Wrong "wrong claim" by ChessGUI

[Matthias Gemuh]

Hi Gabor,
only the debug file can reveal whether the claimer was penalized for not preceeding the draw claim with "offer draw".
No tablebases were used, right ?
In that case, the problem can also be that same-coloured bishops are not correctly recognized as draw. I'll check and eventually bugfix.
Such issues get automatically covered by using at least 4-men tablebases.
Regards,
Matthias.

[Matthias Gemuh]

Hi Gabor,
only the debug file can reveal whether the claimer was penalized for not preceeding the draw claim with "offer draw".
No tablebases were used, right ?
In that case, the problem can also be that same-coloured bishops are not correctly recognized as draw. I'll check and eventually bugfix.
Such issues get automatically covered by using at least 4-men tablebases.
Regards,
Matthias.

Hi Matthias,

Matacz used Shawul EGBB's while nanoSzachy used none (not capable).

I try to find the debug file of the game.

Regards,

I have checked and same-coloured bishops are not correctly recognized as draw.
However, if at least 4-men tablebases were used, they would have "solved" the problem even if the engines don't use tablebases.

Best,
Matthias.

P.S.: I shall upload a bugfix later today.

[Matthias Gemuh]

I have checked and same-coloured bishops are not correctly recognized as draw.
However, if at least 4-men tablebases were used, they would have "solved" the problem even if the engines don't use tablebases.

Best,
Matthias.

P.S.: I shall upload a bugfix later today.

Found debug, it ends with:

SendToEng2Time 0000000125836443 : Eng2 (nanoSzachy 3.5) -> time 11806
.
SendToEng2Time 0000000125836443 : Eng2 (nanoSzachy 3.5) -> otim 12767
.
SendToEng2Time 0000000125836443 : Eng2 (nanoSzachy 3.5) -> f3e2
.
SideToMove = Black.
Scheduling GameMoveTimerEvent() for nanoSzachy 3.5 at 0000000125836443.
Extra time for nanoSzachy 3.5 = 0 milliseconds.
==========================================================================================================================
Eng2 (nanoSzachy 3.5) <- Pipe02Time 0000000125836474
Exact move time for nanoSzachy 3.5
Eng2 (nanoSzachy 3.5) <- 1/2-1/2 {Insufficient material}
Now in WB_EndOfGameClaim() at Time0000000125836521

Now in EndTheGame() at Time0000000125836521

1-0
1-0
{ Game Nr. 176 : nanoSzachy 3.5 loses against Matacz 1.4 by wrong claim (1/2-1/2 {Insufficient material})
GameDuration = 00:16:57, Finalposition 8/8/b7/4K3/8/8/3kB3/8 b - - 0 129 }

OK, the draw already pending, so nanoSzachy was right to claim it without sending a move or "offer draw".

Matthias.

[Matthias Gemuh]

BTW, Matthias, would it be difficult to create a version for 1280x1024 screen resolution? Some parts (e.g. results) do not show up in full on my screen.

Best,

The optimal solution would be one that scales well on all screen resolutions,
but I have not yet found out how to do it with my Borland C++ Builder 5 compiler. Maybe I should experiment more.

When I have time, will try scaling for 1280x1024 screen resolution as a temporary solution.

Best,
Matthias.

[Matthias Gemuh]

And how about this one? Was nanoSzachy rightly penalized for claiming draw? In this position, material is not insufficient for mate.

Current position (Ply 253) : 8/8/7k/5KP1/8/4Bn2/8/8 b - - 0 127 = 8/8/7k/5KP1/8/4Bn2/8/8 b - - 0 127
Eng2 (nanoSzachy 3.5) time lag = 62 milliseconds, max. time lag= 109 milliseconds
Eng2 (nanoSzachy 3.5) thought for 1654 milliseconds, time left = 128177 milliseconds
Almost in ParcelMove(from nanoSzachy 3.5) at Time0000000050947773

DetailsOfMove(CAPTURE_MOVE: move f3g5 = Nf3xg5 = Nxg5 = Nf3-g5) -> nSide = Black, nPieceNr = Knight, nFrom = F3, nTo = G5, nCapturedNr = Pawn, promo = NoPiece, ep = A1, nFlags = 40, (Move = 400000012E190200 == 4611686023495746048)

Nf3xg5
Now in MakeChessMove() at Time0000000050947773, nChsProgNr = 1, ChsStrct->PlyPtr = 253

Now in MakeChessMove() at Time0000000050947773, nChsProgNr = 0, ChsStrct->PlyPtr = 253

Now in DrawGraphics() at Time0000000050947773

Now in DrawGraphics() at Time0000000050947773

Memory: 39 MB + 67 MB (814 MB free)
Total Time Elapsed = 00:16:38
MoveNr. = 127, SideToMove = white.
50 moves to 50-Move-Rule.
Expected Move = f3g5
Now in CheckForEndOfGame() at Time0000000050947789

Now in MakeTimeControlString() at Time0000000050947789

MakeTimeControlString 1 : 040 in 000002:35 (127 made)
MakeTimeControlString 1 : 040 in 000002:35 (127 made)
ChessGUI Total Book Moves Count = 4015.
ChessGUI Valid Book Moves Count = 0
Now in ParcelMove() at Time0000000050947789

Now in SendTimeControlString() at Time0000000050947789

SideToMove=w, nInterval = 4, nmoves = 0, imoves = 7, movestogo = 40, wtime = 155000, btime = 155000, winc = 0, binc = 0, PlyPtr = 254, TimeLeft = 131696 ms
SendToEng1Time 0000000050947789 : Eng1 (Trace 1.37a) -> time 13169
.
SendToEng1Time 0000000050947789 : Eng1 (Trace 1.37a) -> otim 12817
.
SendToEng1Time 0000000050947789 : Eng1 (Trace 1.37a) -> f3g5
.
SideToMove = White.
Scheduling GameMoveTimerEvent() for Trace 1.37a at 0000000050947789.
Extra time for Trace 1.37a = 0 milliseconds.
==========================================================================================================================
Eng2 (nanoSzachy 3.5) <- 1/2-1/2 {Insufficient material}
Now in WB_EndOfGameClaim() at Time0000000050947804

Now in EndTheGame() at Time0000000050947804

1-0
1-0
{ Game Nr. 48 : nanoSzachy 3.5 loses against TRACE 1.37a by wrong claim (1/2-1/2 {Insufficient material})
GameDuration = 00:16:38, Finalposition 8/8/7k/5Kn1/8/4B3/8/8 w - - 0 128 }
.

Yes, nanoSzachy was rightly penalized here.
After claiming a wrong "1/2-1/2", almost all WB engines cannot continue the game if GUI ignores the claim.
So the game has to be terminated (ChessGUI and WinBoard).

Matthias.

[Matthias Gemuh]

Hi Gabor,
only the debug file can reveal whether the claimer was penalized for not preceeding the draw claim with "offer draw".
No tablebases were used, right ?
In that case, the problem can also be that same-coloured bishops are not correctly recognized as draw. I'll check and eventually bugfix.
Such issues get automatically covered by using at least 4-men tablebases.
Regards,
Matthias.

Hi Matthias,

Matacz used Shawul EGBB's while nanoSzachy used none (not capable).

I try to find the debug file of the game.

Regards,

I have checked and same-coloured bishops are not correctly recognized as draw.
However, if at least 4-men tablebases were used, they would have "solved" the problem even if the engines don't use tablebases.

Best,
Matthias.

P.S.: I shall upload a bugfix later today.

ChessGUI 0.165 uploaded.

Matthias.