Opinion: The Chess Community Reaps What They Sow

[matejst]

On the other, it killed other projects that were end-users oriented.

No, they killed themselves by not respecting Stockfish.

Sopel, just like the last time, this is a misunderstanding. I thought here about commercial projects with a GUI, levels, analysis functions, etc. A decade and a half ago it was almost as important as the strength of the engine. I guess you thought I was alluding at Eman, but no -- for me this story is clear and it has ended in an appropriate way for both side.

As a linguist I know how much a written word can be misunderstood. and I have no wish to be confrontational, despite being ready to defend my opinion. When several persons argue in good faith, the truth is somewhere in the middle, and they all have valuable arguments most of the time. So: I am ready to argue in good faith, to try to make myself clear as much as possible, to read the arguments of others and to take them in consideration. I sincerely think you are in the same disposition.

[towforce]

My own conclusion as chess programmer -> go MIT

I see here a lot of new developers, young talents, on the ground who have an alternative interpretation of GPL, use others data for training -> nope, use others code for NNUE implementation -> nope, use others networks for own engine -> nope. Use others ideas -> yes, but you have to implement them in your own way, whatever that means. Maybe this is a matter of Zeitgeist, dunno, things are moving, so I am moving to MIT license to make clear what the spirit behind my code base is.

--
Srdja

IMO, the main difference between MIT and GPL is that under MIT, other users can take your source, modify it, and then keep the new source closed - from you and everyone else.

Why would you want that?

[smatovic]

My own conclusion as chess programmer -> go MIT

I see here a lot of new developers, young talents, on the ground who have an alternative interpretation of GPL, use others data for training -> nope, use others code for NNUE implementation -> nope, use others networks for own engine -> nope. Use others ideas -> yes, but you have to implement them in your own way, whatever that means. Maybe this is a matter of Zeitgeist, dunno, things are moving, so I am moving to MIT license to make clear what the spirit behind my code base is.

--
Srdja

IMO, the main difference between MIT and GPL is that under MIT, other users can take your source, modify it, and then keep the new source closed - from you and everyone else.

Why would you want that?

There is value in a fork even if it is closed source.

--
Srdja

[towforce]

There is value in a fork even if it is closed source.

Wouldn't you at least want to know if your code was being used in another engine?

[Luecx]

My own conclusion as chess programmer -> go MIT

I see here a lot of new developers, young talents, on the ground who have an alternative interpretation of GPL, use others data for training -> nope, use others code for NNUE implementation -> nope, use others networks for own engine -> nope. Use others ideas -> yes, but you have to implement them in your own way, whatever that means. Maybe this is a matter of Zeitgeist, dunno, things are moving, so I am moving to MIT license to make clear what the spirit behind my code base is.

--
Srdja

IMO, the main difference between MIT and GPL is that under MIT, other users can take your source, modify it, and then keep the new source closed - from you and everyone else.

Why would you want that?

There is value in a fork even if it is closed source.

--
Srdja

I think 90% of the community strongly disagree on that. It seems to me that people who fork engines and close the source only want reputation but not people finding out that they actually just use a fork. Beside they do not contribute anything to the chess engine development community.

[smatovic]

There is value in a fork even if it is closed source.

Wouldn't you at least want to know if your code was being used in another engine?

My experience with other gpu chess programmers was mostly positive and friendly, some peek for information, some give something back, regardless of any code base or license. I guess if the exception intends to stay silent on this I will just respect that. But I admit there is a point, if someone is able to speed up my move gen 10 fold I would be really interested in how they accomplished this

--
Srdja

[smatovic]

There is value in a fork even if it is closed source.

--
Srdja

I think 90% of the community strongly disagree on that. It seems to me that people who fork engines and close the source only want reputation but not people finding out that they actually just use a fork. Beside they do not contribute anything to the chess engine development community.

This is a open source philosophy question imo, BSD/MIT vs. GPL. They both play a role in the FOSS world, with MIT (yet) not that present in the computer chess community.

--
Srdja

[smatovic]

There is value in a fork even if it is closed source.

--
Srdja

I think 90% of the community strongly disagree on that. It seems to me that people who fork engines and close the source only want reputation but not people finding out that they actually just use a fork. Beside they do not contribute anything to the chess engine development community.

This is a open source philosophy question imo, BSD/MIT vs. GPL. They both play a role in the FOSS world, with MIT (yet) not that present in the computer chess community.

--
Srdja

MIT chess engines on CPW (Arasan is not listed):

https://www.chessprogramming.org/Category:MIT

and afaik there are public domain implementations for PolyGlot books and MIT implementations for EGTB.

--
Srdja

[JohnWoe]

MIT vs GPLv2 is the reason. Why MacOS and PS5 are FreeBSD instead of Linux.
Or BSD license but there's no difference w/ MIT license. AFAIK

[smatovic]

Clang/LLVM vs. GCC is another example. There are different kind of BSD-clause licenses, so these might not all be straight forward compatible with MIT.

--
Srdja