Hello!
Moustique (JS Schach UCI) version 0.4 is available.
Since the previous release, the code has been polished and optimized. The engine plays exactly like the previous version (I believe), but faster and using less memory.
There are Linux and Windows binaries, but only Linux binaries have been tested.
Regards.
Roland
Moustique 0.4
Moderator: Ras
-
Roland Chastain
- Posts: 687
- Joined: Sat Jun 08, 2013 10:07 am
- Location: France
- Full name: Roland Chastain
-
Roland Chastain
- Posts: 687
- Joined: Sat Jun 08, 2013 10:07 am
- Location: France
- Full name: Roland Chastain
Moustique 0.4.1
Moustique 0.4.1 is available.
Changes:
No Windows binaries for now. If you need it, you can build it like this (with Free Pascal):
Otherwise you can also use the included Lazarus project.
Changes:
- Support command position fen ... moves ...
- Fixes in Book unit
No Windows binaries for now. If you need it, you can build it like this (with Free Pascal):
Code: Select all
fpc -Mobjfpc -Sh moustique.pas -dRELEASE -CX -XX -Xs-
Gabor Szots
- Posts: 1482
- Joined: Sat Jul 21, 2018 7:43 am
- Location: Budapest, Hungary
- Full name: Gabor Szots
-
Roland Chastain
- Posts: 687
- Joined: Sat Jun 08, 2013 10:07 am
- Location: France
- Full name: Roland Chastain
Re: Moustique 0.4
Hi Gabor. No, I didn't abandoned it. I will try to propose a new version in 2023. 
-
Roland Chastain
- Posts: 687
- Joined: Sat Jun 08, 2013 10:07 am
- Location: France
- Full name: Roland Chastain
Re: Moustique 0.4
Uploaded a Windows binary.
-
Gabor Szots
- Posts: 1482
- Joined: Sat Jul 21, 2018 7:43 am
- Location: Budapest, Hungary
- Full name: Gabor Szots
-
Roland Chastain
- Posts: 687
- Joined: Sat Jun 08, 2013 10:07 am
- Location: France
- Full name: Roland Chastain
Re: Moustique 0.4
Yes.
The code having been optimized (some variables removed), maybe I could try to increase the search depth and add a time control. But I am not sure that it work, because the best move search is recursive.
I have to test how the program behaves if we increase search depth. (It's possible to do that by editing the INI file.)
Last edited by Roland Chastain on Thu Jan 05, 2023 11:00 am, edited 1 time in total.
-
Gabor Szots
- Posts: 1482
- Joined: Sat Jul 21, 2018 7:43 am
- Location: Budapest, Hungary
- Full name: Gabor Szots
Re: Moustique 0.4
OK, thanks.Roland Chastain wrote: ↑Thu Jan 05, 2023 10:49 amYes.
The code having been optimized (some variables removed), maybe I could try to increase the search depth and add a time control. But I am not sure that it work, because the best move search is recursive.
Gabor Szots
CCRL testing group
CCRL testing group
-
Roland Chastain
- Posts: 687
- Joined: Sat Jun 08, 2013 10:07 am
- Location: France
- Full name: Roland Chastain
Re: Moustique 0.4
A first test, with mindepth or maxdepth increased by 1. Cute Chess options: tc=40/60 timemargin=10000.Roland Chastain wrote: ↑Thu Jan 05, 2023 10:49 am I have to test how the program behaves if we increase search depth. (It's possible to do that by editing the INI file.)
Code: Select all
480 game(s) loaded, 0 game(s) with unknown result ignored.
Rank Name Elo + - games score oppo. draws
1 Moustique 0.4.1 32-bit [mindepth 4] 32 29 29 192 58% -8 72%
2 Moustique 0.4.1 [mindepth 4] 29 29 29 192 57% -7 76%
3 Moustique 0.4.1 32-bit [maxdepth 8] -11 29 29 192 47% 3 74%
4 Moustique 0.4.1 [maxdepth 8] -23 29 30 192 45% 6 67%
5 Moustique 0.4.1 -26 29 29 192 43% 7 74% Code: Select all
+================+======+=====================================+=======+
| File | Game | Engine | Move |
+================+======+=====================================+=======+
| 20230105131150 | 11 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105131150 | 24 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105131150 | 44 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105131150 | 46 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105131150 | 49 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105131150 | 78 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105131150 | 87 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105131150 | 144 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105131150 | 174 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105131150 | 176 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105131150 | 181 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105131150 | 196 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105131150 | 218 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105131150 | 236 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105121706 | 41 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105121706 | 42 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105121706 | 54 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105121706 | 94 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105121706 | 144 | Moustique 0.4.1 [mindepth 4] | 0000 |
| 20230105121706 | 171 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |
| 20230105121706 | 214 | Moustique 0.4.1 32-bit [mindepth 4] | 0000 |The '0000' move comes from here:
Code: Select all
// player.pas
function TChessPlayer.BestMove(out ACode: TExitCode): string;
// ...
var
// ...
i: integer;
begin
result := '0000';
i := FCurPos.BestEval(FCurPos.ActiveColor, 1, 32000);
i := BestMoveIndex(i);
if i > 0 then
begin
// ...
end else
ACode := ecError;