Rybka's evaluation has been the subject of much speculation ever since its appearance. Various theories have been put forth about the inner workings of the evaluation, but with the publication of Strelka, it was shown just how wrong everyone was. It is perhaps ironic that Rybka's evaluation is its most similar part to Fruit; it contains, in my opinion, the most damning evidence of all.
Whoever wrote that has absolutely no clue what he is talking about. Fruit and Rybka evaluations are worlds apart. I am talking from a chess evaluation tuning point of view as I have no clue how the inner workings coding wise. But the output of Fruit evaluation when compared to Rybka evaluation is so different, as to make any comparison "chess wise" completely ridiculous. The main reason why Rybka gained so much ELO over all other programs is because it has a totally different static evaluation (especially in unbalanced positions) over any other program.
Whoever wrote that has absolutely no clue what he is talking about.
I wonder whether you looked at the code that accompanies the text within that document before you made such a derogatory statement. If you didn’t, I recommend that you do so now.
M ANSARI wrote:
I am talking from a chess evaluation tuning point of view as I have no clue how the inner workings coding wise.
In that case, perhaps a bit more humility and open-mindedness is called for. Someone has done a lot of work preparing that document.
The degree of arrogance, hubris and intellectual dishonesty displayed by some of the more experienced member of CCC is depressing. Sadly, you are not the only culprit, another one of our “esteemed” members not so long ago passed judgment that Zach’s work was insignificant because it only referred to the UCI parser. It turned out that he hadn’t even bothered to look at Zach’s work.
There is no arrogance or intellectual dishonesty, just my opinion based on EVALUATIONS of Fruit and Rybka 1.0. I remember at the time looking at many different positions and being quite surprised at some of the evaluations of Rybka. Chesswise Rybka's evaluation is world's apart from Fruit. The Rybka 1.0 beta was extremely strong but it also had glaring weaknesses and elementary code missing. I remember then thinking that if this thing was polished up it would really take chess engines by storm, and that is exactly what happened. Note that I am not talking about code such as the UCI parser (which I think is open source by the way) where some are claiming it is copied. I wouldn't know if it is copied if both source codes were in front of me. I will never even try to enter that type of debate as I do not have the technical background to even dare make an opinion. But here it is talk about chess EVALUATION, and by that it means the output by the engine of what it thinks of a position ... and here I feel confident enough to say that Fruit 2.1 and Rybka 1.0 beta have extremely different evaluations, and I think most everyone who has tested both engines would agree.
Anyway even amongst the most experienced chess engine coders, there is absolutely no consensus that Rybka 1.0 beta violated Fruit's GPL. CW disagrees with a lot of what Zach says and even the Rebel author seems to doubt some of the points mentioned as proof that Rybka violated GPL. Even Hyatt who seems to follow in Zach's camp seems to think that if Vas had mentioned that he got some ideas from Fruit he should say so and all would be good. Of course you can argue that Vas DID more than give credit to Fruit author as well as many others in his release notes. But the point is this is all about Rybka 1.0 beta ... man this is EONS ago. Rybka 2.2 and Rybka 2.3.2a and then Rybka 3 and now Rybka 4 is here. I don't understand why nobody seems to be willing to admit that Vas has made tremendous advances in how to make a chess engine strong. Even when R3 was disassembled and its guts spilled out for all to see, nobody wants to give any credit to stuff that he created. I really can't figure it out, every day thousands are using his programs and thousands more are using engines that have gained strength from his work (without his consent) and yet some keep trying to make it seem like Vas had nothing to do with it. Even some of the most experienced guys here, who I am absolutely positive knew that Ippolit came from disassembled R3 code, would argue tooth and nail that it had nothing to do with R3 and it was an original work. Human nature is a very strange thing indeed!
M ANSARI wrote:There is no arrogance or intellectual dishonesty, just my opinion based on EVALUATIONS of Fruit and Rybka 1.0. I remember at the time looking at many different positions and being quite surprised at some of the evaluations of Rybka. Chesswise Rybka's evaluation is world's apart from Fruit. The Rybka 1.0 beta was extremely strong but it also had glaring weaknesses and elementary code missing. I remember then thinking that if this thing was polished up it would really take chess engines by storm, and that is exactly what happened. Note that I am not talking about code such as the UCI parser (which I think is open source by the way) where some are claiming it is copied. I wouldn't know if it is copied if both source codes were in front of me. I will never even try to enter that type of debate as I do not have the technical background to even dare make an opinion. But here it is talk about chess EVALUATION, and by that it means the output by the engine of what it thinks of a position ... and here I feel confident enough to say that Fruit 2.1 and Rybka 1.0 beta have extremely different evaluations, and I think most everyone who has tested both engines would agree.
What you are talking about doesn't have much about EVALUATION. This is a very common mistake by laymen, but that still does not excuse you.
What you are calling EVALUATION is in reality search+evaluation+obfuscated Rybka output.
The real evaluation you would get if you put fixed depth search to -2, disable hash + decode obfuscated Rybka output (in this case rescale score values).
And let me give you a small piece of advice. If you don't have anything smart to say, it's better to keep quiet, even though you feel an immense urge to defend Vas whenever his name is mentioned. Ppl would respect you more...
Sean Evans wrote:
There it is in bold just substitute the word Crafty with Fruit and that is the proof In every lie, there is always some truth.
Cordially,
Sean
The 1st version of Rybka was before Fruit. It played in CCT tournaments although it did not do well.
Was it called Rybka at that time. I think he did take a lot from
Crafty before he came out with Rybka beta which he said he
took a lot from Fruit. I was one of the first to test the
Rybka beta versions that had the big elo jump. At that time he
said he took a lot of ideas from Crafty and Fruit.
Maybe Bob H. helped him on some things with Crafty ideas.
Best,
Gerold.
P.S. IMO at that time i think he gave the impression that
Rybka would be free and not be a comm. engine.
You "think" he did take a lot from Crafty. Do you have any analysis that confirms your assertion? Perhaps, Hyatt could assign some of his students a task to reverse engineer Rybka and determine how much Crafty code was copied into Rybka, if any at all.
Sean Evans wrote:
There it is in bold just substitute the word Crafty with Fruit and that is the proof In every lie, there is always some truth.
Cordially,
Sean
The 1st version of Rybka was before Fruit. It played in CCT tournaments although it did not do well.
Was it called Rybka at that time. I think he did take a lot from
Crafty before he came out with Rybka beta which he said he
took a lot from Fruit. I was one of the first to test the
Rybka beta versions that had the big elo jump. At that time he
said he took a lot of ideas from Crafty and Fruit.
Maybe Bob H. helped him on some things with Crafty ideas.
Best,
Gerold.
P.S. IMO at that time i think he gave the impression that
Rybka would be free and not be a comm. engine.
You "think" he did take a lot from Crafty. Do you have any analysis that confirms your assertion? Perhaps, Hyatt could assign some of his students a task to reverse engineer Rybka and determine how much Crafty code was copied into Rybka, if any at all.
Cordially,
Sean
No i have no date from that time. That was a few years ago.
M ANSARI wrote:Whoever wrote that has absolutely no clue what he is talking about. Fruit and Rybka evaluations are worlds apart. I am talking from a chess evaluation tuning point of view as I have no clue how the inner workings coding wise. But the output of Fruit evaluation when compared to Rybka evaluation is so different, as to make any comparison "chess wise" completely ridiculous. The main reason why Rybka gained so much ELO over all other programs is because it has a totally different static evaluation (especially in unbalanced positions) over any other program.
M ANSARI wrote:Whoever wrote that has absolutely no clue what he is talking about. Fruit and Rybka evaluations are worlds apart. I am talking from a chess evaluation tuning point of view as I have no clue how the inner workings coding wise. But the output of Fruit evaluation when compared to Rybka evaluation is so different, as to make any comparison "chess wise" completely ridiculous. The main reason why Rybka gained so much ELO over all other programs is because it has a totally different static evaluation (especially in unbalanced positions) over any other program.
Hi Zach, nice to see you, seems like the Rybka derivative debate is still alive and well. I am surprised after the release of your report on Rybka, that the WCCC did not demand the Rybka source code before being allowed to play at the WCCC. In addition, there are assertions now that Rybka has Crafty code in it too!
Zach, some questions for you:
1. Should the WCCC test Rybka to determine if it is a derivative?
2. If Rybka is allowed to play in the WCCC, should Houdini, Firebird, Ippo, etcetera be allowed to play in the WCCC?
On a side note, as Crafty is a GNU and cannot be sold commercially, does this leave Rybka open to a law suit from Hyatt
M ANSARI wrote:Whoever wrote that has absolutely no clue what he is talking about. Fruit and Rybka evaluations are worlds apart. I am talking from a chess evaluation tuning point of view as I have no clue how the inner workings coding wise. But the output of Fruit evaluation when compared to Rybka evaluation is so different, as to make any comparison "chess wise" completely ridiculous. The main reason why Rybka gained so much ELO over all other programs is because it has a totally different static evaluation (especially in unbalanced positions) over any other program.
Hi Zach, nice to see you, seems like the Rybka derivative debate is still alive and well. I am surprised after the release of your report on Rybka, that the WCCC did not demand the Rybka source code before being allowed to play at the WCCC. In addition, there are assertions now that Rybka has Crafty code in it too!
Zach, some questions for you:
1. Should the WCCC test Rybka to determine if it is a derivative?
2. If Rybka is allowed to play in the WCCC, should Houdini, Firebird, Ippo, etcetera be allowed to play in the WCCC?
On a side note, as Crafty is a GNU and cannot be sold commercially, does this leave Rybka open to a law suit from Hyatt
Cordially,
Sean
This is just silly now. If Bob or Fabien thought that and were bothered they should have challenged when Rybka 1st entered. They did not so case closed for the ICGA. I am pretty sure that if Bob thought it was a Crafty clone he would have complained.
Last edited by Harvey Williamson on Thu Jul 01, 2010 7:59 pm, edited 1 time in total.
Here is my email to the WCCC I sent some time ago. As of now no response. It seems the WCCC prefers to hide under a rock rather than deal with issues like Rybka the derivative It looks like I will have to personally pull the WCCC out from under their rock, kicking and screaming if necessary
-----Original Message-----
From: Sean Evans
Sent: June 28, 2010 5:22 PM
To: 'tournaments@icga.org'
Subject: WCCC - 2010
Hello WCCC,
Would you please advise me whether you will reverse engineer Rybka 4 to confirm it is an original work? There is a significant amount belief and some evidence in the computer chess community that Rybka 4 is a derivative of the open source computer chess program “Fruit”.
For instance, you can see a straw poll at the Computer Chess Club here:
Sean Evans wrote:Here is my email to the WCCC I sent some time ago. As of now no response. It seems the WCCC prefers to hide under a rock than deal with issues like Rybka the derivative
-----Original Message-----
From: Sean Evans
Sent: June 28, 2010 5:22 PM
To: 'tournaments@icga.org'
Subject: WCCC - 2010
Hello WCCC,
Would you please advise me whether you will reverse engineer Rybka 4 to confirm it is an original work? There is a significant amount belief and some evidence in the computer chess community that Rybka 4 is a derivative of the open source computer chess program “Fruit”.
For instance, you can see a straw poll at the Computer Chess Club here:
I look forward to receiving a response from you as soon as possible.
Thank you,
Sean Evans
LOL
Are you an ICGA member? Maybe you should join and then resubmit? Although I doubt they take notice of polls on Forums so your complaint would need to be a little better constructed.
M ANSARI wrote:There is no arrogance or intellectual dishonesty, just my opinion based on EVALUATIONS of Fruit and Rybka 1.0. I remember at the time looking at many different positions and being quite surprised at some of the evaluations of Rybka. Chesswise Rybka's evaluation is world's apart from Fruit. The Rybka 1.0 beta was extremely strong but it also had glaring weaknesses and elementary code missing. I remember then thinking that if this thing was polished up it would really take chess engines by storm, and that is exactly what happened. Note that I am not talking about code such as the UCI parser (which I think is open source by the way) where some are claiming it is copied. I wouldn't know if it is copied if both source codes were in front of me. I will never even try to enter that type of debate as I do not have the technical background to even dare make an opinion. But here it is talk about chess EVALUATION, and by that it means the output by the engine of what it thinks of a position ... and here I feel confident enough to say that Fruit 2.1 and Rybka 1.0 beta have extremely different evaluations, and I think most everyone who has tested both engines would agree.
What you are talking about doesn't have much about EVALUATION. This is a very common mistake by laymen, but that still does not excuse you.
What you are calling EVALUATION is in reality search+evaluation+obfuscated Rybka output.
The real evaluation you would get if you put fixed depth search to -2, disable hash + decode obfuscated Rybka output (in this case rescale score values).
And let me give you a small piece of advice. If you don't have anything smart to say, it's better to keep quiet, even though you feel an immense urge to defend Vas whenever his name is mentioned. Ppl would respect you more...
I don't know what I am talking about? Maybe you don't know what you are talking about. I don't give a rats ass what you consider what evaluation is, but for me evaluation is STATIC EVALUATION of a position. It has been a while since I tested Fruit or Rybka 1.0, but I can assure you that the evaluations of both programs are completely different. Now if for you evaluation means the SEARCH + EVALUATION SKELETAL CODE + OBFUSCATED RYBKA OUTPUT , then that is your problem. As for me being quiet you little moron, that will be in your dreams. As for respect, ... jeeezz where the hell do you come from. Respect for what and by whom??? Is that why you are here ... respect?? If you want respect, then maybe you can do something original instead of copying and pasting. Now be a good boy and go back to your cloner site where you get the "respect" you deserve.
In every lie, there is always some truth. 
