Winboard Forum

by **H.G.Muller** » 18 Jan 2008, 16:56

For those who want to hunt bugs, you can download the executable from

http://home.hccnet.nl/h.g.muller/alpha.tst

Rename it to winboard.exe. You can select FRC from the "new variant ..." menu. It seemed to work, engine-engine as well as Human-engine. (Don't forget to save the debug files!)

by **Guenther Simon** » 18 Jan 2008, 18:26

H.G.Muller wrote:For those who want to hunt bugs, you can download the executable from

http://home.hccnet.nl/h.g.muller/alpha.tst

Rename it to winboard.exe. You can select FRC from the "new variant ..." menu. It seemed to work, engine-engine as well as Human-engine. (Don't forget to save the debug files!)

Hello HG,

Just a quick question. Does the new version also fix the bug I mentioned
to you about not/wrong giving of time for the 40th/80th etc. move
in regarding time controls with moves/time?

Regards,
Guenther

by **H.G.Muller** » 18 Jan 2008, 19:19

No, not yet. I want to overhaul the time-control code anyway, and I postpone ficing the bug till after that.

WinBoard_x already supported time-odds games, where you could write things like 5/2:30 in the time field to give white 5 minutes, and black 2:30. (The time was a string anyway, because of the colon.) I would like to extend that (and change the syntax a little) to allow things like:

40|20 (40 min vs 20 min time odds)
60+40/25 (60 min, andafter that, 40 moves per 25 min repeating)
60|30+40/25|12:30 (60 min followed by 40 moves per 25 min (W) vs 30' + 40m/12.5min (B))
60+40/25+0/5 (60 min followed by 40/25 min and then 5 min for the rest of the game)

The number of moves for the first time-control session already has a separate (numeric) input field (or command-line option), as does the increment (which will be only applied during a 'rest-of-the-game' session). How exactly I have to compensate the time for moves straddling sessions will depend on how I implement that.

by **H.G.Muller** » 18 Jan 2008, 20:02

I already fixed a bug that I discovered myself (it could long-castle through check on the square left of the King). Make sure you use an executable that you downloaded after I posted this message!

by **PulsarMike** » 18 Jan 2008, 20:27

Hi,

I've been waiting to submit chess960 bugs tell shatranj was finished, it may be, i got in to late last night to test, i cant test at work unfortunately, connection to icc is turened off only web pages.

I noticed two issues with winboard_x have not tried f but assume it inherits a lot of code. One is castle moves are sent as O-O O-O-O in home play but as 8 different types of moves like d1g1 etc for icc play. The king typically isnt on the square that the move indicates he is on and you have to check when you get one of these moves if its a legal move for anotehr piece first.

This isnt so bad but adjourned game support is broken. I tried to resume a game and got a could not parst move from ics error when ics loaded the moves of the adjourned game.

Mike

by **Pedro Castro** » 18 Jan 2008, 22:12

Hi H.G.,

I have a problem with setboard command.

Instead, setboard rbnqnkbr/pppppppp/8/8/8/8/PPPPPPPP/RBNQNKBR w HAha - 0 1

I hope to receive, setboard rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1

All engines compatible with Chess960 and Arena receive this.

Code: Select all: recognized 'fischerandom' (-1) as variant fischerandom WinBoard 4.3.13 + danasah341frcpgo Reset(1, 0) from gameMode 0 recognized 'fischerandom' (-1) as variant fischerandom GameEnds(0, (null), 2) GameEnds(0, (null), 2) clock stopped GameEnds(0, (null), 2) after test StartChildProcess (dir="") danasah341frcpgo 656 >first : xboard protover 2 672 <first : DanaSah v.3.41 FRC, web: http://danasah.tk 672 <first : 06/11/2007, Pedro Castro, Spain 672 <first : 672 <first : DanaSah only supports the winboard interface. 672 <first : Type command quit to exit 672 <first : 672 <first : Ini file danasah.ini found. 672 <first : ownbook off 672 <first : Table hash size 128 Mb. 687 <first : Eval cache size 16 Mb. 687 <first : Bitbases path: ./egbb/ 687 <first : Bitbases men: 4 687 <first : Bitbases cache size: 4 687 <first : Kibitz off 687 <first : danasah> feature done=0 687 <first : EgbbProbe 3.1 by Daniel Shawul 687 <first : Cache Size = 4 Mb (517 entries) 1140 <first : Loading egbbs.... 1140 <first : feature sigint=0 sigterm=0 setboard=1 draw=0 colors=0 myname="DanaSah v.3.41 FRC" 1140 >first : accepted sigint 1140 >first : accepted sigterm 1140 >first : accepted setboard 1140 >first : accepted draw 1140 >first : accepted colors 1140 >first : accepted myname 1140 <first : feature done=1 1140 >first : accepted done 1156 >first : new random 1156 >first : variant fischerandom write FEN 50-move: 0 0 0 e0. p=-4 1156 >first : setboard rbnqnkbr/pppppppp/8/8/8/8/PPPPPPPP/RBNQNKBR w HAha - 0 1 1156 >first : level 0 5 0 1156 >first : post 1156 >first : hard 1156 >first : easy release Legality test? c1b3 Legality test: c1b3 Legality test: c1d3 Legality test: e1d3 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: b2b3 Legality test: b2b4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d3 Legality test: d2d4 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 FRC castling, 7 0 5 7 0 5 CoordsToAlgebraic, piece=1 FRC castling, 7 0 5 7 0 5 move: c1b3 , parse: Nb3 ( ) 7125 >first : time 30000 7125 >first : otim 30000 7125 >first : c1b3 GameEnds(0, (null), 2) GameEnds(0, (null), 2) clock stopped GameEnds(0, (null), 2) after test 20828 >first : force 20828 >first : quit

by **Pedro Castro** » 19 Jan 2008, 00:58

Hello again,

I am trying to make DanaSah to play in Winboard FRC. I added quotes at the command variant as you indicated in another message.

I have changed the function setboard at the beginning of a game for having full rights of castling, as I have suggested in another forum for compatibility with all engines with protocol xboard playing in Arena FRC, I will send to these engines in the setboard command "KQkq".

After doing this, I have a small problem when playing as a human and engine with Black. After the commands "time" and "otime", the engine does not start (if it does well in the normal variants)

Code: Select all: recognized 'fischerandom' (-1) as variant fischerandom WinBoard 4.3.13 + danasah359frc Reset(1, 0) from gameMode 0 recognized 'fischerandom' (-1) as variant fischerandom GameEnds(0, (null), 2) GameEnds(0, (null), 2) clock stopped GameEnds(0, (null), 2) after test StartChildProcess (dir="") danasah359frc 672 >first : xboard protover 2 688 <first : DanaSah v.3.59 FRC, web: http://danasah.tk 688 <first : 19/01/2008, Pedro Castro, Spain 688 <first : 688 <first : DanaSah only supports the winboard interface. 688 <first : Type command quit to exit 688 <first : 688 <first : Ini file danasah.ini found. 704 <first : ownbook off 704 <first : Table hash size 128 Mb. 704 <first : Eval cache size 16 Mb. 704 <first : Bitbases path: ./egbb/ 704 <first : Bitbases men: 4 704 <first : Bitbases cache size: 4 704 <first : Kibitz off 704 <first : danasah> feature done=0 704 <first : EgbbProbe 3.1 by Daniel Shawul 704 <first : Cache Size = 4 Mb (517 entries) 1219 <first : Loading egbbs.... 1219 <first : feature setboard=1 draw=0 sigint=0 sigterm=0 myname="DanaSah v.3.59 FRC" variants="normal,fischerandom" colors=0 ics=1 1219 >first : accepted setboard 1219 >first : accepted draw 1219 >first : accepted sigint 1219 >first : accepted sigterm 1219 >first : accepted myname 1219 >first : accepted variants 1219 >first : accepted colors 1219 >first : accepted ics 1219 <first : feature done=1 1235 >first : accepted done 1250 >first : new random 1250 >first : variant fischerandom write FEN 50-move: 0 0 0 e0. p=-4 1250 >first : setboard nrkqnrbb/pppppppp/8/8/8/8/PPPPPPPP/NRKQNRBB w FBfb - 0 1 1250 >first : ics - 1250 >first : level 0 5 0 1250 >first : post 1250 >first : hard 1250 >first : easy 1360 <first : variante = normal 1360 <first : variante = fischerandom release Legality test? d2d3 Legality test: a1b3 Legality test: e1d3 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: b2b3 Legality test: b2b4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d3 Legality test: d2d4 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 FRC castling, 5 1 2 5 1 2 CoordsToAlgebraic, piece=0 Legality test? d2d3 Legality test: a1b3 Legality test: e1d3 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: b2b3 Legality test: b2b4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d3 Legality test: d2d4 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 FRC castling, 5 1 2 5 1 2 move: d2d3 , parse: d3 ( ) 4079 >first : time 30000 4079 >first : otim 30000 4079 >first : d2d3 GameEnds(0, (null), 2) GameEnds(0, (null), 2) clock stopped GameEnds(0, (null), 2) after test 21704 >first : force 21704 >first : quit

If I decide that the engine start a game with White, then I if get the command "go", in this case you will see that the engine begins to play and that castling and recognizes the castling the contrary.

Code: Select all: recognized 'fischerandom' (-1) as variant fischerandom WinBoard 4.3.13 + danasah359frc Reset(1, 0) from gameMode 0 recognized 'fischerandom' (-1) as variant fischerandom GameEnds(0, (null), 2) GameEnds(0, (null), 2) clock stopped GameEnds(0, (null), 2) after test StartChildProcess (dir="") danasah359frc 641 >first : xboard protover 2 672 <first : DanaSah v.3.59 FRC, web: http://danasah.tk 672 <first : 19/01/2008, Pedro Castro, Spain 672 <first : 672 <first : DanaSah only supports the winboard interface. 672 <first : Type command quit to exit 672 <first : 672 <first : Ini file danasah.ini found. 672 <first : ownbook off 672 <first : Table hash size 128 Mb. 672 <first : Eval cache size 16 Mb. 672 <first : Bitbases path: ./egbb/ 672 <first : Bitbases men: 4 672 <first : Bitbases cache size: 4 672 <first : Kibitz off 672 <first : danasah> feature done=0 672 <first : EgbbProbe 3.1 by Daniel Shawul 672 <first : Cache Size = 4 Mb (517 entries) 1141 <first : Loading egbbs.... 1141 <first : feature setboard=1 draw=0 sigint=0 sigterm=0 myname="DanaSah v.3.59 FRC" variants="normal,fischerandom" colors=0 ics=1 1141 >first : accepted setboard 1141 >first : accepted draw 1141 >first : accepted sigint 1141 >first : accepted sigterm 1141 >first : accepted myname 1141 >first : accepted variants 1141 >first : accepted colors 1141 >first : accepted ics 1141 <first : feature done=1 1141 >first : accepted done 1156 >first : new random 1156 >first : variant fischerandom write FEN 50-move: 0 0 0 e0. p=-4 1156 >first : setboard nrkqnrbb/pppppppp/8/8/8/8/PPPPPPPP/NRKQNRBB w FBfb - 0 1 1156 >first : ics - 1156 >first : level 0 5 0 1156 >first : post 1156 >first : hard 1156 >first : easy 1250 <first : variante = normal 1250 <first : variante = fischerandom 12344 >first : time 30000 12344 >first : otim 30000 12344 >first : go 12344 <first : 2 0 0 84 a1b3 a8b6 12344 <first : 3 41 0 216 a1b3 a8b6 e1f3 12344 <first : 4 0 0 809 a1b3 a8b6 e1f3 e8f6 12360 <first : 5 37 0 2662 a1b3 a8b6 e1f3 e8f6 b3c5 12360 <first : 6 0 1 7442 a1b3 a8b6 e1f3 e8f6 b3c5 b6c4 12360 <first : 7 8 1 11117 a1b3 a8b6 e1f3 f7f5 d2d3 g8b3 c2b3 e8f6 12391 <first : 8 16 4 26830 a1b3 a8b6 d2d3 e8f6 f2f4 d7d6 g1b6 c7b6 e1f3 12485 <first : 9 8 14 65972 a1b3 a8b6 d2d3 f7f5 f2f4 e8f6 e1f3 g8b3 c2b3 d7d6 12735 <first : 10 19 39 190458 a1b3 a8b6 d2d3 d7d5 f2f4 e8f6 g1b6 c7b6 e1f3 d8d7 e2e3 13141 <first : 11 22 79 389052 a1b3 a8b6 d2d3 d7d5 e1f3 e8f6 f3e5 d8d6 f2f4 O-O-O b3c5 15219 <first : 12 19 287 1453641 a1b3 a8b6 e2e3 f7f5 e1f3 g8d5 d1e2 e8f6 O-O-O d7d6 f3d4 e7e6 19906 <first : 13 37 756 4010228 a1b3 a8b6 d2d3 d7d5 d1d2 d8d6 f2f4 O-O-O g1c5 d6f6 O-O-O e8d6 e1f3 19906 <first : move a1b3 move to parse: a1b3 Legality test? a1b3 Legality test: a1b3 Legality test: e1d3 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: b2b3 Legality test: b2b4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d3 Legality test: d2d4 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 FRC castling, 5 1 2 5 1 2 CoordsToAlgebraic, piece=1 FRC castling, 5 1 2 5 1 2 move: a1b3 , parse: Nb3 ( ) release Legality test? e7e6 Legality test: a7a6 Legality test: a7a5 Legality test: b7b6 Legality test: b7b5 Legality test: c7c6 Legality test: c7c5 Legality test: d7d6 Legality test: d7d5 Legality test: e7e6 Legality test: e7e5 Legality test: f7f6 Legality test: f7f5 Legality test: g7g6 Legality test: g7g5 Legality test: h7h6 Legality test: h7h5 Legality test: a8b6 Legality test: e8d6 Legality test: e8f6 CoordsToAlgebraic, piece=17 Legality test? e7e6 Legality test: a7a6 Legality test: a7a5 Legality test: b7b6 Legality test: b7b5 Legality test: c7c6 Legality test: c7c5 Legality test: d7d6 Legality test: d7d5 Legality test: e7e6 Legality test: e7e5 Legality test: f7f6 Legality test: f7f5 Legality test: g7g6 Legality test: g7g5 Legality test: h7h6 Legality test: h7h5 Legality test: a8b6 Legality test: e8d6 Legality test: e8f6 FRC castling, 5 1 2 5 1 2 move: e7e6 , parse: e6 ( ) 41172 >first : time 29242 41172 >first : otim 27876 41172 >first : e7e6 41188 <first : 2 -7 0 94 d2d3 a8b6 41188 <first : 2 18 0 140 e1f3 a8b6 41188 <first : 3 42 0 306 d2d3 a8b6 e1f3 41188 <first : 4 28 0 731 f2f4 a8b6 g1b6 c7b6 e1f3 41188 <first : 5 31 0 2253 d2d3 a8b6 f2f4 e8f6 e1f3 41203 <first : 5 43 0 3239 d2d3 a8b6 d1d2 b6d5 O-O-O 41203 <first : 6 34 1 8281 d2d3 a8b6 f2f4 d7d6 g1b6 c7b6 e1f3 41203 <first : 7 39 3 13431 d2d3 a8b6 d1d2 d8f6 O-O-O O-O-O e1f3 41250 <first : 8 22 7 45933 d2d3 a8b6 d1d2 d8f6 d2a5 O-O-O a5a7 e8d6 41313 <first : 9 36 14 89381 d2d3 a8b6 e1f3 d8f6 d1d2 O-O-O O-O-O b6d5 b3c5 41453 <first : 10 46 28 177724 d2d3 a8b6 d1d2 d8f6 d2a5 a7a6 f2f4 b6d5 g1d4 f6h4 41828 <first : 11 44 65 419660 d2d3 a8b6 d1d2 d8f6 d2a5 a7a6 f2f4 b6d5 g1d4 f6h4 d4c5 44922 <first : 12 40 375 2355562 d2d3 d8h4 f2f4 a8b6 d1d2 O-O-O O-O-O h4h5 e1f3 e8f6 g1b6 c7b6 e2e4 48297 <first : 13 53 712 4301504 d2d3 d8h4 f2f4 a8b6 d1d2 O-O-O O-O-O h4h5 e1f3 e8f6 g1f2 d7d6 e2e4 48485 <first : move d2d3 move to parse: d2d3 Legality test? d2d3 Legality test: b1a1 Legality test: e1d3 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d3 Legality test: d2d4 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 Legality test: b3a1 Legality test: b3a5 Legality test: b3d4 Legality test: b3c5 FRC castling, 5 1 2 5 1 2 CoordsToAlgebraic, piece=0 Legality test? d2d3 Legality test: b1a1 Legality test: e1d3 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d3 Legality test: d2d4 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 Legality test: b3a1 Legality test: b3a5 Legality test: b3d4 Legality test: b3c5 FRC castling, 5 1 2 5 1 2 move: d2d3 , parse: d3 ( ) release Legality test? d8e7 Legality test: e6e5 Legality test: a7a6 Legality test: a7a5 Legality test: b7b6 Legality test: b7b5 Legality test: c7c6 Legality test: c7c5 Legality test: d7d6 Legality test: d7d5 Legality test: f7f6 Legality test: f7f5 Legality test: g7g6 Legality test: g7g5 Legality test: h7h6 Legality test: h7h5 Legality test: a8b6 Legality test: d8e7 Legality test: d8f6 Legality test: d8g5 Legality test: d8h4 Legality test: e8d6 Legality test: e8f6 CoordsToAlgebraic, piece=21 FRC castling, 5 1 2 5 1 2 move: d8e7 , parse: Qe7 ( ) 64938 >first : time 28509 64938 >first : otim 26231 64938 >first : d8e7 64953 <first : 2 -4 0 84 d1d2 a8b6 64953 <first : 2 29 0 143 e1f3 a8b6 64953 <first : 3 68 0 347 d1d2 a8b6 O-O-O 64953 <first : 4 37 0 708 d1d2 O-O-O d2a5 a7a6 64969 <first : 5 43 0 2987 d1d2 a8b6 O-O-O O-O-O e1f3 64969 <first : 6 31 1 5074 d1d2 O-O-O d2a5 a8b6 a5a7 e8f6 64969 <first : 7 45 3 15004 d1d2 O-O-O O-O-O c8b8 e1f3 a8b6 c1b1 65000 <first : 8 61 6 51178 d1d2 O-O-O d2a5 a7a6 f2f4 a8b6 g1b6 c7b6 a5b6 e8c7 65063 <first : 9 48 12 96586 d1d2 O-O-O d2a5 c8b8 g2g4 a8b6 O-O-O d7d6 h1e4 65281 <first : 10 64 34 297191 d1d2 O-O-O O-O-O c8b8 d2a5 e8f6 f2f3 a8b6 g1d4 b8a8 66781 <first : 11 81 184 1255365 d1d2 a8b6 O-O-O e8d6 d2a5 g7g6 c1b1 a7a6 e1f3 O-O-O b3d4 68688 <first : 12 85 375 2477330 d1d2 a8b6 d2a5 g7g6 f2f4 e6e5 f4e5 e7e5 a5e5 h8e5 e1f3 e5f6 g1b6 a7b6 O-O-O 72141 <first : move d1d2 move to parse: d1d2 Legality test? d1d2 Legality test: b1a1 Legality test: c1d2 Legality test: d1d2 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: c2c3 Legality test: c2c4 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 Legality test: b3a1 Legality test: b3d2 Legality test: b3a5 Legality test: b3d4 Legality test: b3c5 Legality test: d3d4 FRC castling, 5 1 2 5 1 2 CoordsToAlgebraic, piece=4 FRC castling, 5 1 2 5 1 2 move: d1d2 , parse: Qd2 ( ) release Legality test? c8b8 Legality test: e6e5 Legality test: a7a6 Legality test: a7a5 Legality test: b7b6 Legality test: b7b5 Legality test: c7c6 Legality test: c7c5 Legality test: d7d6 Legality test: d7d5 Legality test: e7d6 Legality test: e7c5 Legality test: e7b4 Legality test: e7a3 Legality test: e7f6 Legality test: e7g5 Legality test: e7h4 Legality test: e7d8 Legality test: f7f6 Legality test: f7f5 Legality test: g7g6 Legality test: g7g5 Legality test: h7h6 Legality test: h7h5 Legality test: a8b6 Legality test: c8d8 Legality test: e8d6 Legality test: e8f6 Legality test: c8b8 CoordsToAlgebraic, piece=33 Legality test? c8b8 Legality test: e6e5 Legality test: a7a6 Legality test: a7a5 Legality test: b7b6 Legality test: b7b5 Legality test: c7c6 Legality test: c7c5 Legality test: d7d6 Legality test: d7d5 Legality test: e7d6 Legality test: e7c5 Legality test: e7b4 Legality test: e7a3 Legality test: e7f6 Legality test: e7g5 Legality test: e7h4 Legality test: e7d8 Legality test: f7f6 Legality test: f7f5 Legality test: g7g6 Legality test: g7g5 Legality test: h7h6 Legality test: h7h5 Legality test: a8b6 Legality test: c8d8 Legality test: e8d6 Legality test: e8f6 Legality test: c8b8 FRC castling, 5 1 2 5 -1 -1 move: c8b8 , parse: O-O-O ( ) 77578 >first : time 27789 77578 >first : otim 25689 77578 >first : O-O-O 77594 <first : 2 -4 0 123 O-O-O a8b6 77594 <first : 2 37 0 291 d2a5 a7a6 77594 <first : 3 43 0 399 O-O-O a8b6 e1f3 77594 <first : 3 68 0 834 d2a5 a8b6 a5a7 77594 <first : 4 19 0 1529 O-O-O a8b6 d2a5 a7a6 77594 <first : 4 36 0 1859 d2a5 a7a6 f2f4 e8f6 77594 <first : 4 57 0 2366 f2f4 a8b6 d2a5 e8f6 a5a7 77594 <first : 5 69 0 2746 O-O-O a8b6 d2a5 a7a6 e1f3 77610 <first : 6 61 0 6261 d2a5 a7a6 f2f4 a8b6 g1b6 c7b6 a5b6 e8c7 77610 <first : 7 78 0 7972 O-O-O a8b6 d2a5 a7a6 f2f4 e7h4 g2g4 77610 <first : 8 78 1 13667 O-O-O a8b6 d2a5 a7a6 f2f4 e7h4 g2g3 h4h3 77625 <first : 8 91 4 37145 d2a5 a7a6 f2f4 a8b6 g1b6 c7b6 a5b6 e7d6 b6d6 e8d6 O-O-O 77656 <first : 9 100 7 74546 O-O-O a8b6 d2a5 a7a6 f2f4 e7h4 g1b6 c7b6 a5b6 h4h2 c1b1 78110 <first : 10 76 53 364108 O-O-O a8b6 d2a5 c8b8 f2f4 g7g6 e1f3 e8d6 g1f2 d6c8 79453 <first : 11 99 187 1286190 O-O-O g7g6 d2a5 e7f6 d3d4 c8b8 e1d3 a8b6 c1b1 a7a6 g2g4 81438 <first : 12 79 386 2607181 O-O-O g7g6 f2f4 c8b8 d2a5 a8b6 e1f3 d7d6 e2e4 f7f5 g1f2 a7a6 f2b6 c7b6 a5b6 f5e4 d3e4 f8f4 84531 <first : move O-O-O move to parse: O-O-O Parser FRC long 2 1 Legality test? c1b1 Legality test: b1a1 Legality test: c1d1 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d1 Legality test: d2c3 Legality test: d2b4 Legality test: d2a5 Legality test: d2e3 Legality test: d2f4 Legality test: d2g5 Legality test: d2h6 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 Legality test: b3a1 Legality test: b3a5 Legality test: b3d4 Legality test: b3c5 Legality test: d3d4 FRC castling, 5 1 2 5 1 2 Legality test: c1b1 CoordsToAlgebraic, piece=16 Legality test? c1b1 Legality test: b1a1 Legality test: c1d1 Legality test: e1f3 Legality test: a2a3 Legality test: a2a4 Legality test: c2c3 Legality test: c2c4 Legality test: d2d1 Legality test: d2c3 Legality test: d2b4 Legality test: d2a5 Legality test: d2e3 Legality test: d2f4 Legality test: d2g5 Legality test: d2h6 Legality test: e2e3 Legality test: e2e4 Legality test: f2f3 Legality test: f2f4 Legality test: g2g3 Legality test: g2g4 Legality test: h2h3 Legality test: h2h4 Legality test: b3a1 Legality test: b3a5 Legality test: b3d4 Legality test: b3c5 Legality test: d3d4 FRC castling, 5 1 2 5 1 2 Legality test: c1b1 move: c1b1 , parse: O-O-O ( ) ...

by **H.G.Muller** » 19 Jan 2008, 13:24

In game 1 it seems your engine suddenly exits, after receiving the user move d2d3. The GemeEnds() call with argument NULL suggest that WinBoard could no longer read from the output pipe of the engine. But from the debug you can see that WinBoard really did send the move.

So I think the problem is in your engine. It somehow does not accept the move, and even crashes from it. It should have known that white was on move from the FEN. Perhaps it gets confused by the 'B' in the castling rights, mistaken it for black-to-move?

I already commented on the FRC castling-rights notation on CCC. The engine has to be able to understand the A-Ha-h notation, or your setboard would fail for non-opening positions like

[diag]3kq3/3pp3/8/8/1PP2PP1/8/8/1K3RR1 w[/diag]

where during the game white moved one Rook from a1 to g1. Even Arena will use F to indicate the short castling right for this position, if the King and Rook on f1 have not yet been moved.

If you only want to work setboard correctly for opening positions, you can indeed ignore all FEN fields after the board, as you can be sure that in the opening white has the move, castling with both Rooks is allowed, e.p. is not possible, the reversible-ply counter is at 0 and it is the first move. Reading the castling field of the FEN is only useful for the purpose of handling positions that might not be openings.

by **Teemu Pudas** » 19 Jan 2008, 14:02

H.G.Muller wrote:Even Arena will use F to indicate the short castling right for this position, if the King and Rook on f1 have not yet been moved.

Actually, it won't.

[FEN "brqknrnb/pppppppp/8/8/8/8/PPPPPPPP/BRQKNRNB w KQkq - 0 1 "]

1.f4 d6 2.Rf3 d5 3.b4 d4 4.a4 a6 5.Ra3 d3 6.Bd4 b6 7.Raa1 *

brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Qkq - 1 7

Not surprising, as Arena goes out of its way to be incompatible with the official UCI protocol by rejecting king-takes-rook and neglecting to set UCI_Chess960.

by **H.G.Muller** » 19 Jan 2008, 14:13

Well, this is definitely wrong. From the Q you cannot see if you can castle with Ra1 or with Rb1. The x-FEN standard clearly specifies that Q would mean the outer-most Rook (so Ra1 here) has the castling rights. As I understood it, Shredder-FEN doesn't ever use K and Q.

If Arena is buggy in this respect, we should not pay attention to what it does, as the aim is to have a system that works, and not to introduce the same bugs to remain compatible!

by **Pedro Castro** » 20 Jan 2008, 14:25

Teemu Pudas wrote:
H.G.Muller wrote:Even Arena will use F to indicate the short castling right for this position, if the King and Rook on f1 have not yet been moved.

Actually, it won't.

[FEN "brqknrnb/pppppppp/8/8/8/8/PPPPPPPP/BRQKNRNB w KQkq - 0 1 "]

1.f4 d6 2.Rf3 d5 3.b4 d4 4.a4 a6 5.Ra3 d3 6.Bd4 b6 7.Raa1 *

brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Qkq - 1 7

Not surprising, as Arena goes out of its way to be incompatible with the official UCI protocol by rejecting king-takes-rook and neglecting to set UCI_Chess960.

Well, first let me say that I remain neutral on the war-x-FEN and Shredder-FEN. My opinion is that to support the protocol you have to change something about your function FEN for x-FEN and for Shredder-FEN.

If an engine plays Teemu moves in the starting position that has given, then the engine will have no problem in knowing how the castle must be, as it has moved a rook, it knows that this rook he can not castle, is the engine internally who leads the account.

If we copy the FEN to the clipboard, we have:
brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b - - 1 7

Arena does not indicate anything about the castles, so it is not where the FEN has gone from Teemu.

We can edit a position chess960 in Arena and give castling rights, which would have the FEN to Teemu.

brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Qkq - 1 7

The FEN is incomplete, as you said Teemu and HG

The correct x-FEN would:
brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Bkq - 1 7

The correct shredder-FEN would:
brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Bbf - 1 7

Well, then Arena fails to edit the position because Arena forget this possibility that there may be 2 rooks on one side. (or because they still do not support X-FEN)

But you (HG) have a error, is that this error in Arena is also an error in Winboard, Winboard if we edit a position not question that the rook is good, the GUI Fritz also no asked its configuration and only GUI Shredder it does.

The only full GUI to play chess960 is Shredder GUI, therefore the engines xboard not currently have a full GUI for it, only to play complete games.

by **Pedro Castro** » 20 Jan 2008, 15:04

When I edit the position (Teemu) in Winboard, I receive:

brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b HAha - 0 1

Of course, this is wrong.

The fault of Arena is that they still do not support XFEN (and ShredderFEN?) and maybe with this they not ask which of the 2 rooks make castling. This is the same mistake made by Fritz GUI although they tend to using ShredderFEN

The Winboard fault is that even his edit has no chance of castling rights, nor can indicate other things ep, number (played) or rule of the 50 moves.

by **H.G.Muller** » 20 Jan 2008, 17:42

The 'edit' command I can actually fix, if there is any demand for this. In normal Chess, engines are supposed to assume that Rooks and King starting on the square of the initial position do have castling rights. For positions where the corresponding rights do not exist, because the piece was moved, biut somehow moved back, I could use a similar kludge as old WinBoard uses to load a position where black is to move. It then sets up a normal position, (with 'new'), plays a2a3, and then uses the 'edit' menu to clear the board and set up the new position.

Similarly, castling rights could be destroyed by never putting Rooks that should not have castling rights in a corner, but set up another position than the requested one, and use 'force' mode after the edit to move the Rooks to their corner (so they get there without having castling rights).

Similarly, I could enter the move number and reversible plies by first 'new', 'force', and then playing the Knights up and down for the required number of moves (throwing in a Pawn move at the right moment to set the 50-move counter), and only then start 'edit'. For e.p. rights I would simply set up the position before the double pawn move, and play the latter in 'force' mode.

A hideous kludge, of course. But it would solve all problems for existing engines, while staying completely within protocol 1.

For FRC it would be more difficult, as the engine is not aware of the initial position. But FRC engines using 'edit' mode should assume that any Rook on the back rank is virgin. When this is not desired, WB could set up the Rooks elsewhere, and move them to the requested square through force.

I wonder if this would be worth it though.

by **H.G.Muller** » 20 Jan 2008, 18:43

Pedro Castro wrote:But you (HG) have a error, is that this error in Arena is also an error in Winboard, Winboard if we edit a position not question that the rook is good, the GUI Fritz also no asked its configuration and only GUI Shredder it does.

The error was in the FEN reader: it assumed that K always meant h-file, and Q always a-file. This is of course not correct in x-FEN, so it was only correctly reading Shredder-FEN. And if the initial rights were not correct, playing the moves and saving the FEN kept the faulty rights.

I fixed it now, by actually searching for the outermost Rook when K or Q are encountered in the castling-rights field. After pasting in the game above, it now saves the FEN:
brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Bfb - 1 7

The corrected WinBoard_F executable can be downloaded from http://home.hccnet.nl/h.g.muller/alpha.tst (as usual).

by **F.Huber** » 20 Jan 2008, 19:30

H.G.Muller wrote:I fixed it now, by actually searching for the outermost Rook when K or Q are encountered in the castling-rights field. After pasting in the game above, it now saves the FEN:
brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Bfb - 1 7

Hmm? Is the outermost (white) rook really on b?

by **H.G.Muller** » 20 Jan 2008, 19:40

It was in the original position, described by the FEN with the Q castling right.

by **Teemu Pudas** » 20 Jan 2008, 19:47

F.Huber wrote:Hmm? Is the outermost (white) rook really on b?

No, and that's why Arena's FEN was wrong.

by **F.Huber** » 20 Jan 2008, 19:47

Ok, it seems I just misunderstood your "actually searching for the outermost Rook".

by **H.G.Muller** » 20 Jan 2008, 20:20

I see. So for clarity: that applies to the routine that reads the FEN. While writing the FEN it just uses the rights (stored internally as file numbers) that WinBoard has kept during the game, and writes those. It would do that even if there are no Rooks at all in those files.

Note that WinBoard_F does not check for castling rights given as A-H if there is a Rook in that file. So in principle it is capable of supporting games where you can castle with pieces other than Rooks. (The current MakeMove does recognize such moves as castlings, though. You would simply capture your own piece!

)

Note that in KnightMate the Royal Knight (represented as Unicorn) can castle with its Rooks. Thinking of it, this might not be properly implemented yet...

by **Pedro Castro** » 20 Jan 2008, 21:36

OK, now I understand.

We can give to Winboard a position Shredder-FEN, we are going to the menu, "File" and "Paste Position From Clipboard".

Then Winboard sent to the command setboard with proper rights of castling in the same format Shredder-FEN.

We can not use in Winboard "Edit Position" of menu "Mode", in this way we can not give all information on the position.

In Arena, the routine that reads the FEN, does not recognize xFEN nor ShredderFEN.

Only one thing, I get the command setboard 2 times, the first with castling rights and the second without them, the second should not have those rights also? and perhaps should not coincide the turn and last 2 numbers with the FEN of clipboard?

Code: Select all: PasteTextFromClipboard(): lock count 0 PasteTextFromClipboard(): lock count 0 PasteTextFromClipboard(): err 0 locked 0 PasteFenFromString(): fen 'brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Bfb - 1 7' FEN castling rights: -1 1 3 5 1 3 56515 >first : force 56515 >first : new random 56515 >first : variant fischerandom write FEN 50-move: 1 0 0 e0. p=-4 56515 >first : setboard brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b Bfb - 1 1 56515 >first : ics - 56515 >first : level 0 5 0 56515 >first : post 56515 >first : hard 56531 >first : easy 56531 >first : force write FEN 50-move: 1 1 1 e1. p=-4 56531 >first : setboard brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b - - 1 1 56593 <first : variante = normal 56593 <first : variante = fischerandom StartChildProcess (dir="") danasah359frc 77281 >second: xboard protover 2 77281 <second: DanaSah v.3.59 FRC, web: http://danasah.tk 77281 <second: 19/01/2008, Pedro Castro, Spain 77281 <second: 77281 <second: DanaSah only supports the winboard interface. 77281 <second: Type command quit to exit 77281 <second: 77281 <second: Ini file danasah.ini found. 77281 <second: ownbook off 77281 <second: Table hash size 128 Mb. 77281 <second: Eval cache size 16 Mb. 77281 <second: Bitbases path: ./egbb/ 77297 <second: Bitbases men: 4 77297 <second: Bitbases cache size: 4 77297 <second: Kibitz off 77297 <second: danasah> feature done=0 77297 <second: EgbbProbe 3.1 by Daniel Shawul 77297 <second: Cache Size = 4 Mb (517 entries) 77750 <second: Loading egbbs.... 77750 <second: feature setboard=1 draw=0 sigint=0 sigterm=0 myname="DanaSah v.3.59 FRC" variants="normal,fischerandom" colors=0 ics=1 77750 >second: accepted setboard 77750 >second: accepted draw 77750 >second: accepted sigint 77750 >second: accepted sigterm 77750 >second: accepted myname 77750 >second: accepted variants 77750 >second: accepted colors 77750 >second: accepted ics 77750 <second: feature done=1 77765 >second: accepted done 77781 >second: new random 77781 >second: variant fischerandom write FEN 50-move: 1 1 1 e1. p=-4 77781 >second: setboard brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB w Bfb - 0 1 77781 >second: ics - 77781 >second: level 0 5 0 77781 >second: post 77781 >second: hard 77781 >second: easy 77781 >second: force write FEN 50-move: 1 1 1 e1. p=-4 77781 >second: setboard brqknrnb/2p1pppp/pp6/8/PP1B1P2/3p4/2PPP1PP/RRQKN1NB b - - 1 1 77781 >first : computer 77781 >second: computer 77781 >second: time 30000 77797 >second: otim 30000 77797 >second: go

Winboard Forum

WinBoard_F now has Chess960 support

WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Re: WinBoard_F now has Chess960 support

Who is online