cutechess-cli 0.7.0

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

User avatar
ilari
Posts: 750
Joined: Mon Mar 27, 2006 7:45 pm
Location: Finland

Re: cutechess-cli 0.7.0

Post by ilari »

voyagerOne wrote:Not related to the GUI...but I got a quick question regarding random opening.

If we have concurrency > 1, do the processes share the same seed?


Thanks in advance.
Cutechess-cli runs in a single process. It uses threads which all share the same random seed. Randomization (indexing) of the openings is done before the match even begins. This guarantees that cutechess-cli doesn't pick the same opening twice until all openings have been played at least once.
Michel
Posts: 2272
Joined: Mon Sep 29, 2008 1:50 am

Re: cutechess-cli 0.7.0

Post by Michel »

Hi Ilari,

Have you seen this

http://talkchess.com/forum/viewtopic.php?t=57465 ?

I have run some simulations and the posted code really works. Perhaps you are interested in simplifying the SPRT implementation in cutechess-cli?

Another advantage is that this is a clean implementation of a standard algorithm (the GSPRT) with proven properties.

Michel
Ideas=science. Simplification=engineering.
Without ideas there is nothing to simplify.
User avatar
Desperado
Posts: 879
Joined: Mon Dec 15, 2008 11:45 am

Re: cutechess-cli 0.7.0

Post by Desperado »

ilari wrote:
Ferdy wrote:Here is the output from updated code. It is all about PGN file cannot be opened.

Code: Select all

Started game 257 of 900 (D2015.1.35.247 vs Arasan_18_x64_popcnt)
Finished game 254 (Gaviota v1.0 64bit vs D2015.1.35.247): 0-1 {Black wins by adjudication}
Started game 258 of 900 (Arasan_18_x64_popcnt vs D2015.1.35.247)
Finished game 256 (SmarThink v1.7 vs D2015.1.35.247): 0-1 {White loses on time}
Finished game 255 (D2015.1.35.247 vs SmarThink v1.7): 1-0 {Black loses on time}
Started game 260 of 900 (DiscoCheck 5.2.1 vs D2015.1.35.247)
Started game 259 of 900 (D2015.1.35.247 vs DiscoCheck 5.2.1)
Started game 261 of 900 (D2015.1.35.247 vs Hakkapeliitta 3.0 x64)
Finished game 258 (Arasan_18_x64_popcnt vs D2015.1.35.247): 1/2-1/2 {Draw by 3-fold repetition}
Warning: PGN file cannot be opened
Warning: Can't write to PGN file .\games\g1.pgn
Warning: PGN file cannot be opened
Warning: Can't write to PGN file .\games\g1.pgn
Warning: PGN file cannot be opened
Warning: Can't write to PGN file .\games\g1.pgn
Finished game 250 (D2015.1.35.247 vs HIARCS 14 WCSC): 0-1 {Black wins by adjudication}
Thanks. I actually managed to reproduce a similar problem but without warning messages: when I used really fast games, at some point the PGN output would just stop. This on Windows 10 with an SSD HD. I just pushed a commit that fixes this problem and makes PGN output more efficient: https://github.com/cutechess/cutechess/ ... a87ed18d1a

Could you try fetching the latest code (which includes the above commit), compile and see if you can still reproduce the "Can't write to PGN file" error? You'll have to recompile the whole project (lib + cli).
Hello.
First, thanks for this great tool. The latest Windows executable does not contain the fix, or does it ?
Well, is someone able to provide a new compile because the version i am using is still not writing all pgn games to the file.
I really would like to upgrade my old but stable version 0.5.1 :) (with version 0.6 i did not have any luck too)
Thank you in advance.
Ferdy
Posts: 4833
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: cutechess-cli 0.7.0

Post by Ferdy »

Desperado wrote:
ilari wrote:
Ferdy wrote:Here is the output from updated code. It is all about PGN file cannot be opened.

Code: Select all

Started game 257 of 900 (D2015.1.35.247 vs Arasan_18_x64_popcnt)
Finished game 254 (Gaviota v1.0 64bit vs D2015.1.35.247): 0-1 {Black wins by adjudication}
Started game 258 of 900 (Arasan_18_x64_popcnt vs D2015.1.35.247)
Finished game 256 (SmarThink v1.7 vs D2015.1.35.247): 0-1 {White loses on time}
Finished game 255 (D2015.1.35.247 vs SmarThink v1.7): 1-0 {Black loses on time}
Started game 260 of 900 (DiscoCheck 5.2.1 vs D2015.1.35.247)
Started game 259 of 900 (D2015.1.35.247 vs DiscoCheck 5.2.1)
Started game 261 of 900 (D2015.1.35.247 vs Hakkapeliitta 3.0 x64)
Finished game 258 (Arasan_18_x64_popcnt vs D2015.1.35.247): 1/2-1/2 {Draw by 3-fold repetition}
Warning: PGN file cannot be opened
Warning: Can't write to PGN file .\games\g1.pgn
Warning: PGN file cannot be opened
Warning: Can't write to PGN file .\games\g1.pgn
Warning: PGN file cannot be opened
Warning: Can't write to PGN file .\games\g1.pgn
Finished game 250 (D2015.1.35.247 vs HIARCS 14 WCSC): 0-1 {Black wins by adjudication}
Thanks. I actually managed to reproduce a similar problem but without warning messages: when I used really fast games, at some point the PGN output would just stop. This on Windows 10 with an SSD HD. I just pushed a commit that fixes this problem and makes PGN output more efficient: https://github.com/cutechess/cutechess/ ... a87ed18d1a

Could you try fetching the latest code (which includes the above commit), compile and see if you can still reproduce the "Can't write to PGN file" error? You'll have to recompile the whole project (lib + cli).
Hello.
First, thanks for this great tool. The latest Windows executable does not contain the fix, or does it ?
Well, is someone able to provide a new compile because the version i am using is still not writing all pgn games to the file.
I really would like to upgrade my old but stable version 0.5.1 :) (with version 0.6 i did not have any luck too)
Thank you in advance.
I could share my compile, but not today, my internet connection is not good.
I am also waiting for the official fix in FEN header tag value where hmvc and fmvn are missing in the output pgn when start position is from epd file. Also the possibility to write the termination header tag in every pgn, be it as draw by 3-fold repetition, white mates, draw by insufficient material and others, time forfeit is already there.
User avatar
ilari
Posts: 750
Joined: Mon Mar 27, 2006 7:45 pm
Location: Finland

Re: cutechess-cli 0.7.0

Post by ilari »

Desperado wrote:Hello.
First, thanks for this great tool. The latest Windows executable does not contain the fix, or does it ?
Well, is someone able to provide a new compile because the version i am using is still not writing all pgn games to the file.
I really would like to upgrade my old but stable version 0.5.1 :) (with version 0.6 i did not have any luck too)
Thank you in advance.
Now the latest Windows executable does contain the fix. It's just not official yet because we haven't updated all the other binaries. But feel free to download cutechess-cli 0.7.1 here: http://koti.mbnet.fi/~ilaripih/bin/cute ... -win32.zip
Ferdy
Posts: 4833
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: cutechess-cli 0.7.0

Post by Ferdy »

ilari wrote:
Desperado wrote:Hello.
First, thanks for this great tool. The latest Windows executable does not contain the fix, or does it ?
Well, is someone able to provide a new compile because the version i am using is still not writing all pgn games to the file.
I really would like to upgrade my old but stable version 0.5.1 :) (with version 0.6 i did not have any luck too)
Thank you in advance.
Now the latest Windows executable does contain the fix. It's just not official yet because we haven't updated all the other binaries. But feel free to download cutechess-cli 0.7.1 here: http://koti.mbnet.fi/~ilaripih/bin/cute ... -win32.zip
Just uploaded my win 32 compile of version 0.7.1, because in your site it was still 0.7.0, under "Installing". I have not uploaded the dll files though.
User avatar
Desperado
Posts: 879
Joined: Mon Dec 15, 2008 11:45 am

Re: cutechess-cli 0.7.0

Post by Desperado »

Good morning,

thx a lot, Ilari + Ferdy, for the new version. :D