Whether Rybka 4 Should Be Tested As A Derivative Of "Fruit"

[Sean Evans]

http://www.grappa.univ-lille3.fr/icga/f ... s-2010.pdf

18th WORLD COMPUTER CHESS CHAMPIONSHIP TOURNAMENT RULES

2. Each program must be the original work of the entering developers. Programming teams
whose code is derived from or including game-playing code written by others must name all
other authors, or the source of such code, in the details of their submission form. Programs
which are discovered to be close derivatives of others (e.g., by playing nearly all moves the
same), may be declared invalid by the Tournament Director after seeking expert advice. For
this purpose a listing of all game-related code running on the system must be available on
demand to the Tournament Director.

[LetoAtreides82]

I see nothing wrong about derivatives so I voted that it should be tested but I am not certain Rybka 4 counts as a derivative of Fruit, I'm not even sure Rybka 1 Beta counts as a derivative of Fruit. Vas has said time and again that he took some ideas from Fruit for Rybka 1 Beta. Taking ideas from something and making something out of it doesn't make it a derivative. A derivative is when you take something and modify it, Toga for example is a derivative of Fruit.

[lmader]

I think the issue with Rybka 1.0 has more to do with the fact that Vas claimed that he only borrowed ideas from Fruit, and that he did not copy actual code.

There is plenty of hard evidence that Vas actually copied code from Rybka (in fact there is another thread going right now where kingliveson shows an actual snippet. Rather timely, but I also know that this stuff has been shown before repeatedly.

Now I don't know if this necessarily means that Vas copied critical pieces of code from Fruit that would make Rybka 1.0 qualify as a clone. But it certainly means that he hasn't been completely honest about how he used Fruit, and it probably means that there is something of a license violation there (I think Fruit was GPL).

At this point though Rybka 4 is a long way away from all of that. I'm not sure there is any point in speculating about the validity/legitimacy of Rybka 4.

[LetoAtreides82]

I think the issue with Rybka 1.0 has more to do with the fact that Vas claimed that he only borrowed ideas from Fruit, and that he did not copy actual code.

There is plenty of hard evidence that Vas actually copied code from Rybka (in fact there is another thread going right now where kingliveson shows an actual snippet. Rather timely, but I also know that this stuff has been shown before repeatedly.

Now I don't know if this necessarily means that Vas copied critical pieces of code from Fruit that would make Rybka 1.0 qualify as a clone. But it certainly means that he hasn't been completely honest about how he used Fruit, and it probably means that there is something of a license violation there (I think Fruit was GPL).

At this point though Rybka 4 is a long way away from all of that. I'm not sure there is any point in speculating about the validity/legitimacy of Rybka 4.

Can you provide a link to the thread with the proof that Rybka 1 Beta contains Fruit code?

[lmader]

Here's the thread here on OpenChess, see kingliveson's post:
http://www.open-chess.org/viewtopic.php ... 1&start=20

Here's Zach's analysis:
https://webspace.utexas.edu/zzw57/rtc/eval/eval.html

[Matthias Gemuh]

Do you mean: "Should Rybka 4 Be Treated As A Derivative Of "Fruit ?" ?

[Eduard Nemeth]

In my opinion, Rybka has not only taken a few ideas of Fruit (supposedly only ~20 ELO) but much more! Without Fruit, Rybka was not even half as strong as it is now. For me, Rybka is a derivat of Fruit!

BTW: And, of course, Rybka is not a original Engine (like Shredder, Junior, HIARCS)!

ED.

[BB+]

18th WORLD COMPUTER CHESS CHAMPIONSHIP TOURNAMENT RULES

If you go to the ICGA page, you will find that ChessBase is a sponsor. So we already know how the WCCC will decide.

Now I don't know if this necessarily means that Vas copied critical pieces of code from Fruit that would make Rybka 1.0 qualify as a clone. But it certainly means that he hasn't been completely honest about how he used Fruit, and it probably means that there is something of a license violation there (I think Fruit was GPL).
At this point though Rybka 4 is a long way away from all of that. I'm not sure there is any point in speculating about the validity/legitimacy of Rybka 4.

I completely agree with the conclusion here.The question of GPL violation seems to be moot by now, as Rybka 1.0 Beta is no longer distributed, and I don't think anyone has claimed that later versions copy the UCI parsing (or clock management).
[I think Ryan Benitez says that the FSF gave the OK to some Rybka version, but I don't know the details. I've known some people who (with other software) have been through a "license-compliance" process with the FSF -- in that case, it was source-code vs source-code (and the question was when the original code went GPLv3 rather than v2+), but they gave we the impression that the FSF was on top of the license-compliance issue quite well. I'm not sure how great they are at ASM/C++ analysis, which would be more relevant with R1].
The question of "critical code" from Fruit in Rybka is also quite unclear, as most (if not all) of the copied code can be argued to be of lesser importance.

[BTO7]

This is all very interesting. The old saying goes....people in glass houses shouldn't throw stones. Vas should have never picked up the rock and threw it because now its looking more and more like he may of been the one that stepped out of line and not the IPPO family at all. Watch what you wish for I'm not sure what I think here just yet so have not voted but what a real twist of fate. The one crying cheater is usually the one cheating themselves

Regards
BT

[Sean Evans]

Do you mean: "Should Rybka 4 Be Treated As A Derivative Of "Fruit ?" ?

No, I mean "TESTED"!

The WCCC first has to test Rybka 4 to determine it is a derivative or not. Innocent until proven guilty. I have a bad feeling the WCCC tests are so weak any program could slip through as long as the moves are not identical. The WCCC will not reverse engineer Rybka 4, which is what is needed to be done here to debunk this issue! If Rybka 4 has nothing to hide then bring it on.

Cordially,

Sean