Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Missing piece of the puzzle
- - By Rebel (****) Date 2015-10-03 08:43
Once upon a time this was a happy forum, the programmer was an active and humorous poster and then stopped posting and only limited himself to the must-answer type of stuff.  Can anyone pin down the exact time line when Vas stopped entertaining the forum members? Did he ever gave a reason for that?

Please PM me if you think your opinion is not appropiate for public consumption.
Parent - By Felix Kling (Gold) [de] Date 2015-10-03 11:25
Well, apparently he noticed that he was spending too much time in the forum and, therefore, less working on Rybka :-)
Parent - - By Labyrinth (*****) [us] Date 2015-10-03 12:12
March of 2011, shortly after Rybka 4.1 release.

I think it had a lot to do with his son being born, just ~8 months prior.
Parent - - By Eelco de Groot (***) Date 2015-10-06 01:57
I don't think Ed really means that, I mean Vas stopped posting frequently earlier, for probably hundreds of good reasons, but is it not more interesting what the latest discussion was. Maybe I'm reading too much into it, but you can pick just maybe five posts from one thread (simply searching for User Vasik Rajlich) and apart from one later acknowledgement of Hans van der Zijden, Jiri and Lucas Cimiotti winning the Leiden 2012 tournament in may, Vas stopped posting after this discussion with Mark Lefler, just read these posts:

Topic What is this case about? By Vasik Rajlich Date 2012-01-12 22:28
Mark,

let me ask you one more question.

David Levy and a bunch of other programmers accused me of copying source code. This is what I assumed was the problem, and what I was prepared to defend myself again.

It seems that you are taking a completely different tack.

Can you write a little bit about this? Is source code copying a dead issue and algorithmic copying now the problem? Or are you stating some sort of minority position?

Vas


By mjlef (**) [us] Date 2012-01-12 23:07
Vas,

Your source code was not available for direct comparison, so the panel reverse engineered your program (much like what was done to create Strelka from Rybka, which you agree is a clone or derivative).  This process allowed us to see how Rybka works and compare it to Fruit.  The investigation was to determine if the Rybka entry violated the rules for the ICGA tournament, especially rule 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."

Notice "derived from".  People keep talking about "code" and ignore the "derived from".  Comparison of the methods used in Rybka showed it to be doing the same things as Fruit, the the uniqueness of the Fruit methods, especially the evaluation function indicates Fruit was copied.

I will assume you agree the following would not be allowed: Take a program and read the evaluation function.  Write down exactly what it does in English, or your language of choice. Pick your favorite programming language.  Write code from the English description to precisely do what the initial program does.  The net result would be a perfect copy of the first program, that would act the same.  Note that the ICGA rules require reporting of derivatives, so even if you made minor changes (say changing the weights of some evaluation functions), it is still a derivative.  The panel concluded you violated the rule of reporting your program as a derivative.  Your program was too close to Fruit 2.1 for the panel members, so they consider it a derivative.  In fact, no one on the panel said it was not a derivative.  All were offered a vote and all that voted agreed it was a derivative.

I know that copying say an idea from a technical paper is allowed, and even encouraged.  That is often the whole point of putting out a technical paper.  But Fruit was not an academic paper. It was owned by its programmer who released it with certain conditions.  It did not say "you are free to use this however you want".  it said "you can use this but anything derived from it requires source code release".  In other words, it would force self documenting of derivatives of it.

In summary, you copied and did not say you copied. With no source code we cannot say for sure you used "copy and paste". Well, copy and paste would hardly work on much of it since Rybka was bitboard and Fruit mailbox, so it would have to be translated.  But translating and changing a few values still makes it a derivative.

Lets summarize the arguments people are making in your defense:
"Rybka has these X differences!  It is not the same!" Adding or changing a few things does not mean the basic functionality was not copied.  And since we showed the basic evaluation of Fruit is in Rybka, then we know it is a derivative. By analogy, repainting a stolen car or changing its tires does not make it yours.

"Vas lost Rybka source, so you can never prove the code was copied". Well, we did prove it was copied since the evaluation did so many of the same things in the same way. It does not matter if you converted it to C by way of Pascal, Cobol and Fortan.  It acts the same way and that is copying.

What is this case about? By Vasik Rajlich Date 2012-01-12 23:31
Mark,

ok. Thanks for the detailed and interesting statement. I don't agree with many of your points. We can return to those later. Let's first finish with the source code copying issue.

Can we say that the accusation that Rybka includes copied Fruit source code is dead?

Vas


By  mjlef (**) [us] Date 2012-01-13 00:48
No, I cannot say that at all.  The evidence is the source code was copied and translated into another form.  It is a copy, but it could not be a literal copy and paste of all similar sections due to the different board representations.  It was a translation to another board representation, and this is well shown in the commented assembly language code.  Rybkas 1.0 through 2.3.2a had much code derived from Fruit.

By Vasik Rajlich (Silver) [pl] Date 2012-01-13 07:52
Mark,

so your definition of "copying source code" is looking through source code and implementing the ideas in one's own framework?

That's a pretty unusual definition, IMHO. And it's not at all what I mean when I say that I didn't copy source code.

Vas


And then one final exchange here, after Graham quoted Fabien:

By mjlef (**) [us] Date 2012-01-13 03:52
I consider this to be normal for Fabien, who is quite humble.  If he really felt it was not original work, he would not have released it with the GNU GENERAL PUBLIC LICENSE.  If it was not original, why require anyone to publish changes to it? The idea he refers to is crediting those that came before us. 

Using an idea like "knights belong in the middle" is just fine.  Copying the specific way of measuring that is now. If it was a matter of general ideas, then the decision would probably have been different.  But the panel found a large amount of copying from a single source.  You can look at the Watkins document, which compares Fruit with a bunch of other programs.  It is not much alike any program except the two versions of Rybka used in the comparison. By our measure of originality, Fruit is quite original.  Other than Rybka, Fruit's other program eval overlap was 34%.  The Fruit - Rybka 1 overlap was over 74%.  So we found that nearly 3/4th of the eval was very similar.  For comparison, the mean was about 31%.

> Other than Rybka, Fruit's other program eval overlap was 34%.  The Fruit - Rybka 1 overlap was over 74%.  So we found that nearly 3/4th of the eval was very similar.  For comparison, the mean was about 31%.

Where did you come up with these numbers? This doesn't sound right at all (to the extent that you can even quantify it).

Vas


Well, just two responses more, from just before. I think it is all here in this thread, it sums it all up and you don't need to read anything else, about what Mark Lefler thinks constituted copying, Vas thinks is not copying but implementing common knowledge in his own program like any good programmer would do it when starting a chessprogram of his own, and not having any qualms whatsoever about it and compared to such a thing as Mark's health, this to Vas is totally a non issue, and he is willing to explain this [Note from Eelco: I am assuming here that Vas still more or less feels this way and would still be willing to answer questions, as long as he feels it is being listened to without prejudice, and not a totally closed case of course]

(Mark Lefler:) > I am having some health issues and just returned from a CT scan

(Vas:) Please get healthy, that's a lot more important!

> when Vas himself refuses to offer a defense


When it doesn't contradict the above, please feel free to ask me anything. I don't know if that constitutes a "defense" on my part, but I hope it will be helpful.

Vas


A Big Thanks and a Small Update By Vasik Rajlich Date 2012-01-12 17:38
(Mark Lefler:) > Vas, where did the ratios/formulas for the Rybka 1.0 PST arrays come from?

(Vas:) I don't remember, but they very well may have come from Fruit. I would have no qualms about using them.

Vas
Parent - By Dr.X (Gold) Date 2015-10-06 06:07 Edited 2015-10-06 16:29
It is  poignantly made clear in his absence!

The answers that Vas  was giving were not what  Mark Lefler, Dr.Hyatt, or anyone of clone minded personages from CCC wanted to hear or were willing entertain as truth.

The questions would only have  gotten tediously repetitious. Or worse, they would be demanding that Vas say exactly what it was that they wanted to hear.He'd be given a script. 

Like I indicated above.

Absolutely, no different than what it is like  now in his absence.

[edit]

Ponder the repetitious postings of Dr. Hyatt on the subject of Vas. With only his desired interpretation of asserting his idea of the  facts where there are no clear cut facts but only  interpretations of asserted assumptions. This kind of thing will go on forever with no resolution, and he knows it.
Parent - - By Rebel (****) Date 2015-10-06 10:06
I don't think Ed really means that, I mean Vas stopped posting frequently earlier, for probably hundreds of good reasons,

Indeed.

As I wasn't around in that period, only peeking in the fora so now and then, my feeling is that Vas got tired of the accusations and gradually lost the fun of posting in sync with the swelling of the accusations. Starting point of accusations is August 2008 right after the R3 release when Zach joined Theron and things went fast after that. I tend to believe there is a relationship.

The 2012 discussion (although interesting) has little to do with it, Vas already left the building. He did it because the VIG camp saw it as a sign of guilt for his alleged non cooperation with the ICGA in order to take away that argument. I say alleged because the Vas-Levy correspondence reads differently confirmed by the FIDE who criticized the ICGA harshly:

In addition, Mr Rajlich was not informed about the existence of real disciplinary proceedings against him nor about the risk to be sanctioned this way. He was informed, and in a very informal way, only about the proceedings concerning “Tournament rule 2” and it is clear –from the exchange of messages between him and Mr Levy- that he was not fully acquainted with the possible multifaceted nature of the proceedings and his right to be heard. One person who is informed that his behaviour could be qualified as a cheating and sanctioned with a lifetime ban is likely to assume different decisions about his defence, in comparison to a person who just know about the risk to be disqualified from a tournament.
Parent - By Dr.X (Gold) Date 2015-10-06 17:36
Following up to the ensuing release of Fritz 15, by way of a suggestion  would be to have the standing,  "Complaint to FIDE Ethics Commission thread"remain,  all other threads deleted. Or just sent to the edge!

Attempts at starting new threads that start resuming the ICGA flame war be removed to the edge,  which hasn't been used in years.

Most of the arguments are so similar in nature that they can even be branched into one on going argument.

At least that is one suggestion in an attempt to clean up this forum.
Parent - - By onursurme (***) [tr] Date 2015-10-09 13:26
I can't stop checking this forum , with the hope of the old good days coming back.
Can someone please delete this whole forum to stop the pain?
Parent - - By turbojuice1122 (Gold) [us] Date 2015-10-09 19:44
Let's see what happens later this month when (if?) Fritz 15 is released...
Parent - - By Felix Kling (Gold) [de] Date 2015-11-08 16:13
It would be a bad sign if it would be released this month. It wouldn't be Vas' engine if there is no slip in the schedule ;-)
Parent - - By turbojuice1122 (Gold) [us] Date 2015-11-11 21:22
I guess the next question is if there will ever be a Fritz 15. :twisted:
Parent - By Felix Kling (Gold) [de] Date 2015-11-19 11:42
looks like a release on friday:

http://de.chessbase.com/post/fritz-15-startet-in-magdeburg

at least that's what they write there :-)
Parent - By Dr.X (Gold) Date 2015-11-08 23:56
Just substitute Vas for Gary Cooper walking in the door and break into ...

Up Topic Rybka Support & Discussion / Rybka Discussion / Missing piece of the puzzle

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill