Are these persons incompetent ?

[Sylwy]

Hi all !

Strelka-this very good & particular chess engine-was the subject of a big cloning scandal some weeks ago. The source code was examined by:
Sergei Markoff (SmarThink)
Bryan Hoffmann
Dann Corbit
Alexander Schmidt was another person defeating this engine with solid reasons.
The verdict was: STRELKA IS NOT A CLONE.
Are all these persons incompetents ?

If some authors working from years to built weak chess engines are nervous or some supporters of cool war are angry or some Rybka fanatics are excited these are another problems.

Please give us some clear samples not philosophical speculations.If not-please stop this discussion .

Regards,
Sylwy

[F.Huber]

Strelka-this very good & particular chess engine-was the subject of a big cloning scandal some weeks ago. The source code was examined by:
Sergei Markoff (SmarThink)
Bryan Hoffmann
Dann Corbit
Alexander Schmidt was another person defeating this engine with solid reasons.
The verdict was: STRELKA IS NOT A CLONE.
Are all these persons incompetents ?

Certainly not incompetent, but they just were mistaken in that case!
Well, that happens sometimes ...

Please give us some clear samples not philosophical speculations.If not-please stop this discussion .

See here:
http://64.68.157.89/forum/viewtopic.php?p=129282#129282

Enough clear samples?

[Sylwy]

Hi !

Do you know the content of this data block ? The size proves nothing.Maybe is only a standard amount of dates in this kind of engines(clearly Strelka has a similar structure with Rybka ).

All the best,
Sylwy

[Sylwy]

1.To be a clone involves the very first source code of Rybka to be in circulation in some occult spheres.Vasik Rajlik must know better this problem. Why he is in silence ??????

2.Uri Blass says Osipov corrected some bugs in Rybka (Strelka knows about underpromotion).
Great-why not Vas ???
Osipov seems to be a great expert in Rybka's code.Nice !
I think the assertion of Mr.Uri Blass that Strelka is a clone without a solid demonstration ( not only the play with various positions) is under the criminal law !

Sylwy

[Uri Blass]

1.To be a clone involves the very first source code of Rybka to be in circulation in some occult spheres.Vasik Rajlik must know better this problem. Why he is in silence ??????

2.Uri Blass says Osipov corrected some bugs in Rybka (Strelka knows about underpromotion).
Great-why not Vas ???
Osipov seems to be a great expert in Rybka's code.Nice !
I think the assertion of Mr.Uri Blass that Strelka is a clone without a solid demonstration ( not only the play with various positions) is under the criminal law !

Sylwy

No

Osipov is the person who did illegal things.

If you publish a book when you copy some chapters from another book and change other chapters you cannot say that what you did is legal.

http://en.wikipedia.org/wiki/Hacker

"A hacker is a person who illegally breaks into computer and network systems, according to the negative meaning of the term popularized by the mass media."

It is possible that Osipov is the hacker who stole the C code from Vas's computer by this way.

I do not know if this is what happened but it seems to me the most logical explanation to the fact that it seems that Osipov had C code of something that is so similiar to rybka based on Dann Corbit and other experts(the fact that he fixed the underpromotion problem suggests also that he has a C code of Rybka because I doubt if he could do that only based on the exe file).

Uri

[Eelco de Groot]

It is possible that Osipov is the hacker who stole the C code from Vas's computer by this way.

I do not know if this is what happened but it seems to me the most logical explanation to the fact that it seems that Osipov had C code of something that is so similiar to rybka based on Dann Corbit and other experts(the fact that he fixed the underpromotion problem suggests also that he has a C code of Rybka because I doubt if he could do that only based on the exe file).

Uri

Come on Uri. Strelka is to all outward appearances a normal chessprogram, written in C. Or else Dann Corbit and Bryan Hofmann would never have declared it, they have seen its sources, as not a clone. Of course it is possible to fix any underpromotion issues in such a program. And change a Winboard interface to a UCI interface, make it generate its datastructures at start-up in hash so it does not have to reside in the exe as a big block of data.

Osipov knows what this program is doing very well I think. Why do you have to make up a hacker who stole Vas' code? Would anybody who did that subsequently sent this somehow stolen file -and I think Vas guards his sources better than that, surely not on his Internet computer without proper precautions - to be examined by other chess programmers as his own original work? That does not make any sense, don't you agree?

Eelco

[Uri Blass]

It is possible that Osipov is the hacker who stole the C code from Vas's computer by this way.

I do not know if this is what happened but it seems to me the most logical explanation to the fact that it seems that Osipov had C code of something that is so similiar to rybka based on Dann Corbit and other experts(the fact that he fixed the underpromotion problem suggests also that he has a C code of Rybka because I doubt if he could do that only based on the exe file).

Uri

Come on Uri. Strelka is to all outward appearances a normal chessprogram, written in C. Or else Dann Corbit and Bryan Hofmann would never have declared it, they have seen its sources, as not a clone. Of course it is possible to fix any underpromotion issues in such a program. And change a Winboard interface to a UCI interface, make it generate its datastructures at start-up in hash so it does not have to reside in the exe as a big block of data.

Osipov knows what this program is doing very well I think. Why do you have to make up a hacker who stole Vas' code? Would anybody who did that subsequently sent this somehow stolen file -and I think Vas guards his sources better than that, surely not on his Internet computer without proper precautions - to be examined by other chess programmers as his own original work? That does not make any sense, don't you agree?

Eelco

I agree only about the fact that Dann Corbit and Bryan Hofmann saw that Strelka is written in C.

You will not find 2 programs with the same analysis of many different positions when they are not clones so the conclusion is that the only way that he got C code is or by stealing Vas code or by translating the exe code of Rybka to C.

Translating the exe file of rybka to C is very hard to do so my guess is that he simply stole rybka's code.

I doubt if Vasik guard his computer better than it at the time of rybka beta.
After all we have not many cases of people who steal code of other and it is possible that he expected that it is not going to happen.

Uri

[Karlo Bala]

Hi Uri,
I think think that the key is _almost identical evals_. We all know that Vas has large material table in Rybka. It is not hard to find it in code and copy-paste in C program. Also it is not hard to find piece square tables, hash codes, and some other well known aspect of eval. You can not find everything, and this is reason why eval is not same in all positions. But, in simple positions (blocked pawns ending, where for example, only PST values are counted) evals are same. Now, it is harder to explain why in some position search depth are identical (rubka depth + 2 = Strelka depth). First thing come on my mind is that author of Strelka could read Rybka asm and find the logic of search. That is hard task, but for someone who is a real expert for asm it is possible (for example Frans Morsh).
If I would like to clone Rybka, first step would be to find same compiler like Vas has, and Vas told already which compiler he uses. Second step would be to try to write something in C, following logic from Rubka asm and then compare output. You try X times and finally if you are lucky, you get something similar like Rybka.
And one maybe simplest scenario: If you know what is input and output of the search, you could cut complete search and link into C program.

[Alexander Schmidt]

Hi,

Strelka is (still) very different, so it is hard to find a real evidence for similaries.

For me one is unguilty until proven the opposite.

And I didn't say Strelka is no clone, I said the possibility is the same as for other engines

Best,
Alex

[Eelco de Groot]

Hi Uri,

Well I don't agree about the "simply stole". I don't think there is anything simple about that and besides it would be a criminal act. Maybe even in Russia, they must have some laws there.

I wrote some more. But I think I'd better stop speculating now, as that serves not much purpose.

Regards, Eelco