Can I read here that you have looked at the posted evidence and that your opinion is that large parts of the code were re-used?
I (briefly) looked at the evidence on page 1 of this thread and the posts from Zach in another thread. I didn't look myself into the sources of Strelka or Fruit.
If you didn't attack Uri about not being neutral about this case because of using pgn code from crafty, then forget about that passage in my previous post.
It's not an attack. Yes, to sum it up I told him that he appeared to contradict the process of exposing evidence just because he believes he may have infringed on some copyright himself. I don't even know if it is the case. I even believe that if he did, then this discussion and what is going to come out of it is very interesting information for him.
tiger wrote:
OK, then as Fruit 2.1 is GPL and as far as we know there has been no special autorization given by Fabien to anyone, I think we are asking legitimate questions.
Of course. I just wanted to make a statement that it is not the same if somebody uses some open source code with the permission of the author with his own license (aka Crafty) or if somebody uses some code form a GPL project (aka Fruit2.1).
I consider some lines of code used with permission of the author not per se as bad, it depends on the circumstances.
Therefore I think your attack against Uri was not fair in respect of the pgn code from crafty. However I don't know how much fruit code is in Movei, if that was the code Movei had to be under GPL.
(By the was, while reading this discussion I realized that I used the same random numbers as fruit, which are GPL, maybe I have to change that should my engine ever leave my harddisk.)
And if Fabien has given or give a special authorization at any time in the future, I hope the same authorization will be given to everybody, not just one commercial entity.
Well, this is dubious. I'm not enough of an expert. Can the same code be GPL and commercial? I mean we are taking about a considerable number of lines of code, not just a function. Can the source of a GPL project used as a starting point to a commercial product? I don't know the answer to this.
// Christophe
I am going to say only that I do not use code from fruit.
I use some ideas from fruit in the evaluation like average between opening evaluation and endgame evaluation based on stage of the game
but not copy and paste and I simply understood the evaluation of fruit in english words and translated modified version of what I understood to code.
Uri
It's OK then. You must have noticed that at no point the re-use of ideas has been questionned.
The verbatim re-use of GPL code that you modify and use as if it was your own has been questionned because it is a clear infringement of the licence under which the original code has been published.
If it is not a mistake to use open source and then claim it as his own code, which has maybe happened and which is the point of posting the evidence, then there is a lesson to learn for all chess programmers: everybody is allowed to do it.
That makes a real difference and creates a precedent and sets new rules.
At least we will know that there is absolutely no risk when you take code that is not yours, even if it is protected, and we will be able to expand the scope of our work by picking parts where we want.
So maybe a significant advance in computer chess is going to come out of this.
Well, but this not realistic. Even if Fabien doesn't take any legal action against Vas, he may do so against the next person that uses Fruit 2.1 as a starting point and breaks the GPL. And there is nothing we can do about it, because he is the copyright owner and the person to decide whether to take legal action or not.
Doesn't help at court much to say: Yes, but this other guy broke the license before me.
I do not plan to continue movei as it is and I did not work on movei in the last half year.
I may write a new engine but at this point of time I still did not start to do it.
When you were only one of the three people who had the Strelka code, you wrote several times that you were planning on taking the ideas from Strelka yourself and use it in Movei.
Does that (and this discussion) have anything to do with your decision to start a new engine now ?
If it is not a mistake to use open source and then claim it as his own code, which has maybe happened and which is the point of posting the evidence, then there is a lesson to learn for all chess programmers: everybody is allowed to do it.
That makes a real difference and creates a precedent and sets new rules.
At least we will know that there is absolutely no risk when you take code that is not yours, even if it is protected, and we will be able to expand the scope of our work by picking parts where we want.
So maybe a significant advance in computer chess is going to come out of this.
Well, but this not realistic. Even if Fabien doesn't take any legal action against Vas, he may do so against the next person that uses Fruit 2.1 as a starting point and breaks the GPL. And there is nothing we can do about it, because he is the copyright owner and the person to decide whether to take legal action or not.
Doesn't help at court much to say: Yes, but this other guy broke the license before me.
For this you must assume that Fabien would be unethical and would favor one person over another one.
In any case I still find it very interesting that the truth about possible illegal source code re-use is exposed, then denied or confirmed.
I do not plan to continue movei as it is and I did not work on movei in the last half year.
I may write a new engine but at this point of time I still did not start to do it.
When you were only one of the three people who had the Strelka code, you wrote several times that you were planning on taking the ideas from Strelka yourself and use it in Movei.
Does that (and this discussion) have anything to do with your decision to start a new engine now ?
Tony
I think to write a bitboard engine and strelka code was productive for me to learn about bitboards.
If I start then the first thing to do is writing a new move generator based on bitboards.
tiger wrote:
For this you must assume that Fabien would be unethical and would favor one person over another one.
In any case I still find it very interesting that the truth about possible illegal source code re-use is exposed, then denied or confirmed.
// Christophe
I too find it very interesting to find the truth about Strelka.
If Strelka is derived from Fruit, would it be legal to make a open source chess engine project (published under the GPL) using the Strelka code as a starting point?
Would be a possibility to catch the fish faster...
tiger wrote:
For this you must assume that Fabien would be unethical and would favor one person over another one.
In any case I still find it very interesting that the truth about possible illegal source code re-use is exposed, then denied or confirmed.
// Christophe
I too find it very interesting to find the truth about Strelka.
If Strelka is derived from Fruit, would it be legal to make a open source chess engine project (published under the GPL) using the Strelka code as a starting point?
Would be a possibility to catch the fish faster...
At first glance I would say it would be of dubious legal status, so it would not be safe unless a court enforced the GPL on Rybka 1.0.
However using the ideas in Strelka is not illegal, so the legal way to achieve what you suggest is to start over from Fruit and then include the ideas found in Strelka by modifying Fruit and by not copying code from Strelka.
The best way to achieve this in the open is to have one guy reading the Strelka source code and describing the ideas he sees in plain english exclusively by email to a second guy who is in charge of modifying Fruit in order to implement the ideas. I think it's called the "clean room" approach. By publishing the resulting source code (even better the complete revision history of the source) and the emails that have been exchanged you are able to provide enough evidence to an expert that no copy/paste of possibly "dirty" code has happened. The method will be extremely difficult to challenge in court.
The resulting code must naturally be published under the GPL in order to be legal.
tiger wrote:
For this you must assume that Fabien would be unethical and would favor one person over another one.
In any case I still find it very interesting that the truth about possible illegal source code re-use is exposed, then denied or confirmed.
// Christophe
I too find it very interesting to find the truth about Strelka.
If Strelka is derived from Fruit, would it be legal to make a open source chess engine project (published under the GPL) using the Strelka code as a starting point?
Would be a possibility to catch the fish faster...
Hi Andreas-
The problem at this moment is that the Strelka source is being distributed without documentation of any kind. There's no readme, no mention of authorship (except the UCI author string = Jury Osipov). other than that nothing is attributed to anybody, and there's no GPL. Jury wants it. Vas claims it.
My advice would be: certainly don't start anything until the smoke clears and you see a distribution with GPL.
The problem at this moment is that the Strelka source is being distributed without documentation of any kind. There's no readme, no mention of authorship (except the UCI author string = Jury Osipov). other than that nothing is attributed to anybody, and there's no GPL. Jury wants it. Vas claims it.
My advice would be: certainly don't start anything until the smoke clears and you see a distribution with GPL.
regards-
Norm
Well, it was a theoretical question. I would never find the motivation to do this, because I don't find it interesting enough.
I started a chess engine form scratch some time ago. I like to find out things myself, and I don't care if my engine will ever play strong. It would never be as satisfying for me to implement stuff that somebody else coded. And by the way, I'm not very good a reading code of others, I have already a hard time to read my own code after some weeks.
But maybe somebody else will do it.