Plagiarism and Rybka

[Zach Wegner]

It is possible that you are right. But it is inconceivable to me that anyone can be sure of it.

Well, you might start by thinking what it would take to be sure.

A rhetorical question: If the author of Sloppy did not say that he copied his evaluation from Fruit, and did not make his engine open source, would you consider that to be plagiarism?

[Uri Blass]

zach, are you refering to some new evicdence, when you say "plagerism definitely applies." The old "evidence" seems to be insufficient, at least in the minds of at least four programers on this forum and the original author of fruit. buy maybe you have uncovered something else?

It seems you have some insider info.
Can you share with us, please, if Fabien said the "old" evidence you reffered to is insufficient?

He was probably referring to Fabian's decision on Strelka. Fabain has never said anything about Rybka so far as I know. And since he has signed over the rights to FSF, I guess it is FSF who should be consulted.

But Strelka != Rybka, right? So his argument is flawed.

I agree with that.
However, I *guess* Fabian would say that there is nothing wrong with the similarity of his UCI parser and the UCI parser in Rybka 1.0.
On the other hand, the FSF could possibly take exception since they now own the copyright.
So Fabian is something of a red herring here, despite the code's origin.

Personally, I think it would be nice to have all code in the world as open source. But it seems that when innovation becomes difficult we seek out another target that is reachable.

My opinion is that:
1. It is very clear that Vas has inspiration from Fruit in Rybka 1.0 beta.
2. It is not at all clear to me that the way that Vas applied what he learned was somehow different than every other chess programmer on the planet who makes an honest attempt at innovation when incorporating new ideas that they learned from outside sources.

4. Vas clearly gave credit to Fruit for some of the ideas in Rybka (a model which other programmers should follow but very few do).
5. I am not an expert on legal stuff, so I could be wrong about everything.

7. I honestly do not know if Vas has done anything wrong or not, and I think the people who say that they know that Vas is in the wrong are extruding some kind of strange clarivoyance that makes no sense to me.

I know: he did something wrong.
Not from a moral viewpoint (I mean he may not be found guilty in a tribunal). He did it from the ethical viewpoint.

2. It is not at all clear to me that the way that Vas applied what he learned was somehow different than every other chess programmer on the planet who makes an honest attempt at innovation when incorporating new ideas that they learned from outside sources.

Well, then you think that original attempts of Thomas Gaksch and Norman Schmidt had been honest attempts. I cannot agree with on that. My conviction is that now their attempts are honest, but in the beginning they had not been.

4. Vas clearly gave credit to Fruit for some of the ideas in Rybka (a model which other programmers should follow but very few do).

Do you really think he gave credit to Fruit´s author? I don´t think so.
Let me explain my point:
Let´s suppose that Daniel Shawul, Scorpio´s (open source engine´s) author, give his written credit to Fruit´s author, Fabien Letouzey.

They both use the "same language": "open source code language", and a simple written credit is enough to settle things.

On the other hand, when a commercial engine´s author (engine with closed source code) try to do the same: to settle things with an open source engine author by a simple written credit, it is not correct anymore. Why? Because they do not speak the same language.

A commercial engine´s author speaks a "money language", which is not the same as "open source code language".

Conclusion: if a commercial engine´s author really wants to give credits to an open source code author, he should give written credits and "money credits" in a form of royalties - authorial rights.

When a commercial engine´s author gives only written credits to an open source code author, it is a mockery.


5. I am not an expert on legal stuff, so I could be wrong about everything.

No need to be "an expert on legal stuff". Enough to know what you would not like that others do to yourself.

I think that the comparison with toga is clearly wrong.

Toga has more than 90% of fruit and it is not the case with Rybka.
Rybka clearly does not have knowledge that fruit has unlike toga that knows to evaluate simple endgames.

Uri

[Tony Thomas]

It is possible that you are right. But it is inconceivable to me that anyone can be sure of it.

Well, you might start by thinking what it would take to be sure.

A rhetorical question: If the author of Sloppy did not say that he copied his evaluation from Fruit, and did not make his engine open source, would you consider that to be plagiarism?

I thought he only copied a material value table from Fruit, and not the whole eval. He thought that it would be considered plagiarism and hence released it as open source, so the answer would be yes.

[Zach Wegner]

I thought he only copied a material value table from Fruit, and not the whole eval. He thought that it would be considered plagiarism and hence released it as open source, so the answer would be yes.

I thought it was the whole eval, but I could be mistaken. He definitely took significant portions of the eval, and it caused many to say that Sloppy was not "original".

[Uri Blass]

I thought he only copied a material value table from Fruit, and not the whole eval. He thought that it would be considered plagiarism and hence released it as open source, so the answer would be yes.

I thought it was the whole eval, but I could be mistaken. He definitely took significant portions of the eval, and it caused many to say that Sloppy was not "original".

If Sloppy is not original what can we say about other engines that started their life as tscp or other engines?

Sloppy at least did not start his life as fruit and has original bitboard move generator and other original parts.

I believe that
People criticized sloopy only because the programmer made it open source and in case of not making it open source and saying nothing you could expect nobody to complain because Sloppy is not espacially strong engine so you could expect people not to investigate too much about it.

Note that I am not against trace and my point is that if people are against sloppy they should be also against other engines for the same reason and for some reason I did not hear them complaining against other engines.

It seems that if you take parts from weak engines and are honest about what you do then nobody complains but if you take parts from strong engines people start to complain.

http://www.members.optusnet.com.au/~john.boyd/

Nobody of the people who complained against sloppy asked to forbid the participation of trace in tournaments.

I am also not against trace but I am also not against sloppy.

Uri

[Zach Wegner]

I believe that
People criticized sloopy only because the programmer made it open source and in case of not making it open source and saying nothing you could expect nobody to complain because Sloppy is not espacially strong engine so you could expect people not to investigate too much about it.

But in that case, even if nobody noticed, would that make it right?

I will note that I am not against Sloppy either, precisely because he explained exactly what he did and gave due credit.

[Dirt]

Actually I was wondering about the same thing. I hope that plagiarism is not seen as "the way to get things done."

It seems to work ok in mathematics.