New engine releases 2020

[Luecx]

Koivisto 1.0

Some strangers and I have written this engine for about 4 months now, and we estimate its strength to be between 2800 and 2900.
We're still figuring out the build process, so for now, you might have to compile yourself. It's a single makefile.

https://github.com/Luecx/Koivisto



The engine supports all important UCI commands together with additional print (to print the board) and eval for a detailed overview of the evaluation.

We'd also like to have this run in CCRL, but we're not very familiar with the process. We hope there is interest in testing our engine.

Not clear what figuring out the build process means. And, at your home page you write:

A few compiler warnings about unused functions might pop up. Those functions are only from the syzygy code which have been copied from Fathom

If you don't use syzygy/*.c in your make file, only syzygy/tbprobe.c then all compiler warnings disappear.

Oh thank you very much on this one!
What we mean is that we are not yet able to build a static executable due to marche=native which is required for our SSE/AVX code to run.
Therefor one needs to compile the project in order to get a working executable. We are trying to solve this issue at the moment.

[Jamal Bubker]

Thanks Graham for announcing this new engine on the scene and thanks to its author Jonathan for releasing publicly this new strong engine

[quote="Graham Banks" post_id=858807 time=1599020466 user_id=6]
Copied from CCRL public forum:

[i]Hi all,

I've spent the last couple months shut in at home making my own chess engine, and I think it's finally ready for release and rating!

Name: Jonathan Tseng
Engine name: Beef
Nationality: US
Protocol: UCI
Open source: https://github.com/jtseng20/Beef/releases/tag/v0.2

I estimate from limited fast testing that Beef is around 3000.

I work on a linux machine, so the makefile should work for linux, and my windows compiles are made with visual c++, but I am not completely sure yet if everything is correct. Please tell me if there are any problems with the given binaries. Thanks![/i]
[/quote]

[Jamal Bubker]

Thank you Terje great job for this huge improvment !!

[quote=Terje post_id=858862 time=1599063297 user_id=12185]
Weiss 1.1 released https://github.com/TerjeKir/weiss/releases/tag/v1.1

55 self-play elo at normal time controls, about double that at cyclic time control (1.0 did not handle cyclic well)

Windows binaries provided as usual - you can likely get better performance compiling yourself with "make pgo". This assumes a machine that supports pext, if anyone wants to try compiling with pgo for older machines I can give directions on how to do it, it's simple
[/quote]

[Jamal Bubker]

Thanks Guenther for reporting these new releases

[quote=Guenther post_id=858101 time=1598693470 user_id=4222]
[b]Eubos 1.09[/b]
https://github.com/cjbolt/EubosChess/releases

[b]LC0 0.26.2 rc1[/b]
https://github.com/LeelaChessZero/lc0/releases
[/quote]

[Jamal Bubker]

Thanks Gabor for reporting this new update

[quote="Gabor Szots" post_id=858733 time=1598979940 user_id=10665]
Eubos 1.1.0

New engine: Supernova 1.0.1 by Minkai Yang, USA - https://github.com/MichaeltheCoder7/Supernova
[/quote]

[tmokonen]

Oh thank you very much on this one!
What we mean is that we are not yet able to build a static executable due to marche=native which is required for our SSE/AVX code to run.
Therefor one needs to compile the project in order to get a working executable. We are trying to solve this issue at the moment.

I thought it was recommended not to use march=native for compiles you intend to distribute, but instead to target a specific platform (eg march=haswell for a BMI2 compile).

[Jamal Bubker]

Thank you Finn !!
I hope there will be some windows binary compilations for various architecture
i'm eager to test your engine !

[quote=Luecx post_id=858899 time=1599071871 user_id=12600]
[size=150][b]Koivisto 1.0[/b][/size]

Some strangers and I have written this engine for about 4 months now, and we estimate its strength to be between 2800 and 2900.
We're still figuring out the build process, so for now, you might have to compile yourself. It's a single makefile.

[url]https://github.com/Luecx/Koivisto[/url]



The engine supports all important UCI commands together with additional [b]print [/b](to print the board) and [b]eval[/b] for a detailed overview of the evaluation.

We'd also like to have this run in CCRL, but we're not very familiar with the process. We hope there is interest in testing our engine.
[/quote]

[Luecx]

Oh thank you very much on this one!
What we mean is that we are not yet able to build a static executable due to marche=native which is required for our SSE/AVX code to run.
Therefor one needs to compile the project in order to get a working executable. We are trying to solve this issue at the moment.

I thought it was recommended not to use march=native for compiles you intend to distribute, but instead to target a specific platform (eg march=haswell for a BMI2 compile).

march=native is recommended if you compile yourself. Sadly there seems to be an issue with -static in general. Hopefully, this will be resolved by the end of this week.

[Luecx]

Thank you Finn !!
I hope there will be some windows binary compilations for various architecture
i'm eager to test your engine !

Koivisto 1.0

Some strangers and I have written this engine for about 4 months now, and we estimate its strength to be between 2800 and 2900.
We're still figuring out the build process, so for now, you might have to compile yourself. It's a single makefile.

https://github.com/Luecx/Koivisto



The engine supports all important UCI commands together with additional print (to print the board) and eval for a detailed overview of the evaluation.

We'd also like to have this run in CCRL, but we're not very familiar with the process. We hope there is interest in testing our engine.

Thank you very much! I hope there will be binaries by the end of this week

[Gabor Szots]

We'd also like to have this run in CCRL, but we're not very familiar with the process. We hope there is interest in testing our engine.

There is no process. CCRL tries to test all engines that generate interest, let it be public or personal. Every member makes his choice using his discretion.