Open-source improvements released

[kranium]

Hi all-

I've been playing quite a bit with 3 open-source programs, made some progress and have put it all up on Github:

Fridolin 2.00
- code optimizations, eval tuning, etc.
published on GitHub as Jinx 1.0 (~ +146 Elo)
https://github.com/FireFather/jinx

Bobcat 8.0
- code optimizations, eval tuning, etc.
published on GitHub as Tomcat 1.0 (~ +45 Elo)
https://github.com/FireFather/tomcat

Gull 3
published on GitHub as Seagull 1.0 (~ +5 Elo)
https://github.com/FireFather/seagull

Seagull changes:
- source code cleaned up, simplified, and broken up into various source and header files
- a couple of VS code analysis fixes (ex: gen_kpk() was causing stack exceed error)
- compiler warnings resolved up to level 4
- benchmark and perft utilities added (type 'bench' and engine will write a date-stamped text file with results)
- support for syzygy tablebases

All include Visual Studio 2015 project files and x64 binaries.

Best Regards,
Norm

[Gusev]

Thank you very much, Norm, for your contributions to the open-source community! This is very nice of you!

[cdani]

Many thanks for all this nice work!!

[Norbert Raimund Leisner]

It seems that two other current open-source UCI compatibles engines have problems with their time management. Unter the graphical interfaces ChessGUI/Arena during a 40/2 round robin tournament the termination is often called "time (forfeit).

RuyDos 1.0.14 SSE4.2 x64
https://bitbucket.org/alonamaloh/ruydos/downloads/

WyldChess 1.5 x64
https://github.com/Mk-Chan/WyldChess/releases

I have sent the pgn-games including debug-files/output to Manik and Álvaro.

Who of you (esspecially CEGT/CCRL testers) made the same experiences?

Best wishes,
Norbert

[Graham Banks]

It seems that two other current open-source UCI compatibles engines have problems with their time management. Unter the graphical interfaces ChessGUI/Arena during a 40/2 round robin tournament the termination is often called "time (forfeit).

RuyDos 1.0.14 SSE4.2 x64
https://bitbucket.org/alonamaloh/ruydos/downloads/

WyldChess 1.5 x64
https://github.com/Mk-Chan/WyldChess/releases

I have sent the pgn-games including debug-files/output to Manik and Álvaro.

Who of you (esspecially CEGT/CCRL testers) made the same experiences?

Best wishes,
Norbert

No time losses for Wyldchess 1.5 64-bit yet.
Still testing RuyDos 1.0.2 64-bit. No time loss issues there either.

[jdart]

Nice job, but the changelog doesn't really tell me what you did (I am looking at Tomcat). Practically all the commit messages just say "Changes".

--Jon

[kranium]

Nice job, but the changelog doesn't really tell me what you did (I am looking at Tomcat). Practically all the commit messages just say "Changes".

--Jon

Hi Jon-

I'm not sure to what you're referring as "changelog".
I didn't include any such document with the upload.

The commit messages offer very little because I uploaded the whole thing for the 1st time yesterday.

I did create a short README.md for the upload:
- code optimizations and eval tuning
(off the top of my head in an effort to be more specific, I remember: bit operations, psq tuning, mobility tuning, etc)

I track versions using a right-click directory/file sync context diff utility...
comparing it to the original release in this manner would be the easiest way to see the changes.

Norm

[jdart]

I guess then those are Bobcat commits I am looking at.

Still, what most people do if fork a repository in git and then make incremental commits to that. If you make a bunch of changes offline and then upload them (or do something like merge --squash in git) then you have no history of your changes for others to see.

--Jon

[kranium]

I guess then those are Bobcat commits I am looking at.

Still, what most people do if fork a repository in git and then make incremental commits to that. If you make a bunch of changes offline and then upload them (or do something like merge --squash in git) then you have no history of your changes for others to see.

--Jon

Completely agree.
Just always admired Bobcat and Fridolin coding style and techniques used and thought they may have some real potential.
Unfortunately, I never planned to make these efforts public until recently.

But I do still have all the preceding versions (30 for Tomcat), so in fact I could create a repository that incrementally reflected the changes...
just not sure if it's worth the effort, but I'm willing to consider it.

Norm

[JVMerlino]

Marvelous! If I make Myrddin open-source will you also make it not awful?