Strelka

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 11:14 pm
Location: Munster, Nuremberg, Princeton

Re: S. Fern

Post by Rolf »

bob wrote:
fern wrote:Yesterday I downloaded last version of Strelka, I played it and of course I saw it is a fantastically strong engine.
Now, my question:
Was it probed as a clone of Rybka or not?
I did not follow that discussion entirely when it happened so I am in the open about this.
If it is probed as clone, I will uninstall it.
I do not want to harm Vass.

Fern
Vas certainly claimed it was a clone.
Yes, but he meant something else than what you implied with your few buddies. Let me make this very clear, it can either say, yes I have thoroughly examined the source, compared it with my own and judge it as equal and therefore I say it's mine, or - it can mean and this is what I think it means, I claim that it's practically mine and who ever uses it in our juridical sphere in the Western US World, he gets problems. However what Bob said and others believed it, Vas has stated himself that this is his thing and because the hidden author of this thing has claimed that he took it from FRUIT, therefore, Bob claimed repeatedly, that Vasik had made a big mistake - because this way he had admitted that he took his own stuff from FRUIT.

I think that nobody should say such a nonsense against Vas. Therefore I find the naivety, played, pretended or not of Fern, a bit too outlandish to be taken for serious. But in the end he will admit it himself. I for one find the whole topic a bit indecent, to say the least, but then I am just an observer of the scene that once saw its offspring in science.

Sad to see those days have gone with the wind...
-Popper and Lakatos are good but I'm stuck on Leibowitz
Uri Blass
Posts: 10890
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Strelka

Post by Uri Blass »

GenoM wrote:Strelka is mainly based on Fruit. Osipov claims that in the last version (Strelka 3 R) search of Strelka is combined with Rybka 3 evaluation as a kind of scientific experiment.
Very interesting engine, not anything close to Rybka 1.0 beta (regarding style things).
Just my 2 cents.
I disagree.
I think that strelka is mainly based on rybka.
The move generator is clearly based on rybka and the material imbalance tables are also based on rybka.

The evaluation is also based on rybka's evaluation and it was designed to be similiar to rybka and not similiar to fruit

I think that similiarity to rybka is clearly bigger than the similiarity to fruit(not only in playing style but also in code).

Uri
Uri Blass
Posts: 10890
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Strelka

Post by Uri Blass »

fern wrote:I will wait some more opinions, but for the time being I will stop to use it. Anyway I already have Rybka.

thanks
fern
If you have rybka3 I see no demage that you do by using strelka and Vas cannot complain that you did not buy from him because of strelka.

I also see no demage to Vas by using it assuming that strelka is weaker than the free version of rybka(rybka2.2n2).

In that case people who do not buy rybka3 because of strelka will also not buy rybka3 because of rybka2.2n2 so Vas is going to lose nothing.

Uri
Osipov Jury
Posts: 186
Joined: Mon Jan 21, 2008 2:07 pm
Location: Russia

Re: Strelka

Post by Osipov Jury »

The purpose of the first version of Strelka was to achieve maximum correlation with Rybka.
Today I am doing a new version, trying to achieve two goals:
1. Stronger by more than 50 points compared with Rybka 3.
2. Minimum correlation with Rybka 3.
I think that in June a new version of Strelka will be ready.
User avatar
Graham Banks
Posts: 44607
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Strelka

Post by Graham Banks »

Osipov Jury wrote:The purpose of the first version of Strelka was to achieve maximum correlation with Rybka.
Today I am doing a new version, trying to achieve two goals:
1. Stronger by more than 50 points compared with Rybka 3.
2. Minimum correlation with Rybka 3.
I think that in June a new version of Strelka will be ready.
Will it be entirely original?
gbanksnz at gmail.com
User avatar
Matthias Gemuh
Posts: 3245
Joined: Thu Mar 09, 2006 9:10 am

Re: Strelka

Post by Matthias Gemuh »

Osipov Jury wrote:The purpose of the first version of Strelka was to achieve maximum correlation with Rybka.
Today I am doing a new version, trying to achieve two goals:
1. Stronger by more than 50 points compared with Rybka 3.
2. Minimum correlation with Rybka 3.
I think that in June a new version of Strelka will be ready.

Hi Osipov,

if you give the world a chess engine stronger than Rybka 3, people will stop selling chess engines and some people will hate you.

Matthias.
My engine was quite strong till I added knowledge to it.
http://www.chess.hylogic.de
Osipov Jury
Posts: 186
Joined: Mon Jan 21, 2008 2:07 pm
Location: Russia

Re: Strelka

Post by Osipov Jury »

Graham Banks wrote:Will it be entirely original?
It will not be totally original. Lots of ideas taken from the Rybka and other engines. But the style of play will be most different.
Osipov Jury
Posts: 186
Joined: Mon Jan 21, 2008 2:07 pm
Location: Russia

Re: Strelka

Post by Osipov Jury »

Matthias Gemuh wrote:Hi Osipov,

if you give the world a chess engine stronger than Rybka 3, people will stop selling chess engines and some people will hate you.

Matthias.
I hate a lot of people for a long time.
What can I do in this case?
User avatar
Graham Banks
Posts: 44607
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Strelka

Post by Graham Banks »

Osipov Jury wrote:
Graham Banks wrote:Will it be entirely original?
It will not be totally original. Lots of ideas taken from the Rybka and other engines. But the style of play will be most different.
Will it be open source (or will the source be able to be examined by somebody if not)?
gbanksnz at gmail.com
Uri Blass
Posts: 10890
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Strelka

Post by Uri Blass »

Matthias Gemuh wrote:
Osipov Jury wrote:The purpose of the first version of Strelka was to achieve maximum correlation with Rybka.
Today I am doing a new version, trying to achieve two goals:
1. Stronger by more than 50 points compared with Rybka 3.
2. Minimum correlation with Rybka 3.
I think that in June a new version of Strelka will be ready.

Hi Osipov,

if you give the world a chess engine stronger than Rybka 3, people will stop selling chess engines and some people will hate you.

Matthias.
I think that it is not correct that people will stop selling chess engines and one of the reasons is that I expect rybka of June of this year to be stronger than strelka of June of this year and it is possible that Vas is going to release Rybka4 in case that Osipov release strelka that is stronger than rybka3.

Uri