cutechess-cli for macOS

[ydebilloez]

Dear all,
I am looking for a cutechess-cli binary for macOS.

I came across a version 1.0.0 in a fish test project, but it relies on some dylibs and on a local install of the QT framework and creating additional user folders.

Looking for version 1.2 without any dependencies.

[amanjpro]

Dear all,
I am looking for a cutechess-cli binary for macOS.

I came across a version 1.0.0 in a fish test project, but it relies on some dylibs and on a local install of the QT framework and creating additional user folders.

Looking for version 1.2 without any dependencies.

Compile from the source, that is what I did

[Ras]

Looking for version 1.2 without any dependencies.

Did you consider c-chess-cli that was designed to not have dependencies? https://github.com/lucasart/c-chess-cli

[purechess]

Looking for version 1.2 without any dependencies.

Did you consider c-chess-cli that was designed to not have dependencies? https://github.com/lucasart/c-chess-cli

Problem with c-chess-cli is that engines are hanging very often. So it's not very reliable for engine matches.

[Ras]

Problem with c-chess-cli is that engines are hanging very often.

I've ever only seen this with buggy engines because they discard input during search, but mishandle the race condition between resetting the discard flag and new valid input already arriving. That becomes especially visible at fast time controls with all logical CPU cores fully loaded.

[purechess]

Problem with c-chess-cli is that engines are hanging very often.

I've ever only seen this with buggy engines because they discard input during search, but mishandle the race condition between resetting the discard flag and new valid input already arriving. That becomes especially visible at fast time controls with all logical CPU cores fully loaded.

The issue is posted on Github . Check it here

https://github.com/lucasart/c-chess-cli/issues/33

[Ras]

The issue is posted on Github.

From the issue discussion: "I think it would be nice to be able to treat that engine failure as a single defeat rather than aborting the whole tournament." So the root cause is an unresponsive engine, not problem with c-chess-cli, which is why DIE() is called in the first place. I've had the same with buggy engines, and ideally, that would be a bug report for the engine author - typically caused by the race condition I mentioned, within the engine.