A Very Novel Idea Concerning Vas- BE FAIR

[ozziejoe]

I changed my mind about rybka. I bought earlier version and most recently rybka 4. I even defended rybka repeatedly on this forum, based on the observation that it produced such different assessments than fruit (which i also bought btw).

However, gradually the evidence has built against rybka. I learned that tiny changes to the programing can result in radically different assessments,so can't use output to assess plagerism. I saw experts and people i respected presenting evidence of overlaping code. I saw this evidence argued against again and again. then strelka came on the scene and the same kind of evidence that had been presented against rybka was now presented against strelka, but now people accepted it. Vas said it was a clone and suddenly that was enough.

Fabien's statement was the last straw for me. I don't care what improvement vas made afterwards. My buest guess is that he plagerized (don't know what the official legal term is but who cares).

Translating the novel " War and Peace" to spanish does not mean that you wrote war and peace. If you pretend that you wrote it, that would be plagerism. Even if you added a chapter, or improved some sections, you would still be plagerizing war and peace.

if all this is wrong, I am happy to hear how. Particularly, maybe Vas can show his code to a couple of experts and this can all be resolved. Or he can say something on the subject. I'll stay open. Until then, I will not by any more rybkas

[K I Hyams]

Who is sane enough to think that Vas only won the titles because he copied Fruit and not his extraordinary talent as a programmer?

Me... it does not (appear) to take "extraordinary talent" to take a World Class chess engine and convert it to Bitboard (with a few personal changes in the code). Knowledge and talent... sure... but it is the kind I can do without.

Is this the kind of talent demonstrated in a University? Dollars to Donuts he would have been booted from a Grad program.

He proved his talents after the alleged crime. Being at the top for a long time and by far is no accident. It can't be achieved by reverse engeneer a program 150 Elo weaker than Rybka. Fruit itself was never the strongest engine. Every contender could have used all the Fruit ideas in a completely legal way by rewriting every single idea. Should have been easy to be at the top also for others, if there would not have been that extraordinary talent of the Rybka developer. A talent which the ippo cloners lack. A talent which R. Houdart seems to have as we can judge by now.

Please stop mixing the issues "copy crime" and "non-talent". Looking back at Rbyka's history there is no self-explaining connection.

He may have proved less than you imply "after the alleged crime". If you want to claim that he added extraordinary strength to Fruit, you have to compare the increase in strength that he added to Fruit to the increase in strength that other programmers were able to add to add to Fruit. In other words, instead of comparing the strength of Rybka 1 to Fruit, compare it to the amateur spin-off, Toga. The strength difference between the two was not significant.

I understand that conversion to bitboard is a largely technical process; mechanical and laborious rather than requiring flair. It is also a process that gives an almost certain increase in strength. I wonder whether the amateur programmer of Toga actually made that conversion in the early versions of Toga. If he didn't, time may not have been the only advantage that Rajlich held.

Fabien admitted that he makes cryptic comments. Perhaps the fact that he chose not to expand that statement is part of the game. Either way, he has on at least two occasions made an interestingly unclear comment about the programming of Toga.

[michiguel]

What is surprising in these posts is that there is this short term memory loss issue.


(1) Strelka was outed as a Rybka clone by the one man who I think is an expert on Rybka - its author! There was no court case on this issue. A declaration was made and accepted by most.

(2) Strelka was then revealed to have a lot of Fruit in it. Again, I do not hear this being denied. What I hear is the fantastic proposition that an chess engine can be a clone of Rybka BUT have the Fruit parts which it contains excluded.

This is not fantastic, and it is not a speculation, this was admitted by Strelka's author few days ago.

Miguel
you can scroll some messages to find it
http://www.talkchess.com/forum/viewtopi ... bka#392680

(3) I hear all sorts of standards of testimony - legal - being deemed necessary to settle this matter. Why exactly? I think that it does a signal disservice to the other clones of weaker engines which were outed and their authors banished to computerchess limbo with no legal fanfare at all.

(4) Who really lives their real life this way? I see someone in my neighbour's yard and hear my neighbour yell "Thief!" The person attempts to flee but is apprehended and various items from the neighbour's yard are taken from a bag in the possession of this person. Now you all can wait for the trial if you like but that man is a thief. I know it and anyone in that position knows it too.

I am NOT declaring that Vas is or not. I really am not technically competent to evaluate the code snippets etc. What I want is the same field for the supposed clones about which it seems that less evidence is being exposed....

The issue now becomes funny to me because one of the arguments of the no Fruit involved in Rybka side was that Rybka was so much stronger than Fruit - or anything else incidentally - on the serious, professional rating lists that if anything was taken it was at best, minimal.

Now the argument is that as Rybka is the strongest engine on said lists, any engine which is demonstrably stronger than Rybka must be a Rybka clone.

Huh?

I think that the hypocrisy and nonsense should stop and the two sides agree that science or evidence or facts has nothing to do with it. Both sides believe in what they believe and that belief will not accept a contrary view.

It is a lot simpler to dismiss that which does not accord with our world view - whatever that is.

Maybe if the two sides did that we could cease this pretence of considering evidence when what is being done is a considering of evidence which lines up with what we believe.

Later.


Ps. Rob Osborne, I am congratulating you, not for the final view you came to but at least for having the guts to admit publicly that you feel that you were wrong and that you can change. The process is the thing to me, not the final result.

[Tom Barrister]

Who is sane enough to think that Vas only won the titles because he copied Fruit and not his extraordinary talent as a programmer?

I'm that sane. Fruit was at or near the top. Naturally, if Rybka was a near-copy of Fruit, supposedly rewritten for bitboards, and whatever else, it would be logical that it would start near the top, too. Of course it would need to improve over time. How it improved is open to debate.

I couldn't say how talented Mr. Rajlich is as a programmer. Frankly, my belief is that his biggest asset might well be the ability to aggregate the best ideas/code/routines/whatever of others and combine them into Rybka. That could possibly be done without any significant unique ideas of his own. That's not a novel concept.

Frankly, it's a moot point. It doesn't matter to me how talented Mr. Rajlich is as a programmer. It doesn't matter if Rybka 4 was rewritten from scratch and contains no Fruit code whatsoever (beyond the generic things that all chess programs may need). It doesn't matter if Mr. Rajlich is the greatest chess programmer of all time. Based on what I've read and seen, Rybka at some point in its closed-source existence (perhaps even to a large degree at the present) is/was an almost direct derivative of Fruit. If it wasn't a clone/copy, it was nearly so, rewritten or not for whatever new technologies/code/etc. existed. That's against the rules. If Mr. Rajlich writes an entirely new program from scratch with an ELO of 4000, then good for him... but HE STILL BROKE THE RULES. Breaking the rules, then lying and saying he didn't, then trying to cover up by saying that he lost the source code, and also refusing to let a trusted independent source verify once and for all his claims that he didn't cheat..... well, if it looks like bull manure and smells like bull manure, then it's probably bull manure. It smacks of charlatanry, and the fact that his forum is so heavily moderated doesn't help his credibility. Why is there a need to hide anything if he's innocent.

The "He doesn't have to prove himself to anybody" line is another heap of bull manure. This isn't one crank with an isolated complaint. There's substantial doubt among many that he's broken the rules. Somebody in that position doesn't do his credibility any good by sweeping it all under the rug. He's made no real effort to prove us all wrong. Why not? If he's innocent, he can prove it easily enough. In fact, if he were innocent, I would dare guess that he'd have done so instantly, just to show us all how wrong we were. In my book and that of others, he's guilty as charged, and saying "he's a great programmer who improved Rybka and set the bar higher for all others to improve" (assuming that's true) doesn't change the fact that he's guilty.

And now to address some other issues.

Some Rybka defenders keep quoting that Rybka has advanced X number of ELO (the figure changes). These people don't point out that a lot of other programs also advanced in ELO during the same time period. They also neglect to say that some of the ELO increase might be due to bitboards or whatever else changed, multiple cores, 64 bit vs 32 bit, improved functionality put forth by somebody and shared by whoever wished, and any other improvements that were available to the programming masses.

I'm a civil person by nature, provided I'm treated in the same manner. Those who are civil to me get a civil reply, regardless of whether I agree with them. Those who start unwarranted attacks lose the right to civility and deserve the same ball bat that they tried to use on me, except that I probably won't be as crude with them as they were with me.

One other thing: I don't need to be a programmer to understand what's going on here. If a=b and b=c, then a=c. If Rybka=Strelka, and Strelka=Fruit, then Rybka=Fruit. It's that simple to me. Trying to muddle matters by saying that they aren't exactly alike, Rybka retooled many Fruit things or added some things, etc., is just putting another diversionary spin.

Calling others names, even if said names are true, doesn't justify any wrongdoing by the accused. Changing the topic, accusing others of wrongdoing, attacking the credibility of the accusers: these are all common tactics used when there's no sound defense.

[Roger Brown]

This is not fantastic, and it is not a speculation, this was admitted by Strelka's author few days ago.

Miguel
you can scroll some messages to find it
http://www.talkchess.com/forum/viewtopi ... bka#392680

Hello Miguel,

You are aware that the same author also says

Rybka has a lot of matches with Fruit

,right? What does that mean?

I am not an author of a chess engine so I will await developments as to how the Fruit parts and the Rybka parts worked so well together to produce a smoothly functioning and strong engine - Strelka - which somehow manages to be a clone of one engine but not the other.

Later.

[Milos]

This is not fantastic, and it is not a speculation, this was admitted by Strelka's author few days ago.

Please stop that nonsense. The guy currently writing on CCC under pseudonym of Juri Osipov in perfect english is not more probable to be the author of Strelka than you.

[Rolf]

Who is sane enough to think that Vas only won the titles because he copied Fruit and not his extraordinary talent as a programmer?

I'm that sane. Fruit was at or near the top. Naturally, if Rybka was a near-copy of Fruit, supposedly rewritten for bitboards, and whatever else, it would be logical that it would start near the top, too. Of course it would need to improve over time. How it improved is open to debate.

I couldn't say how talented Mr. Rajlich is as a programmer. Frankly, my belief is that his biggest asset might well be the ability to aggregate the best ideas/code/routines/whatever of others and combine them into Rybka. That could possibly be done without any significant unique ideas of his own. That's not a novel concept.

Frankly, it's a moot point. It doesn't matter to me how talented Mr. Rajlich is as a programmer. It doesn't matter if Rybka 4 was rewritten from scratch and contains no Fruit code whatsoever (beyond the generic things that all chess programs may need). It doesn't matter if Mr. Rajlich is the greatest chess programmer of all time. Based on what I've read and seen, Rybka at some point in its closed-source existence (perhaps even to a large degree at the present) is/was an almost direct derivative of Fruit. If it wasn't a clone/copy, it was nearly so, rewritten or not for whatever new technologies/code/etc. existed. That's against the rules. If Mr. Rajlich writes an entirely new program from scratch with an ELO of 4000, then good for him... but HE STILL BROKE THE RULES. Breaking the rules, then lying and saying he didn't, then trying to cover up by saying that he lost the source code, and also refusing to let a trusted independent source verify once and for all his claims that he didn't cheat..... well, if it looks like bull manure and smells like bull manure, then it's probably bull manure. It smacks of charlatanry, and the fact that his forum is so heavily moderated doesn't help his credibility. Why is there a need to hide anything if he's innocent.

The "He doesn't have to prove himself to anybody" line is another heap of bull manure. This isn't one crank with an isolated complaint. There's substantial doubt among many that he's broken the rules. Somebody in that position doesn't do his credibility any good by sweeping it all under the rug. He's made no real effort to prove us all wrong. Why not? If he's innocent, he can prove it easily enough. In fact, if he were innocent, I would dare guess that he'd have done so instantly, just to show us all how wrong we were. In my book and that of others, he's guilty as charged, and saying "he's a great programmer who improved Rybka and set the bar higher for all others to improve" (assuming that's true) doesn't change the fact that he's guilty.

And now to address some other issues.

Some Rybka defenders keep quoting that Rybka has advanced X number of ELO (the figure changes). These people don't point out that a lot of other programs also advanced in ELO during the same time period. They also neglect to say that some of the ELO increase might be due to bitboards or whatever else changed, multiple cores, 64 bit vs 32 bit, improved functionality put forth by somebody and shared by whoever wished, and any other improvements that were available to the programming masses.

I'm a civil person by nature, provided I'm treated in the same manner. Those who are civil to me get a civil reply, regardless of whether I agree with them. Those who start unwarranted attacks lose the right to civility and deserve the same ball bat that they tried to use on me, except that I probably won't be as crude with them as they were with me.

One other thing: I don't need to be a programmer to understand what's going on here. If a=b and b=c, then a=c. If Rybka=Strelka, and Strelka=Fruit, then Rybka=Fruit. It's that simple to me. Trying to muddle matters by saying that they aren't exactly alike, Rybka retooled many Fruit things or added some things, etc., is just putting another diversionary spin.

Calling others names, even if said names are true, doesn't justify any wrongdoing by the accused. Changing the topic, accusing others of wrongdoing, attacking the credibility of the accusers: these are all common tactics used when there's no sound defense.

Hi Tom,

I broadened your in your thoughts strongest accusations and reflections. My short answer is this:

a) computerchess for the Rybka author is a business

b) he isnt a fool opening his secrets of his talent how he did what he did, because his collegues are proof that copying Fruit alone wouldnt be enough for the top

c) the strongest injustice against Vas is to say that he had to open his secrets but anybody else had NOT to do that

d) I asked that years ago and got no satisfying answer from Bob Hyatt. He just waved hands and said actually we have the Rybka case in our focus

e) Logically would be to assume that the whole little scene of commercial authors go back and through the content of other strong programs without that would be cheating but then the same by Vas isnt cheating too

f) so first of all we must face the somewhat irritating news that commercial chess isnt science with its openess, but this similar to the practice of military projects, to single out someone would be likewise a somewhat outlandish nono

g) so, if you prove to me that Shredder, Junior and Hydra are absolutely sane and innocent thenwe had a different situation, but since that wont happen, Vas is innocent too

i) we just dont have a legal systen that allows private people to insinuate that others must prove that they did something against the law; if you want to do this you must obey to certain practical issues; until now nobody did this

j) you among others want to impress by a certain realism in saying that although Vas is the best that this isnt a big thing; fine, believe into that; if we qould analyse the deepest details of Einstein's theories, they would lose their flair too; the same with all creative talents, it's all hard work and the digesting of former achievements and ideas

k) as you and especially Bob likes to argue: if it smells like a champion, looks like it too, has that outstanding superior decency, it is decency and superor genius, but aside that, we all could be able to do it in the very same manner - not exactly but then in our own genial way - me too (sigh)

l) we have a problem in CC if someone like scientist Hyatt thinks that commercial closedness is evil, but we must decide if we like that sort of ancient enjoyment to look at such a circus, if we all do, then a certain secrecy is part of the game, crying about the lack of openess like it's used in science is just hypocritical

m) it's beyond me why some 5 people plus claque think it's reasonable to single out specific individuals, forgetting about tradition in sports and to hold extended lynch courts instead of just showing some style and taking the irregular bugs as the essence of the whole feature if it enjoys a little community

n) with true chess computerchess is still a bit overasked actually, I hope we agree on that one too, so what is your problem, real life will never be able to compete with the delusions of the internet...

[frcha]

Translating the novel " War and Peace" to spanish does not mean that you wrote war and peace. If you pretend that you wrote it, that would be plagerism. Even if you added a chapter, or improved some sections, you would still be plagerizing war and peace.

This is wrong...At worst Vas translated War and Peace - then hacked off a few chapters and added his own ...

You have no clue how much Vas has taken from fruit -- there is some disagreement among experts though quite a few have said he has violated the gpl.

EVEN IF Vas has violated the GPL -- -1. it would probably never be brought to court and tried successfully 2. Vas created a much stronger engine (in the end) -- something NO ONE else was able to do in the open source realm despite the fruit code being available for use.

A lot of people here do not understand what it takes to create and sell a product successfully .. Sure -- Vas may have been unethical but so is Bill Gates (when he was a capitalist ) -- Look up what Bill Gates thought of Open Source software ..

[Ant_Gugdin]

EVEN IF Vas has violated the GPL -- -1. it would probably never be brought to court and tried successfully

Could you please explain the basis for this claim. How do you know this? I remember other Rybka fans claiming Rybka was a "legal derivative" of Fruit - without even having determined what the applicable law was, let alone having read it.

On the strength of what programmers who have reviewed the R1 executable are saying and on a cursory review of Polish copyright law, it looks like Fabien/ the FSF would have a strong case against Vas in the Polish civil courts. Furthermore, there are several Polish law criminal offences which look to be in point.

[Tom Barrister]

I see the same tired arguments in Mr. Rajlich's defense.

I'm going to sum my views up before moving on.

As far as I'm concerned, Mr. Rajlich used significant portions of Fruit, in violation of the GPL, in the creation of Rybka. I take that as true beyond reasonable doubt.

That being the case, Mr. Rajlich broke the rules/laws/sanctions/whatever of the GPL. The rest of this uses that as a given fact.

Even if any/every other commercial chess programmer also broke the rules, it does not justify Mr. Rajlich's breaking the rules. I haven't seen arguments regarding any commercial engine as containing anything that was stolen unlawfully or in violation of GPL. I can find plenty about Rybka. Even if they exist, it doesn't justify another wrongdoing.

Mr. Rajlich may be the greatest programmer in the universe, but he still broke the rules.

Rybka 4 may have been written from scratch, using programming alogrithms/techniques/routines/whatever that nobody else has dreamed of, and every last byte may be completely original code, but Mr. Rajlich still broke the rules.

It doesn't matter if Mr. Rajlich does or does not suffer a conviction and/or civil judgement from his actions. He still broke the rules.

You can try to justify Mr. Rajlich's behavior by pointing fingers at everybody and everything else in the world, from Osama bin Laden to the hole in the ozone layer, as valid reasons for such behavior, but he still broke the rules.

You can divert the issue all you want with every Spin Doctor tactic known, but Mr. Rajlich still broke the rules.

That said, it's been my experience that when somebody is liberal with rule-breaking (and I would consider wholesale copying of an engine to be liberal rule-breaking), the same is generally true thereafter. As a conjecture, it wouldn't surprise me if the current version of Rybka (yes, Rybka 4) contained bits and pieces of 40 other engines/programs, many of which fall under the GPL umbrella, and perhaps even some from reverse-engineered commercial engines. When somebody has no qualms about taking the work of another, rewriting it for some other routine/technology/whatever, and calling it his own, he isn't likely to have qualms about doing it in the future, either.

Now moving on to the Ippolit series, if their program was created initially from a reverse-engineered Rybka 3, then they broke the rules, as well, since it's possible that a percentage of Rybka 3 actually belongs to Mr. Rajlich free and clear of GPL violations.

If Mr. Houdart used the Ippolit series in the creation of Houdini, and if any part of it was at one time covered by the GPL, then he broke the rules as well, since it's wrong to have a closed-source project that contains GPL-protected code without offering to release the source (correct me if that's wrong). Also, if he knowingly used any commercial code belonging to Mr. Rajlich or anybody else (without permission), then he broke the rules.

For all we know, some/most/all other commercial engines contain GPL-protected code, or perhaps their programmers all reverse-engineered each other's programs and stole from them. The entire chess programming fraternity, or any part of it, may be a pack of thieves, or some/most of them may be innocent. There's certainly enough of a parallel in the rest of the programming world, as is evidenced by the lawsuits the big corporations file against each other. Thievery/plagiarism and theft of intellectual property certainly didn't begin with computer chess programming, or even with any computer programming ever.

Any or all of the above three paragraphs being true or false doesn't change this: Mr. Rajlich broke the rules, and if he took some/most/all of Fruit and rewrote it to fit a new technology/whatever, it's very bald-faced rule-breaking. It can't be justified.

The rule-breaking doesn't tick me off as much as the categorical denying of same, the lame attempts to put new spins on things, point fingers at others, justify the behavior by saying "everybody else does it, so it must be okay", the silly game of dodge-ball that's being played, etc. in an effort to draw attention away from the rule-breaking. That's the reason that the entire post to date has been single-minded.

I have another point to make, but it deserves its own thread.

You may now resume the ping pong match.