Sachy by Martin Macok retouched for CuteChess

[Guenther]

...While I was working on time management, I believe I understood why you couldn't get it to work: I believe it's because of the time control that you used. For now Sachy only supports classical TC (not incremental). Maybe I will change that in the next version...
Regards.

Roland

A lot of old xboard programs did not support Fischer time controls.

[Roland Chastain]

I had zero problems with the current makefile (with march=native) in msys2 + gcc

Good to know. I will make another attempt later. Would you share your binary? I would put it on the release page.

A lot of old xboard programs did not support Fischer time controls.

Indeed.

[Guenther]

I had zero problems with the current makefile (with march=native) in msys2 + gcc

Good to know. I will make another attempt later. Would you share your binary? I would put it on the release page.

A lot of old xboard programs did not support Fischer time controls.

Indeed.

Roland, I'll send you a PM with a link soon.

[Carlos777]

Good news!

I made a quick test on Windows: the engine works. While I was working on time management, I believe I understood why you couldn't get it to work: I believe it's because of the time control that you used. For now Sachy only supports classical TC (not incremental). Maybe I will change that in the next version.

I made a new release: Sachy 0.2.0.4

You are right about Sachy not supporting incremental time controls (although it lost one game on time in standard tc, but after 37 moves)
I downloaded both versions from your site.

[pgn][Event "Computer Chess Game"]
[Site "DESKTOP-VK18VPD"]
[Date "2021.03.10"]
[Round "1"]
[White "Sachy 0.2.0.4 w64"]
[Black "Sachy 0.2.0.4 w32"]
[Result "0-1"]
[TimeControl "40/120"]
[Annotator "3. -0.10 3... +0.10"]

1. d4 e6 2. c4 Bb4+ 3. Bd2 {-0.10/6 2.6} Bxd2+ {+0.10/5 1.1} 4. Nxd2
{-0.10/6 0.9} Qf6 {+0.10/5 0.9} 5. e3 {-0.10/6 8} Nc6 {+0.10/5 0.9} 6. Qa4
{+0.16/5 0.6} Qd8 {-0.16/4 0.7} 7. e4 {+0.16/5 1.3} Nxd4 {+0.20/5 2.3} 8.
Ngf3 {-0.20/6 6} Qf6 {+0.20/5 2.5} 9. e5 {+0.03/5 1.9} Nxf3+ {+0.22/5 4}
10. Nxf3 {-0.27/6 1.4} Qf5 {+0.27/5 0.8} 11. Qb3 {+0.03/5 1.0} Ne7
{+0.22/5 0.8} 12. Qe3 {-0.20/6 5} Qc2 {+0.20/5 3} 13. Qd2 {-0.20/6 2.5}
Qxd2+ {+0.08/6 2.3} 14. Nxd2 {-0.08/7 5} Nc6 {+0.08/6 7} 15. Nf3
{-0.19/6 2.7} Ne7 {+0.19/5 0.8} 16. Bd3 {-0.10/6 1.0} f5 {+0.10/5 0.8} 17.
exf6 {-0.11/6 7} gxf6 {+0.00/6 6} 18. Nd4 {-0.11/6 5} a6 {+0.11/5 4} 19.
Be4 {-0.11/4 1.0} Rg8 {+0.17/5 1.0} 20. f4 {+0.04/5 1.0} c5 {+0.20/5 4} 21.
Nb3 {+0.03/5 0.8} d6 {+0.19/5 1.1} 22. Bf3 {+0.03/5 0.7} d5 {-0.03/4 2.8}
23. Nxc5 {+0.04/5 1.1} dxc4 {+0.09/5 2.1} 24. Ne4 {-0.09/4 0.6} Kd8
{-0.15/4 1.4} 25. Nxf6 {+0.14/5 3} Rf8 {-0.12/6 6} 26. Nxh7 {-0.04/6 8}
Rxf4 {-0.10/6 1.8} 27. b3 {-0.08/6 9} Rh4 {+0.08/5 0.8} 28. Nf6 {-0.02/6 5}
Rf4 {+0.02/5 2.4} 29. Ne4 {+0.13/5 2.3} Nf5 {-0.14/6 12} 30. bxc4
{-0.06/6 9} Nd4 {+0.06/5 4} 31. Rd1 {+0.00/6 3} e5 {+0.00/5 1.6} 32. Nc3
{+0.03/6 5} Bg4 {-0.03/5 5} 33. Bxb7 {+0.09/6 5} Ra7 {-0.09/5 0.6} 34. Bf3
{+0.09/6 5} Rd7 {-0.09/5 1.7} 35. Bxg4 {+0.51/5 3} Rxg4 {-0.10/5 3} 36. Rd2
{+0.25/5 2.1} Rd6 {-0.09/5 8} 37. c5 {+0.54/5 1.0} Nc2+ {-0.54/4 2.8}
{Black wins on time} 0-1

[Event "Computer Chess Game"]
[Site "DESKTOP-VK18VPD"]
[Date "2021.03.10"]
[Round "2"]
[White "Sachy 0.2.0.4 w32"]
[Black "Sachy 0.2.0.4 w64"]
[Result "1/2-1/2"]
[TimeControl "40/120"]
[Annotator "3. +0.04 3... +0.17"]

1. d4 e6 2. c4 Bb4+ 3. Nc3 {+0.04/5 1.6} c6 {+0.17/5 1.1} 4. Qa4
{+0.05/5 4} a5 {+0.19/5 0.9} 5. a3 {+0.14/5 1.0} Bxc3+ {+0.09/5 4} 6. bxc3
{-0.10/6 4} d5 {+0.10/5 1.4} 7. Nf3 {+0.15/5 2.3} Qf6 {-0.15/4 1.7} 8. g4
{+0.15/5 10} dxc4 {-0.15/4 1.2} 9. Bg5 {+0.15/5 0.9} Qg6 {+0.23/5 1.3} 10.
Qxc4 {+0.14/5 10} f6 {-0.04/6 10} 11. Bf4 {+0.04/5 4} Qxg4 {-0.02/6 6} 12.
Bxb8 {-0.23/4 0.9} Rxb8 {+0.22/5 1.5} 13. Rg1 {+0.10/5 2.2} Qe4
{+0.20/5 0.9} 14. Rxg7 {+0.27/5 1.1} Ne7 {+0.17/5 1.0} 15. Qc5 {+0.28/5 4}
Nf5 {+0.24/5 1.9} 16. Qa7 {+0.25/5 8} Qf4 {+0.19/5 1.9} 17. Rg1
{+0.29/5 13} b6 {+0.15/5 12} 18. e3 {-0.15/4 1.6} Nxe3 {+0.42/5 0.8} 19.
Bd3 {+0.07/5 25} Rb7 {-0.07/4 1.4} 20. Qa8 {-0.14/4 1.1} Rc7 {-0.16/4 1.0}
21. c4 {-0.27/4 1.6} Qxf3 {+0.15/6 3} 22. Qb8 {-0.15/5 4} Nc2+
{+0.64/5 0.8} 23. Bxc2 {-0.55/6 1.4} Qc3+ {+0.43/6 2.1} 24. Ke2
{-0.69/6 1.1} Qxc2+ {+0.43/6 1.8} 25. Ke3 {-0.92/6 2.0} Qc3+ {+0.43/6 2.0}
26. Ke4 {-0.92/6 3} Qc2+ {+0.43/6 2.3} 27. Ke3 {-0.92/6 1.2} Qc3+
{+0.43/6 3} 28. Ke4 {-0.92/6 2.7} Qc2+ {+0.43/6 2.7} 29. Ke3 {-0.92/6 1.8}
{XBoard adjudication: repetition draw} 1/2-1/2

[Event "Computer Chess Game"]
[Site "DESKTOP-VK18VPD"]
[Date "2021.03.10"]
[Round "1"]
[White "Sachy 0.2.0.4 w64"]
[Black "Sachy 0.2.0.4 w32"]
[Result "1-0"]
[TimeControl "60+1"]

1. e4 e6 2. Nf3
{White wins on time} 1-0

[Event "Computer Chess Game"]
[Site "DESKTOP-VK18VPD"]
[Date "2021.03.10"]
[Round "2"]
[White "Sachy 0.2.0.4 w32"]
[Black "Sachy 0.2.0.4 w64"]
[Result "0-1"]
[TimeControl "60+1"]

1. d4 f5
{Black wins on time} 0-1[/pgn]

[Roland Chastain]

You are right about Sachy not supporting incremental time controls (although it lost one game on time in standard tc, but after 37 moves)
I downloaded both versions from your site.

Thank you for the test. Yes, the engine now loses some games on time. I will try to find a solution.

But it seems to play better.

Code: Select all

Rank Name                          Elo    +    - games score oppo. draws 
...
  22 Sachy 0.2.0.4                 844   57   56   160   62%   737   17% 
  23 Iota 0.1                      834   41   40   264   61%   774   22% 
  24 Nero 6.0 32-bit               830   72   70    88   63%   698   15% 
  25 Nero 6.0                      823   67   64   102   67%   662   15% 
  26 Belofte 2.1.0                 756   56   55   144   52%   774   16% 
  27 Monchester 1.0                636   48   48   160   44%   698   28% 
  28 TuxedoCat 0.0.2               635   71   73    88   32%   867   24% 
  29 Hippocampe 0.4.2.0.1          629   67   67    96   44%   719   18% 
  30 Hippocampe 0.4.2.0.1 32-bit   619  149  145    16   50%   647   38% 
  31 Moustique 0.3                 559   43   43   224   42%   711   30% 
  32 Moustique 0.3 32-bit          553   71   72    72   42%   712   36% 
  33 Hippocampe 0.4.2.0.2          547   72   77    80   27%   785   16% 
  34 Sachy 0.2.0.3 32-bit          503   62   61   100   57%   445   22% 
  35 Sachy 0.2.0.3                 477   31   31   388   47%   521   29% 
  36 Sachy 0.2.0.2                 463   42   42   236   48%   506   25% 

[Roland Chastain]

Hello!

A new revision of Sachy is available. It handles correctly a zero value for the first parameter of the level CECP command. Thanks to Carlos who reported the bug.

You can find source code and Linux binaries here. (I am moving all my projects to GitLab.) I don't have the possibility to make Windows binaries for now.

Regards.

Roland

[Guenther]

Hello!

A new revision of Sachy is available. It handles correctly a zero value for the first parameter of the level CECP command. Thanks to Carlos who reported the bug.

You can find source code and Linux binaries here. (I am moving all my projects to GitLab.) I don't have the possibility to make Windows binaries for now.

Regards.

Roland

This time I had to use makefile1 (instead of the original makefile) because of linker errors in msys2 gcc.
I compiled a native and an amdfam10 one for Carlos in case my native one doesn't run on his machine.

Also this time I got two warnings.

Code: Select all

Guenther Simon@CAPPUCCINO MINGW64 ~/sachy_0205
$ make
gcc -O3 -march=native -Wall   -c -o main.o main.c
gcc -O3 -march=native -Wall   -c -o inic.o inic.c
gcc -O3 -march=native -Wall   -c -o interf.o interf.c
gcc -O3 -march=native -Wall   -c -o movegen.o movegen.c
gcc -O3 -march=native -Wall   -c -o value.o value.c
gcc -O3 -march=native -Wall   -c -o rules.o rules.c
gcc -O3 -march=native -Wall   -c -o makemove.o makemove.c
makemove.c:432:34: warning: argument 2 of type 'char[6]' with mismatched bound [-Warray-parameter=]
  432 | int BestMove2(TGame *aGame, char aMoveStr[/*5*/6], int aOutput)
      |                             ~~~~~^~~~~~~~~~~~~~~~
In file included from makemove.c:11:
data.h:145:24: note: previously declared as 'char *'
  145 | int BestMove2(TGame *, char *, int); /* najde a provede nejlepsi tah */
      |                        ^~~~~~
gcc -O3 -march=native -Wall   -c -o search.o search.c
gcc -O3 -march=native -Wall   -c -o test.o test.c
gcc -O3 -march=native -Wall   -c -o sort.o sort.c
gcc -O3 -march=native -Wall   -c -o book.o book.c
gcc -O3 -march=native -o sachy main.o inic.o interf.o movegen.o value.o rules.o makemove.o search.o test.o sort.o book.o

I can send the two binaries today.

[Roland Chastain]

Thank you Günther.

This time I had to use makefile1 (instead of the original makefile) because of linker errors in msys2 gcc.

Weird.

Also this time I got two warnings.

This is weird too. I only changed a line in the code. Or maybe I forget something that I did... Anyway I will try to fix the warnings. Thank you for the compiler output. (If someone who knows C better than I do has an idea, he is welcome.)

I can send the two binaries today.

Nice! But maybe it would be simpler that you publish them on your website? So we don't waste time to exchange files. And it would probably be more visible than on my git repository.

[Guenther]

Thank you Günther.

This time I had to use makefile1 (instead of the original makefile) because of linker errors in msys2 gcc.

Weird. :?

May be not so weird, because I have updated gcc a few times and installed a very lot of things in my msys2 since then
(also for being able to compile rust/cargo etc) and who knows about backward compatibility to very old code, or whatever.
Warnings also might change of course if things get deprecated etc... (for your point below).

Also this time I got two warnings.

This is weird too. I only changed a line in the code. Or maybe I forget something that I did... Anyway I will try to fix the warnings. Thank you for the compiler output. (If someone who knows C better than I do has an idea, he is welcome.)

I can send the two binaries today.

Nice! But maybe it would be simpler that you publish them on your website? So we don't waste time to exchange files. And it would probably be more visible than on my git repository. :)

No problem, then it would make sense though I also make the previous ones available.

BTW will you close your github after your migration to gitlab?

Thanks for the update BTW!

[Roland Chastain]

May be not so weird, because I have updated gcc a few times and installed a very lot of things in my msys2 since then
(also for being able to compile rust/cargo etc) and who knows about backward compatibility to very old code, or whatever.

I see.

BTW will you close your github after your migration to gitlab?

For now, I will let a README with a redirection. The old releases stay available. But maybe later I will delete the repository. So if you wish to host some files, it could be a good idea.

Thanks for the update BTW!

You're welcome. Thank you for your kind support.