New engine releases 2020

[Gabor Szots]

After 2 years hyatus:

Devel 3.0.0a: http://www.develchessengine.no/

Can't download with Chrome on Win10, it detects a virus in the zip file.

I've checked it via VirusTotal. All but one anti-virus software has found it healthy. The one exception was Microsoft: Trojan:Win32/Wacatac.C!ml.

Reported the issue to the author.

Version 3.0.0b is out. Direct link (link on site not updated yet): http://www.develchessengine.no/Devel-3.0.0b.zip. Per used a different compiler this time, I myself have no problem downloading and using the program.

Well, I have updated my Win10 to 1909 and updated Windows Defender as well and now they do not complain about version 3.0.0a.

[supersharp77]

https://github.com/tschmoderer/iathena

IATHENA

Free Open Source UCI Chess Engine

gitHub release Build Status license

IATHENA is a chess engine built in C++ using the UCI protocol. IATHENA is deployed on Lichess. If online, please feel free to drop a challenge.

Getting Started

You are welcome to use IATHENA for your own personnal projects, to clone the repo :

git clone https://github.com/tschmoderer/iathena.git
IATHENA code is located in CPP_IATHENA folder. Run make to compile your own version of IATHENA.

cd src
make
If you want to deploy IATHENA on Lichess please check the Lichess API documentation. And use lichess-bot wrapper to deploy.

You can deploy IATHENA in any chess server you like, please let the author know you have deployed IATHENA, see the Contact section below.

Documentation

The project is fully documented with doxygen and is available at https://tschmoderer.github.io/iathena/.

Ratings

Evolution of IATHENA rating

Authors

Timothée Schmoderer - Initial work - https://github.com/tschmoderer
See also the list of contributors who participated in this project.

License

This project is licensed under the GNU General Public License - see the LICENSE file for details.

Acknowledgments

IATHENA first version is inspired by the tutorial of Bluefever Software on YouTube.
The authors would like to thank all the Lichess developpement team for their inspiring work. We are glad to deploy our engine on this webside.
IATHENA plays on Lichess via the lichess-bot framework
Testing framework based on cutechess-cli
ELO computing based on ordo
Skeleton of this README is based on template by PurpleBooth.
Great inspirations from Stockfish, Shallow Blue and ToppleChess

[supersharp77]

https://github.com/codeandkey/neocortex

neocortex

about

neocortex is a UCI chess engine, designed to be efficient and powerful.

You can play against it here!

architecture

Prinipcal variation search with iterative deepening
Bitboard representation, magic bitboard move generation
Transposition table with Zobrist hashing
Incremental position update, lazy PST evaluation
Time control management
UCI interface
dependencies

neocortex will build on any gcc supporting c99.

To build docs, you will need Doxygen installed and available in your PATH.

building

To build neocortex, execute make in the project root.

To install the neocortex binary to /usr/bin, execute make install as superuser.

docs

To build the html documentation, execute make doc in the project root. The documentation is then available at doc/html/index.html.

[fabianVDW]

I have releasd a new version of FabChess at https://github.com/fabianvdW/FabChess/r ... /tag/v1.14 :

This version increases the playing strength of FabChess. It is the same version which played at TCEC, as development hasn't been much in the last months. It is considerably stronger than the last released version so that it is worth to release. In the meantime, thanks for all the testers having shown interest in FabChess.

Changes made include:

- Decreased aspiration window size
- The whole TT has changed with TCEC. It is now lockless and the legality of each tt move is guaranteed, as it led to crashes
- Only probe TT in root of qsearch
- TT is allocated in parallel if threads >= 1 set in UCI
- Increase see pruning depth, change see capture values
- Fail lows will now never be used as best moves, and if the current best move fails low on an thread, more time will be used
- Correct reporting of mate scores
- Move ordering now done in more stages, and with MVVLVA score first, less SEE

Strength increase via selfplay around 50 elo

Fabi

[Gabor Szots]

Thanks Fabi. I also think it is worth testing.

[silentshark]

Francesca 0.28 is released today. You can download from www.silentshark.co.uk

From the readme file:

"More small changes. Bugfix to move ordering. Slight tweak to king eval. Slight search enhancements,
such as not trying a null move on the PV. Better endgame (draw etc.) recognition. Trick of searching
fail highs to reduced depth. Self-testing showing +60 ELO improvement. Testing vs other engines,
somewhat less."

[Gabor Szots]

Francesca 0.28 is released today. You can download from www.silentshark.co.uk

From the readme file:

"More small changes. Bugfix to move ordering. Slight tweak to king eval. Slight search enhancements,
such as not trying a null move on the PV. Better endgame (draw etc.) recognition. Trick of searching
fail highs to reduced depth. Self-testing showing +60 ELO improvement. Testing vs other engines,
somewhat less."

Thank you Tom. Seems you are making nice progress.

[silentshark]

Francesca 0.28 is released today. You can download from www.silentshark.co.uk

From the readme file:

"More small changes. Bugfix to move ordering. Slight tweak to king eval. Slight search enhancements,
such as not trying a null move on the PV. Better endgame (draw etc.) recognition. Trick of searching
fail highs to reduced depth. Self-testing showing +60 ELO improvement. Testing vs other engines,
somewhat less."

Thank you Tom. Seems you are making nice progress.

Cheers, some progress, yes. +60 ELO is probably a bit optimistic. But 0.28 is definitely, measurably stronger than 0.27.

[CMCanavessi]

Did something happen between Devel 3.0.0a and 3.0.0b ? There's a difference of 100 elo in CCRL between the 2, with 3.0.0a being stronger. How can that be the case? 3.0.0b is even weaker than Devel 2. Weird.

[Gabor Szots]

Did something happen between Devel 3.0.0a and 3.0.0b ? There's a difference of 100 elo in CCRL between the 2, with 3.0.0a being stronger. How can that be the case? 3.0.0b is even weaker than Devel 2. Weird.

We are investigating the case. I can't tell you more, we don't understand it either.