The Real #2 Chess Engine

[chrisw]

Stockfish is, of course, very impressive. strongest chess engine in the world for a very long time.
But I think Torch and Reckless are a lot more impressive when you consider the strength as a function of team size and total resources.
The fact that a small team can compete with a gargantuan project with enormous engineering and testing resources is truly astonishing.

If someone produces the fastest car in the world together with detailed instructions of how to put it all together, plus all other details, why is it either surprising or impressive that someone else or many else’s will do more or less the same, either by source code copying (which we tend to disapprove of) or by conceptually understanding it so well (thank you documentation authors btw) that they’re able to do more or less the same thing albeit a bit different (which we tend to find okay)?

[Dann Corbit]

Stockfish is, of course, very impressive. strongest chess engine in the world for a very long time.
But I think Torch and Reckless are a lot more impressive when you consider the strength as a function of team size and total resources.
The fact that a small team can compete with a gargantuan project with enormous engineering and testing resources is truly astonishing.

If someone produces the fastest car in the world together with detailed instructions of how to put it all together, plus all other details, why is it either surprising or impressive that someone else or many else’s will do more or less the same, either by source code copying (which we tend to disapprove of) or by conceptually understanding it so well (thank you documentation authors btw) that they’re able to do more or less the same thing albeit a bit different (which we tend to find okay)?

I am a dinosaur relic from the past. As an ACM member, I used to look at the journal transactions where people would post their algorithms in the hopes that other people would use them and improve them. I think it is the competitive nature of chess that makes people angry when their ideas get used.
To me, I never understood why people got mad about similar evaluation terms. After all, evaluation terms are nothing more than explanations GMs made about chess which have been codified into books and then into code. An early example of this kind of thing is "My System" by Hans Berliner where the same person went right from the chess analysis to the code with explanations. There were also lots of interesting things in the early publications where the ideas were presented clearly for scholarly examination. The work of Ed Scjrpder in describing how chess programs work is another interesting and old way this sort of thing was communicated (Bruce Moreland's page also springs to mind).
Somewhere along the way people decided that competition was more important than advancement, perhaps. Of course, chess programmers have always needed to obey laws like copyright protections, GPL license restrictions, etc.
But I have also clearly seen examples where people thought that they owned algorithms (without a patent) that I find distasteflu.

I don't mean to start a war on cloning and not cloning and those sorts of ideas. But I have personally always thought it was better to share than to protect ideas.