GS wrote:Well, _no one_ has it yet compared to the released version!
Huh? Dann has done so in this very thread. The scores, the PV and the node counts were identical for the released version and the binary Dann compiled himself.
It was much time to clean it up meanwhile to let it look
less cloney, even if I don't know why someone should do this,
because the 'author' himself claimed he has cloned parts of Fruit
and Rybka...
The source code I have is certainly far from a clone of Fruit or any other program I know (although it resembles Fruit somewhat, in a similar way that Glaurung 1.x resembles Phalanx). Whatever Osipov has written in a message translated from Russian and taken out of context does not change this fact.
The whole new discussion seems fruitless to me,
To me, too. Every single person who has seen the source code agrees that Strelka is not a clone of any program they know. When this is not enough to convince people, I'm afraid nothing will.
Tord
Tord, this was my point from the beginning. Whe you have 5 Hall of Fame
computer chess experts- you, Hoffman, Dan, Uri and Alessandro, all claiming they have checked it and it is not a clone- what in God's name do they keep this up for. Im really sick and tired of all this. And thanks for your input.
Best to you always,
Note only that I did not claim that it is not a clone but only that it is not a fruit clone.
I did not claim that it is legal but only that if it is illegal then it is illegal because of rybka and not because of fruit.
I believe that the programmer do reversed engineering to rybka and took parts from it and I did not express an opinion if it is legal.
It seems clear to me that the target of the programmer was not to make a stronger engine but to make something that is similiar to rybka and the programmer probably could easily make it 100 elo better but was not interested in it.
matejst wrote:It seems that when somebody wants to obtain the code of a new interesting engine, he's got only to claim that it is a clone. If somebody else got the code and claims it is not a clone, you continue asserting that he wants to use that code in his own engine, and he's, of course, wrong or dishonnest. If you can't get the code, you don't stop till the engine becomes open source.
So...
I am quite certain that Shredder, Rybka and Hiarcs are clones. They have the same PV when mating in three.javascript:emoticon(':D') They also have the same PV when mating in two. javascript:emoticon(':D')
Of course we never had any clones. I stay with Tony Werten now, who
can follow me. BTW I am sure Vas hasn't seen that source yet?
Guenther
Hi Guenther,
I do not remember a programmer of a clone engine did send his code to other programmers, do you?
What if even Chris would say now: I did a look at the tables - this is not a clone?
I remember several ;)
(e.g. Fafis, DanChess,ElChinito... and so on)
Note also that e.g. WBEC requested the source of programs in doubt
for being checked by independent experts.
(Of course some refused for obvious reasons and were excluded)
I don't know if he (Leo) still does it or if he resigned to the clone problem,
because people who are sceptical against progams are in minority and
even hated by the crowd of engine 'fans' who collect and use whatever is
available...
We have no clear definition of "clone", and we do not have access to Rybka, so these things are murky. If a "clone" is a program that takes ideas from other programs, that would mean pretty much every chess program out there is a clone. A "clone" label should probably be limited to something using actual code from other programs. I doubt that we could ever easily prove a program uses a huge number of ideas of another program. Maybe we need a new term. "Highly plagerized" maybe?
Since many people have seen the source code of Strelka now, and all seem to agree it might be a little similar to Fruit, but vastly different, I think we can safely say it is not a clone. How much of it is original work is what interests me the most. And if the author is willing to share his ideas.
Vas says Rybka code is on one machine and it is well controlled. So this should mean none of us has seen it. Without seeing the Rybka code and the ideas it has, not only can we not say if another program is a clone of it, but we cannot even say how much of it is alike. SImilar PVs and score are interesting and suggest similarities, but are no real proof.
Copyright law protect actual words. Patent law does protect some "processes". Using ideas from a disassenbled program is very merky...you cannot patent an idea or information. This was all well settled by the courts in reverse phone directories (created by companies who scanned in a phone book, then sorted by phone number). So even data tables in a program are probably not patentable or copyrightable. Unless those tables are text.
If Strelka is mostly just a program copying ideas in another program, it is not so interesting. If it has new ideas, it is. I wish the author would tell us more.
The fact that the code was released openly suggests the author has nothing to hide and proven his honesty. Why should we say that whenever a strong engine comes up, everybody accuses it as a clone? Everyone should be very careful before accusing anybody of this.
GenoM wrote:Guenther
first of all -- I am not russian, I am bulgarian
second -- IMO there was a very fine psichological play in Osipov appearences in KasparovChess forum. Its too complicated to explain it, but its my feeling:
Osipov wanted that ComputerChess Community believed that his Strelka was a clone. But in fact -- Strelka is not a clone. He was joking. He made fools all of CCC-members that selfestimate theirselves as computer-chess experts and who believed Strelka is a clone. That was his game He just laughed at you.
Just my opinion of course
Best regards,
Geno
I guess that there is a simpler explanation. I guess that Yuri cannot speak English at all. I guess he does not know what 'clone' means.
I don't see another logical explanation.
Anyway, Fabian has the sources for Strelka now (with permission) and he is going to examine them. I think it will put to bed any remaining doubt about Strelka being a 'fruit clone' [one way or the other!]
So log as you realise that in no way proves anything with regard to Rybka that would be good.
As to whether it has Rybka in it well......
1. It's already been shown to have this in its disassembly.
2. Osipov admitted it anyway whether you like it or not.
I deal only in fact and logic with regard to this. In fact there is nothing I would like more than for strong original engines to appear even daily. It's rare but sometimes they do.
Anyway as I said before....it's being checked against Rybka here and Rybka only. I don't give a toss about Fruit. Fruit is not what will determine whether this thing is a clone.
It's one thing and one thing only......Rybka.
If checking both 1.0 and 1.8. One may be a clone and the other may be changed so much or even completely rewitten that it is not a clone.
GenoM wrote:Guenther
first of all -- I am not russian, I am bulgarian
second -- IMO there was a very fine psichological play in Osipov appearences in KasparovChess forum. Its too complicated to explain it, but its my feeling:
Osipov wanted that ComputerChess Community believed that his Strelka was a clone. But in fact -- Strelka is not a clone. He was joking. He made fools all of CCC-members that selfestimate theirselves as computer-chess experts and who believed Strelka is a clone. That was his game He just laughed at you.
Just my opinion of course
Best regards,
Geno
I guess that there is a simpler explanation. I guess that Yuri cannot speak English at all. I guess he does not know what 'clone' means.
I don't see another logical explanation.
Anyway, Fabian has the sources for Strelka now (with permission) and he is going to examine them. I think it will put to bed any remaining doubt about Strelka being a 'fruit clone' [one way or the other!]
So log as you realise that in no way proves anything with regard to Rybka that would be good.
As to whether it has Rybka in it well......
1. It's already been shown to have this in its disassembly.
2. Osipov admitted it anyway whether you like it or not.
I deal only in fact and logic with regard to this. In fact there is nothing I would like more than for strong original engines to appear even daily. It's rare but sometimes they do.
Anyway as I said before....it's being checked against Rybka here and Rybka only. I don't give a toss about Fruit. Fruit is not what will determine whether this thing is a clone.
It's one thing and one thing only......Rybka.
If checking both 1.0 and 1.8. One may be a clone and the other may be changed so much or even completely rewitten that it is not a clone.
Christopher
Hi Chris
Everyones definition of a "Clone" differs. What's your definition?
GenoM wrote:Guenther
first of all -- I am not russian, I am bulgarian
second -- IMO there was a very fine psichological play in Osipov appearences in KasparovChess forum. Its too complicated to explain it, but its my feeling:
Osipov wanted that ComputerChess Community believed that his Strelka was a clone. But in fact -- Strelka is not a clone. He was joking. He made fools all of CCC-members that selfestimate theirselves as computer-chess experts and who believed Strelka is a clone. That was his game He just laughed at you.
Just my opinion of course
Best regards,
Geno
I guess that there is a simpler explanation. I guess that Yuri cannot speak English at all. I guess he does not know what 'clone' means.
I don't see another logical explanation.
Anyway, Fabian has the sources for Strelka now (with permission) and he is going to examine them. I think it will put to bed any remaining doubt about Strelka being a 'fruit clone' [one way or the other!]
So log as you realise that in no way proves anything with regard to Rybka that would be good.
As to whether it has Rybka in it well......
1. It's already been shown to have this in its disassembly.
2. Osipov admitted it anyway whether you like it or not.
I deal only in fact and logic with regard to this. In fact there is nothing I would like more than for strong original engines to appear even daily. It's rare but sometimes they do.
Anyway as I said before....it's being checked against Rybka here and Rybka only. I don't give a toss about Fruit. Fruit is not what will determine whether this thing is a clone.
It's one thing and one thing only......Rybka.
If checking both 1.0 and 1.8. One may be a clone and the other may be changed so much or even completely rewitten that it is not a clone.
Christopher
Hi Chris
Everyones definition of a "Clone" differs. What's your definition?
GenoM wrote:Guenther
first of all -- I am not russian, I am bulgarian
second -- IMO there was a very fine psichological play in Osipov appearences in KasparovChess forum. Its too complicated to explain it, but its my feeling:
Osipov wanted that ComputerChess Community believed that his Strelka was a clone. But in fact -- Strelka is not a clone. He was joking. He made fools all of CCC-members that selfestimate theirselves as computer-chess experts and who believed Strelka is a clone. That was his game He just laughed at you.
Just my opinion of course
Best regards,
Geno
I guess that there is a simpler explanation. I guess that Yuri cannot speak English at all. I guess he does not know what 'clone' means.
I don't see another logical explanation.
Anyway, Fabian has the sources for Strelka now (with permission) and he is going to examine them. I think it will put to bed any remaining doubt about Strelka being a 'fruit clone' [one way or the other!]
So log as you realise that in no way proves anything with regard to Rybka that would be good.
As to whether it has Rybka in it well......
1. It's already been shown to have this in its disassembly.
2. Osipov admitted it anyway whether you like it or not.
I deal only in fact and logic with regard to this. In fact there is nothing I would like more than for strong original engines to appear even daily. It's rare but sometimes they do.
Anyway as I said before....it's being checked against Rybka here and Rybka only. I don't give a toss about Fruit. Fruit is not what will determine whether this thing is a clone.
It's one thing and one thing only......Rybka.
If checking both 1.0 and 1.8. One may be a clone and the other may be changed so much or even completely rewitten that it is not a clone.
Christopher
Hi Chris
Everyones definition of a "Clone" differs. What's your definition?
So to me it's search and eval. I believe most programmers might agree with me that these are the heart and soul of a chess program.
Eval is what is of primary concern in the Strelka case.
Regards
Christopher
Chris, you're a great guy and i consider you a friend. But Hoffman, Dann, Uri, Alessandro and Tord have all said Strelka is not a clone. Hoffman and Dann, at least, were including Rybka in this also. So forgive me, but what ever your find is completely irrelevant to me. I shall go with what they say and keep testing it. I think all that can be said has been said- and im satisfied with what those 5 (i would call experts) say. Have a nice day and best regards. PS: I can show you a commercial engine- and its NOT Rybka, that will score worse and look more suspicious in your tests than Strelka. That's right, commercial- we paid for it.
geots wrote:
Chris, you're a great guy and i consider you a friend. But Hoffman, Dann, Uri, Alessandro and Tord have all said Strelka is not a clone. Hoffman and Dann, at least, were including Rybka in this also. So forgive me, but what ever your find is completely irrelevant to me. I shall go with what they say and keep testing it. I think all that can be said has been said- and im satisfied with what those 5 (i would call experts) say. Have a nice day and best regards.
George, only because you wrote it twice meanwhile, I want to tell you that
Alessandro does _not_ have Strelkas' source and he never gave that
impression in that thread.
