, 09:04
Location: Belgium

Re: PolyGlot and book

Postby Fabien Letouzey » 28 May 2005, 09:09

Marc Lacrosse wrote:I think a few features could make opening book management within polyglot very interesting.
I will try to give you more precise specifications.
I have extensively tested a few programs book management features but most of them are not completely satysfying in my humble opinion.
May be I should send a few comments to you in french (ouside the forum) on this topic because I am not to fluent in english for precise description ... :(


Hi Marc,

Sure, feel free to send an email.
It should not be difficult to find my address.

Fabien.
Fabien Letouzey
 
Posts: 110
Joined: 03 Dec 2004, 10:17
Location: France

Re: Tinkering with Winboard

Postby Daniel Mehrmann » 12 Jun 2005, 11:13

Well, i'm just beginning now to backport some addons by Alessandro Scotti to the X-/Winboard mainline. You can see the latest changes on http://savannah.gnu.org/projects/xboard/

I allso added a service for users, who can't compile the latest CVS sources or just don't want that, on my own homepage: http://www.homer-chess.com

Here can you download the latest offical CVS snapshot of Winboard.
I hope this will be usefull.
greetings
Daniel
User avatar
Daniel Mehrmann
 
Posts: 127
Joined: 02 Oct 2004, 06:10
Location: Germany

Little bug found

Postby Guenther Simon » 09 Aug 2005, 14:51

Hi Alessandro,

I have found a little bug in your WinboardX. The first one after
quite long ;) It seems it cannot attach the score/depth
for the first ply currently ?
I noticed this during my _bookless_ long time control matches.

Best wishes,
Guenther

Oh and as I am already writing, would it be difficult to let
the user change the standard PGN event header? :)

Example:
Processed with Georges tool
[Event "RWBC LongTimeTest1"]
[Site "ESPRESSO P4 2.67Ghz 128MB hash PO"]
[Date "2005.08.02"]
[Round "1"]
[White "Fruit_21"]
[Black "Shredder_91T"]
[Result "0-1"]
[TimeControl "14400+60"]

1. e4{+0.21/16 514s 309170000n} e5{-0.42/20 905s 263145283n} 2. Nf3
{+0.36/16 531s 315860444n} d5{-0.32/19 1393s 422286452n} 3. exd5

WBx
[Event "Computer chess game"]
[Site "ESPRESSO"]
[Date "2005.08.02"]
[Round "1"]
[White "Fruit_21"]
[Black "Shredder_91T"]
[Result "0-1"]
[TimeControl "14400+60"]

1. e4 e5 {-0.42/20} 2. Nf3 {+0.36/16} d5 {-0.32/19} 3. exd5 {+0.71/15}
User avatar
Guenther Simon
 
Posts: 794
Joined: 26 Sep 2004, 19:49
Location: Regensburg, Germany

Re: Tinkering with Winboard

Postby Somnus » 20 Aug 2005, 05:58

Daniel Mehrmann wrote:Well, i'm just beginning now to backport some addons by Alessandro Scotti to the X-/Winboard mainline. You can see the latest changes on http://savannah.gnu.org/projects/xboard/

I allso added a service for users, who can't compile the latest CVS sources or just don't want that, on my own homepage: http://www.homer-chess.com

Here can you download the latest offical CVS snapshot of Winboard.
I hope this will be usefull.


What did you use to compile the latest CVS build on your website?
Somnus
 
Posts: 54
Joined: 26 Dec 2004, 03:12
Location: Edmonton, Alberta, Canada

Re: Tinkering with Winboard

Postby Daniel Mehrmann » 20 Aug 2005, 23:06

Somnus wrote:
Daniel Mehrmann wrote:Well, i'm just beginning now to backport some addons by Alessandro Scotti to the X-/Winboard mainline. You can see the latest changes on http://savannah.gnu.org/projects/xboard/

I allso added a service for users, who can't compile the latest CVS sources or just don't want that, on my own homepage: http://www.homer-chess.com

Here can you download the latest offical CVS snapshot of Winboard.
I hope this will be usefull.


What did you use to compile the latest CVS build on your website?


VC 6.0 professional with SP6
greetings
Daniel
User avatar
Daniel Mehrmann
 
Posts: 127
Joined: 02 Oct 2004, 06:10
Location: Germany

FEN bug in latest WBx release?

Postby Guenther Simon » 01 Nov 2005, 14:58

Hi Alessandro,

It seems there is some quirk here with FENs in WB?
You can see that the FEN is changed to a different FEN by WB
(adding castling rights suddenly)
and e.g. Fruit 2.21 crashes with that FEN.

My wild guess is that it has to do with changes for FRC handling?

Best regards,
Guenther

[diag]2rq1rk1/1b1nbpp1/pn1pp2p/1p5P/3NP1P1/P1N1BP2/1PP2Q2/1K1R1B1R w - - 0 1[/diag]
WBx debug:
Code: Select all
PasteFenFromString(): fen '2rq1rk1/1b1nbpp1/pn1pp2p/1p5P/3NP1P1/P1N1BP2/1PP2Q2/1K1R1B1R w - - 0 1'
11055 >first : force
11055 >first : new
random
11055 >first : ics -
11055 >first : level 40 120 0
11055 >first : post
11055 >first : hard
11055 >first : easy
11055 >first : ping 2
11055 >first : force
11055 >first : setboard 2rq1rk1/1b1nbpp1/pn1pp2p/1p5P/3NP1P1/P1N1BP2/1PP2Q2/1K1R1B1R w Kq - 0 1
11126 <first : pong 2
13419 >first : name Guenther Simon
13419 >first : time 720000
otim 720000
13419 >first : go
Fatal Error: Error: first chess program (Fruit_221) exited unexpectedly
GameEnds(0, (null), 2)
13629 >first : force
13629 >first : ping 3
13739 >first : quit


Polyglot log excerpt:
Code: Select all
< XBOARD level 40 120 0
< XBOARD post
< XBOARD hard
POLYGLOT WAIT
< XBOARD easy
POLYGLOT WAIT
< XBOARD ping 1
> XBOARD pong 1
< XBOARD force
POLYGLOT WAIT
< XBOARD new
POLYGLOT NEW GAME
POLYGLOT WAIT
> ENGINE ucinewgame
< XBOARD random
< XBOARD ics -
< XBOARD level 40 120 0
< XBOARD post
< XBOARD hard
POLYGLOT WAIT
< XBOARD easy
POLYGLOT WAIT
< XBOARD ping 2
> XBOARD pong 2
< XBOARD force
POLYGLOT WAIT
< XBOARD setboard 2rq1rk1/1b1nbpp1/pn1pp2p/1p5P/3NP1P1/P1N1BP2/1PP2Q2/1K1R1B1R w Kq - 0 1
POLYGLOT FEN 2rq1rk1/1b1nbpp1/pn1pp2p/1p5P/3NP1P1/P1N1BP2/1PP2Q2/1K1R1B1R w Kq - 0 1
POLYGLOT WAIT
< XBOARD name Guenther Simon
< XBOARD time 720000
< XBOARD otim 720000
< XBOARD go
POLYGLOT THINK
POLYGLOT START SEARCH
POLYGLOT FEN 2rq1rk1/1b1nbpp1/pn1pp2p/1p5P/3NP1P1/P1N1BP2/1PP2Q2/1K1R1B1R w Kq - 0 1
> ENGINE position fen 2rq1rk1/1b1nbpp1/pn1pp2p/1p5P/3NP1P1/P1N1BP2/1PP2Q2/1K1R1B1R w Kq - 0 1
> ENGINE go wtime 7200000 btime 7200000 movestogo 40
< ENGINE board_from_fen(): bad FEN (pos=64)
< ENGINE EOF
User avatar
Guenther Simon
 
Posts: 794
Joined: 26 Sep 2004, 19:49
Location: Regensburg, Germany

Re: Tinkering with Winboard

Postby Alessandro Scotti » 04 Nov 2005, 19:21

Hi Guenther,
the problem is indeed related to the FRC code.
By default Winboard doesn't keep track of castling rights so when the current position is converted to FEN it tries to reconstruct them (doing some tests on king and rook positions).
This code was updated in the FRC version so that castling rights are generated according to FRC rules and with the UCI syntax, and those cannot be parsed by standard engines.
I overlooked this part of code and so I inserted it into the "standard" version. If we want to support both standard and FRC in the same executable then the proper encoding method must be selected in some way... let me think about it and suggestions are always well accepted of course!
User avatar
Alessandro Scotti
 
Posts: 306
Joined: 20 Nov 2004, 00:10
Location: Rome, Italy

Re: Tinkering with Winboard

Postby Fabien Letouzey » 06 Nov 2005, 09:55

Hi Alessandro,

Alessandro Scotti wrote:I overlooked this part of code and so I inserted it into the "standard" version. If we want to support both standard and FRC in the same executable then the proper encoding method must be selected in some way... let me think about it and suggestions are always well accepted of course!

I don't remember for sure whether the version from Tord and I works correctly in this case, but we use the "fischerandom" variant to distinguish FRC from normal chess.

Fabien.
Fabien Letouzey
 
Posts: 110
Joined: 03 Dec 2004, 10:17
Location: France

Re: Tinkering with Winboard

Postby Alessandro Scotti » 30 Nov 2005, 00:17

Hi,
there is a new Winboard version at this page:

http://www.ascotti.org/programming/chess/winboard_x.htm

There are two main features in this release: an enhanced game list window and full support for setting up a FRC start position.

It is now possible to filter games in the game list, and also to get some stats about the games. If you use Winboard as a fast PGN browser, I think you'll like this.

Also, you can now start a FRC game from the last position, one of the different 960 positions (by number) or get a random position each time you start a new game.

As usual there are some other minor features and bug fixes. I have had to move a few lines of code around, most noticeably those related to game adjudication. I have performed many test games without problems but it's not a bad idea to play some test games yourself if you plan to use this new version.

What's coming next? Definitely, a move history list and the evaluation graph. In fact, they should have been there already but I was not completely happy with the design (too "ambitious") so I've scrapped the existing code. Hopefully, it won't take too much to get something working again.
User avatar
Alessandro Scotti
 
Posts: 306
Joined: 20 Nov 2004, 00:10
Location: Rome, Italy

Re: Tinkering with Winboard

Postby Guenther Simon » 30 Nov 2005, 16:18

Thanks Alessandro!

Strange that I found something exactly a few hours ago ;-)
I guess though this was there since long and may be even
since always in WB.
It is likely also a prob with the output of the engine,
(not quite sure as it works partly)

I noticed that SSEChess' output isn't read correctly for
both the thinking line output in WB itself _and_ in the
PGN files with depth/score.
It reports PVs and depths for both which are not the _real_
depths/scores reached according to the debug file!

Somehow the reported output is always at a less depth
as SSEChess really had during thinking.
That was the reason I noticed it at all, because I wondered
about the low reported depths.

Below I have posted two different PGNs for the same game.
The first one is the original PGN saved by WB and the second
one is extracted from the debug file by LGDEBPGN.
You can see the differences for SSEChess.

Ah! After posting also two thinking examples from the debug
it seems I know the reason already :)
If I am right the problem is that SSEChess sometimes reports
the time not as integers but as floating points with one digit!

I don't know right now what the protocol says and leave it to you,
if this is something which should be changed/fixed in WB :)
At least I know it won't be fixed anymore in SSEChess.

Best wishes,
Guenther

[Event "RWBC_Class_D_6th_Edition_Test"]
[Site "ESPRESSO"]
[Date "2005.11.29"]
[Round "1"]
[Number "24"]
[White "SSEChess_2045"]
[Black "Natwarlal_012"]
[Result "0-1"]
[TimeControl "40/900"]
1. e4 e5 2. Nf3 d6 3. d4 exd4 4. Nxd4 Nf6 5. Nc3 Be7 6. Be2 O-O {-0.46/12}
7. O-O d5 {-0.43/12} 8. exd5 {+0.58/7} Nxd5 {-0.48/12} 9. Nxd5 {+0.47/11}
Qxd5 {-0.33/11} 10. Bf3 {+0.53/6} Qc5 {-0.50/10} 11. Re1 {+0.50/6}
Rd8 {-0.63/10} 12. c3 {+0.56/6} Bd6 {-0.86/10} 13. h3 {+0.43/9}
Nd7 {-0.27/10} 14. Re2 {+0.39/9} Nf6 {-0.01/10} 15. Qb3 {+0.42/6}
c6 {+0.08/10} 16. Kf1 {+0.32/6} a5 {+0.29/10} 17. a4 {+0.49/6}
Bd7 {+0.21/9} 18. Bd2 {+1.02/4} Rab8 {+0.23/9} 19. Rd1 {+0.07/6}
Rdc8 {+0.03/9} 20. Qc2 {+0.21/8} Re8 {+0.08/10} 21. Be3 {+0.32/6}
Qe5 {+0.45/10} 22. Red2 {+0.33/5} Bc7 {+0.46/10} 23. Ne2 {+0.29/6}
Bf5 {+0.89/11} 24. Qc1 {-0.27/9} Qh2 {+0.89/10} 25. b3 {+0.13/5}
Ne4 {+1.19/11} 26. Rb2 {-0.34/5} Rbd8 {+1.46/11} 27. Rxd8 {-0.33/6}
Rxd8 {+1.81/11} 28. Ng1 {-0.41/6} Be5 {+1.63/10} 29. Bb6 {-0.26/6}
Rd5 {+1.97/10} 30. Bxe4 {-0.46/6} Bxe4 {+3.07/11} 31. f3 {-0.93/7}
Bf4 {+5.02/12} 32. Qe1 {-0.56/7} Bd3+ {+5.71/12} 33. Re2 {-0.60/6}
c5 {+6.54/13} 34. Bxc5 {-2.05/6} Rxc5 {+13.90/13} 35. Qh4 {-4.88/7}
g5 {+19.62/13} 36. Qf2 {-5.96/7} Rxc3 {+327.60/9} 37. Qe1 {-10.40/8}
Be3 {+327.62/6} 38. Qf2 {-99999.94/6} Rc1+ {+327.64/4} 39.
Qe1 {-99999.96/1} Qxg1# {+327.66/4}
{Black mates} 0-1

[Event "RWBC_Class_D_6th_Edition_Test"]
[Site "ESPRESSO"]
[Date "2005.11.29"]
[Round "1"]
[Number "24"]
[White "SSEChess_2045"]
[Black "Natwarlal_012"]
[Result "0-1"]
[TimeControl "40/900"]
1. e4 e5 2. Nf3 d6 3. d4 exd4 4. Nxd4 Nf6 5. Nc3 Be7 6. Be2 O-O {-0.46/12} 7. O-O
d5 {-0.43/12} 8. exd5 {+0.53/11} Nxd5 {-0.48/12} 9. Nxd5 {+0.47/12} Qxd5 {-0.33/11}
10. Bf3 {+0.47/10} Qc5 {-0.50/10} 11. Re1 {+0.49/10} Rd8 {-0.63/10} 12. c3
{+0.47/10} Bd6 {-0.86/10} 13. h3 {+0.43/9} Nd7 {-0.27/10} 14. Re2 {+0.31/10} Nf6
{-0.01/10} 15. Qb3 {+0.30/10} c6 {+0.08/10} 16. Kf1 {+0.23/10} a5 {+0.29/10} 17. a4
{+0.12/10} Bd7 {+0.21/9} 18. Bd2 {+0.14/9} Rab8 {+0.23/9} 19. Rd1 {+0.12/9} Rdc8
{+0.03/9} 20. Qc2 {+0.26/9} Re8 {+0.08/10} 21. Be3 {+0.14/9} Qe5 {+0.45/10} 22. Red2
{+0.14/9} Bc7 {+0.46/10} 23. Ne2 {+0.00/9} Bf5 {+0.89/11} 24. Qc1 {-0.24/10} Qh2
{+0.89/10} 25. b3 {-0.23/9} Ne4 {+1.19/11} 26. Rb2 {-0.45/9} Rbd8 {+1.46/11} 27.
Rxd8 {-0.37/9} Rxd8 {+1.81/11} 28. Ng1 {-0.64/9} Be5 {+1.63/10} 29. Bb6 {-0.35/9} Rd5
{+1.97/10} 30. Bxe4 {-1.17/10} Bxe4 {+3.07/11} 31. f3 {-2.12/11} Bf4 {+5.02/12} 32.
Qe1 {-2.75/10} Bd3+ {+5.71/12} 33. Re2 {-2.28/9} c5 {+6.54/13} 34. Bxc5 {-5.84/10}
Rxc5 {+13.90/13} 35. Qh4 {-8.60/11} g5 {+19.62/13} 36. Qf2 {-10.07/11} Rxc3
{+327.60/9} 37. Qe1 {-99999.92/10} Be3 {+327.62/6} 38. Qf2 {-99999.94/6} Rc1+
{+327.64/4} 39. Qe1 {-99999.96/1} Qxg1# {+327.66/4}
{Black mates} 0-1

(debug snippet for 17.a4)
Code: Select all
605281 >first : a7a5
605301 <first : 1 59 0 64 Be3
605301 <first : 2 57 0 290 Be3 h6
605301 <first : 3 59 1 2262 Be3 a4 Qd1
605301 <first : 4 29 1 4898 Be3 a4 Qd1 Bd7
605331 <first : 5 56 4 19107 Be3 a4 Qc2 Nd5 Rae1 Nxe3+ Rxe3
605441 <first : 6 49 15 72201 Be3 a4 Qd1 Nd5 Bxd5 Qxd5 Kg1
606202 <first : 7 32 90.1 454567 Be3 a4 Qc2 Nd5 Bxd5 Qxd5 c4 Qe5
607874 <first : 8 36 257.3 1329623 Be3 a4 Qc2 Re8 Rd1 Bd7 Nb5 Rxe3 Nxd6
616046 <first : 9 32 1075.5 5529417 a4 Be5 Be3 Qd6 Rd2 Qe7 Rad1 Nd5 Re1 Nxe3+ Rxe3
628364 <first : 10 12 2307.3 11605825 a4 Bd7 Be3 Qe5 Rae1 Qh2 Rd2 c5 Nb5 Be6 Qd1
628364 <first : move a2a4


(debug snippet for 24.Qc1)
Code: Select all
907265 >first : d7f5
907275 <first : 1 8 0 90 Qc1
907275 <first : 2 -1 1 370 Qb3 Ne4
907275 <first : 3 0 1 1662 Qc1 Ne4 Rd3
907285 <first : 4 9 2 5700 Qc1 Be4 Bxe4 Nxe4 Rd7
907305 <first : 5 -8 4 14400 Qc1 Ne4 Bxe4 Qxe4 b3 b6
907375 <first : 6 4 11 45618 Qc1 Ne4 Bxe4 Qxe4 b3 b6 Nd4
907545 <first : 7 0 28 141157 Qc1 Ne4 Bxe4 Qxe4 b3 b6 Nd4 Rbd8 Nxf5 Qxf5 Rxd8 Bxd8
908226 <first : 8 0 96.1 502849 Qc1 Ne4 Bxe4 Qxe4 b3 b6 Nd4 Rbd8 Nxf5 Qxf5 Rxd8 Bxd8
914045 <first : 9 -27 677 3356366 Qc1 Qh2 Ng1 Ne4 Rd3 Rbd8 Rxd8 Bxd8 b3 b6
921616 <first : 10 -24 1435.1 7051874 Qc1 Qh2 Ng1 Ne4 Rd4 b5 axb5 cxb5 b3 Rbd8 Ne2
927364 <first : move c2c1
User avatar
Guenther Simon
 
Posts: 794
Joined: 26 Sep 2004, 19:49
Location: Regensburg, Germany

Re: Tinkering with Winboard

Postby Dann Corbit » 30 Nov 2005, 17:55

Guenther Simon wrote:Thanks Alessandro!
{snip}
I don't know right now what the protocol says and leave it to you,
if this is something which should be changed/fixed in WB :)
At least I know it won't be fixed anymore in SSEChess.
{snip}


I have the source for SSEChess. I could make the needed changes if you like.
Dann Corbit
 

Re: Tinkering with Winboard

Postby Guenther Simon » 30 Nov 2005, 18:10

<Dann Corbit>
I have the source for SSEChess. I could make the needed changes if you like.


Thanks Dann, that's a nice offer!
I would like to know first though, if it is really a missinterpretation
of the protocol by SSEChess, because if it is not, WB must
be fixed of course.
Matacz BTW also has a problem with the output formatting,
thus WB doesn't show _anything_ for it in the thinking lines
and in the extended PGN files.
It uses 2 digits floating points for all scores...WB skips all
lines which contain non integers it seems.


<Matacz debug snippet>

Code: Select all
178507 >second: d1e2
178557 <second:     2     -0.17     0     121  c5d4 c3d4
178557 <second:     3     0.23     0     307  c5d4 c3d4 a8c8
178557 <second:     4     -0.01     0     787  c5d4 c3d4 a8c8 c1f4
178557 <second:     5     -0.01     0     10569  c5d4 c3d4 a8c8 c1f4 g7g5
178647 <second:     6     -0.01     0     22871  c5d4 c3d4 a8c8 c1f4 g7g5 f3e5
178978 <second:     7     0.00     0     129202  c5d4 c3d4 a8c8 c1f4 g7g5 f3e5 g5f4
179639 <second:     8     -0.01     0     252706  c5d4 c3d4 a8c8 c1f4 g7g5 f3e5 g5f4 a1b1
181151 <second:     9     0.01     0     778375  c5d4 c3d4 a8c8 c1d2 f8e8 e4e5 b7d5 d3e4 a7a5
191936 <second:     10     -0.08     0     3425144  a8c8 c1d2 a7a5 a3a4 g7g5 h2h3 c5c4 d3c4 b7e4 e2e4
204925 <second: move a8c8


Best regards,
Guenther
User avatar
Guenther Simon
 
Posts: 794
Joined: 26 Sep 2004, 19:49
Location: Regensburg, Germany

Re: Tinkering with Winboard

Postby Guenther Simon » 30 Nov 2005, 18:20

Ok I looked into the WB protocol and it seems clear to me
that WB expects integers for all values, thus I would really
welcome your code change for SSEChess Dann :)

For the case of Matacz, I will simply mail the author about
the necessary change.

Best regards,
Guenther


Excerpt from Tim Manns WB protocol:
Code: Select all
10. Thinking Output

If the user asks your engine to "show thinking", xboard sends your engine the "post" command. It sends "nopost" to turn thinking off. In post mode, your engine sends output lines to show the progress of its thinking. The engine can send as many or few of these lines as it wants to, whenever it wants to. Typically they would be sent when the PV (principal variation) changes or the depth changes. The thinking output should be in the following format:

ply score time nodes pv

Where:
ply   Integer giving current search depth.
score   Integer giving current evaluation in centipawns.
time   Current search time in centiseconds (ex: 1028 = 10.28 seconds).
nodes   Nodes searched.
pv   Freeform text giving current "best" line. You can continue the pv onto another line if you start each continuation line with at least four space characters.

Example:

  9 156 1084 48000 Nf3 Nc6 Nc3 Nf6
User avatar
Guenther Simon
 
Posts: 794
Joined: 26 Sep 2004, 19:49
Location: Regensburg, Germany

Re: Tinkering with Winboard

Postby Thomas Mayer » 30 Nov 2005, 18:21

Hi G?nther,

the reason for that problem is simple:

SSE_Chess output:

Code: Select all
10 12 2307.3 11605825 a4 Bd7 Be3 Qe5 Rae1 Qh2 Rd2 c5 Nb5 Be6 Qd1


the third number has a "." inbetween. It seems that Winboard checks each number for having just numeric chars in it, therefor it does not accept the line as a thinking line. Afaik this is not exactly defined in the standard maybe one could adjust WinBoard in a way here to keep that more relaxed. (On the other hand this could affect compatibilty to other engines)

Same reason for Matacz:

Code: Select all
179639 <second:     8     -0.01     0     252706  c5d4 c3d4 a8c8 c1f4 g7g5 f3e5 g5f4 a1b1


here it is the score parameter which is given with a "dot" -> in fact according to the winboard standard the score should be given in centipawns. So here definitely Matacz makes something wrong. Maybe the author can change that ?

Greets, Thomas

P.S.: Edited because format was wrong
User avatar
Thomas Mayer
 
Posts: 40
Joined: 26 Oct 2004, 13:45
Location: Germany

Re: Tinkering with Winboard

Postby Thomas Mayer » 30 Nov 2005, 18:23

too slow... :)
User avatar
Thomas Mayer
 
Posts: 40
Joined: 26 Oct 2004, 13:45
Location: Germany

Re: Tinkering with Winboard

Postby Guenther Simon » 30 Nov 2005, 18:24

<Thomas>
Hi G?nther,

the reason for that problem is simple:

SSE_Chess output:

Code:
10 12 2307.3 11605825 a4 Bd7 Be3 Qe5 Rae1 Qh2 Rd2 c5 Nb5 Be6 Qd1


the third number has a "." inbetween. It seems that Winboard checks each number for having just numeric chars in it, therefor it does not accept the line as a thinking line. Afaik this is not exactly defined in the standard maybe one could adjust WinBoard in a way here to keep that more relaxed. (On the other hand this could affect compatibilty to other engines)


Hi Thomas,

I thought this is exactly what I had written in my first post ;-)
(integers vs. floating point numbers)

Best,
Guenther
User avatar
Guenther Simon
 
Posts: 794
Joined: 26 Sep 2004, 19:49
Location: Regensburg, Germany


Re: Tinkering with Winboard

Postby Guenther Simon » 01 Dec 2005, 12:40

<quote Dann>
subject: SSEChess with integral time in seconds and centipawn eval


Thanks Dann, the fixed version already won its first game in RWBC :)
I propose you announce it perhaps in a special message or at
Leos too, because I don't know how much read here with us.

I noticed BTW it calls itself 2.050, does this mean it is based on
a slightly newer source than the last official release 2.045, or did
you name it that way for distinction?

Best regards,
Guenther

[Event "RWBC_Class_D_6th_Edition"]
[Site "ESPRESSO"]
[Date "2005.12.01"]
[Round "2"]
[Number "40"]
[White "Monarch2002_04c"]
[Black "SSEChess_2050DC"]
[Result "0-1"]
[TimeControl "40/900"]

1. b3 e5 2. Bb2 Nc6 3. e3 Nf6 4. Bb5 {+0.08/11} d6 5. Nf3 Bd7 {+0.38/11} 6.
O-O {-0.02/11} a6 {+0.35/11} 7. Be2 {-0.08/12} e4 {+0.27/11} 8.
Ne1 {-0.13/11} Be7 {+0.22/11} 9. c4 {+0.03/12} O-O {+0.12/11} 10.
Nc3 {+0.06/12} Bf5 {+0.05/11} 11. g4 {-0.23/13} Be6 {+0.47/10} 12.
g5 {+0.20/12} Nd7 {+0.50/11} 13. Nxe4 {+0.24/11} Bf5 {+0.40/10} 14.
Nc3 {+0.34/11} Bh3 {+0.52/11} 15. Ng2 {-0.15/13} Bxg5 {+0.60/11} 16.
d4 {+0.35/11} Nf6 {+0.56/10} 17. f4 {+0.47/12} Bh6 {+0.62/10} 18.
e4 {+0.47/11} Qc8 {+0.63/9} 19. f5 {+0.46/10} Qe8 {+0.78/10} 20.
Qd3 {+0.54/11} Bxg2 {+1.05/12} 21. Kxg2 {+0.51/11} Nb4 {+1.05/10} 22.
Qb1 {+0.49/11} Be3 {+0.88/10} 23. d5 {+0.52/11} Bd4 {+1.29/10} 24.
a3 {+0.50/11} Qe5 {+0.64/9} 25. axb4 {+0.48/11} Bxc3 {+1.92/11} 26.
Bxc3 {-21474205.11/12} Qxc3 {+1.82/11} 27. Ra4 {-0.24/11} Rfe8 {+1.69/10}
28. Bf3 {+0.32/10} b5 {+2.13/10} 29. cxb5 {-0.44/9} axb5 {+1.76/8} 30.
Rxa8 {-0.57/11} Rxa8 {+2.15/9} 31. Rf2 {-0.61/10} Qxb4 {+2.00/10} 32.
Qc2 {-0.63/11} h6 {+2.18/10} 33. Re2 {-0.54/11} Ra7 {+2.13/10} 34.
Re3 {-0.50/10} Qd4 {+2.11/10} 35. Rc3 {-0.76/10} Ra2 {+2.43/10} 36.
Qxa2 {-0.85/12} Qxc3 {+2.32/10} 37. Qb1 {-0.94/11} Qd2+ {+2.87/10} 38.
Kg1 {-0.85/10} Qe3+ {+2.77/10} 39. Kg2 {-1.33/12} Nh7 {+2.77/10} 40.
h4 {-0.50/12} Qd2+ {+2.94/10} 41. Kg3 {-1.11/13} Nf6 {+3.03/11} 42.
Kh3 {-1.26/13} b4 {+3.11/11} 43. Kg3 {-1.37/13} Nd7 {+3.25/11} 44.
Qd1 {-1.39/13} Qc3 {+3.15/12} 45. Kf2 {-1.50/12} Nc5 {+3.38/11} 46.
Bg2 {-1.56/12} Nxb3 {+3.52/10} 47. Qg4 {-1.61/11} Qd4+ {+3.66/8} 48.
Kg3 {-2.37/12} Nd2 {+3.65/9} 49. f6 {-4.06/12} Qc3+ {+3.65/6} 50.
Kf2 {-2.51/12} Qxf6+ {+3.46/7} 51. Ke2 {-2.48/12} Nb3 {+3.86/10} 52.
Qc8+ {-2.40/12} Kh7 {+4.28/9} 53. Qxc7 {-2.61/11} Nd4+ {+3.97/9} 54.
Ke3 {-3.36/12} b3 {+4.88/10} 55. Qc1 {-5.44/11} Nc2+ {+5.95/10} 56.
Ke2 {-4.44/13} Qc3 {+5.93/10} 57. e5 {-6.96/11} Qc4+ {+5.75/8} 58.
Kf2 {-4.27/12} Qxh4+ {+5.79/10} 59. Kg1 {-4.32/13} Qd4+ {+5.88/10} 60.
Kh1 {-5.13/12} Qxe5 {+5.86/11} 61. Qb1 {-6.07/11} Qe3 {+5.86/10} 62.
Bf1 {-6.09/12} Qg3 {+6.24/10} 63. Bb5 {-5.00/11} Qh4+ {+7.29/10} 64.
Kg2 {-5.36/12} Qg5+ {+7.29/10} 65. Kh2 {-4.97/12} Qe5+ {+7.29/10} 66.
Kg1 {-5.99/12} Qxd5 {+7.31/10} 67. Bf1 {-5.55/12} Kh8 {+7.27/11} 68.
Bg2 {-7.46/11} Qd4+ {+7.64/10} 69. Kh2 {-7.76/11} Qe5+ {+7.33/10} 70.
Kg1 {-7.76/11} b2 {+8.29/11} 71. Be4 {-8.35/12} Qxe4 {+8.29/10} 72.
Qxb2 {-7.85/11} Qg4+ {+8.30/9} 73. Kf2 {-7.71/11} Qf4+ {+8.29/9} 74.
Ke2 {-7.72/11} Qe5+ {+8.21/9} 75. Qxe5 {-8.01/15} dxe5 {+8.49/13} 76.
Kd1 {-11.25/15} Ne3+ {+8.81/13} 77. Ke2 {-8.11/15} Nd5 {+8.99/13} 78.
Kf3 {-8.75/14} f5 {+9.07/12} 79. Kg2 {-12.29/14} e4 {+9.51/12} 80.
Kf1 {-14.77/14} e3 {+13.62/12} 81. Ke2 {-20.20/15} h5 {+17.89/13} 82.
Kd3 {-20.54/15} h4 {+18.68/13} 83. Kd4 {-21474205.11/14} e2 {+22.64/11} 84.
Kxd5 {+1672.19/12} e1=Q {+22.77/11} 85. Kc4 {+1672.21/12} Qe6+ {+22.82/9}
86. Kb5 {+1672.21/12} h3 {+22.86/9} 87. Kb4 {+1672.23/13} h2 {+99999.89/9}
88. Kc3 {+1672.25/13} Qb6 {+99999.91/8} 89. Kc4 {+1672.77/12} f4 {+34.44/9}
90. Kd3 {+1672.27/10} Qc5 {+99999.91/8} 91. Kd2 {+1672.27/10}
Qc4 {+99999.93/6} 92. Kd1 {+1672.29/8} h1=Q+ {+99999.95/4} 93.
Kd2 {+1672.31/4} Qhc1# {+99999.97/1}
{Black Mates} 0-1
User avatar
Guenther Simon
 
Posts: 794
Joined: 26 Sep 2004, 19:49
Location: Regensburg, Germany

Re: Tinkering with Winboard

Postby Dann Corbit » 01 Dec 2005, 19:58

Guenther Simon wrote:<quote Dann>
subject: SSEChess with integral time in seconds and centipawn eval


Thanks Dann, the fixed version already won its first game in RWBC :)
I propose you announce it perhaps in a special message or at
Leos too, because I don't know how much read here with us.

I noticed BTW it calls itself 2.050, does this mean it is based on
a slightly newer source than the last official release 2.045, or did
you name it that way for distinction?

Best regards,
Guenther


I don't remember if I modified the version number or if it was Sam. It was 2001 when he sent me his code. We talked about some ideas for improvements but I do not remember if we implemented any of them. So it might be a little stronger or it might just report its thinking a bit differently.
Dann Corbit
 

PreviousNext

Return to Winboard and related Topics

Who is online

Users browsing this forum: Google [Bot] and 12 guests