Fabien's open letter to the community

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

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
hgm
Posts: 22581
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: Fabien's open letter to the community

Post by hgm » Tue Jan 25, 2011 1:04 pm

SzG wrote:
Sven Schüle wrote:Even in court, there is such a thing as expert testimony.
Wrong person. The quote is from GCP.
Oops! :oops: Wrongly edited it. Bt I was still in time to fix it, thanks to your warning. 8-)

Gian-Carlo Pascutto
Posts: 1107
Joined: Sat Dec 13, 2008 6:00 pm
Contact:

Re: Fabien's open letter to the community

Post by Gian-Carlo Pascutto » Tue Jan 25, 2011 1:06 pm

hgm wrote:
GCP wrote:Even in court, there is such a thing as expert testimony.
Sure, but the consulted experts are usually not the parties involved in the dispute.
I didn't mean to imply that, though that might not have been clear.

I was saying that for the FSF, to decide whether to go to court or not, I would expect Fabien's opinion to matter. His opinion does not change the facts, but it will affect the actions of the parties.

You're claiming Fabien's opinion is irrelevant. That is true for the facts, but not for the actions!

User avatar
hgm
Posts: 22581
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: Fabien's open letter to the community

Post by hgm » Tue Jan 25, 2011 1:21 pm

Gian-Carlo Pascutto wrote:I was saying that for the FSF, to decide whether to go to court or not, I would expect Fabien's opinion to matter. His opinion does not change the facts, but it will affect the actions of the parties.

You're claiming Fabien's opinion is irrelevant. That is true for the facts, but not for the actions!
We might have a different view on this. I think the FSF would prosecute any violation of the GPL, once someone took the trouble to point it out to them. Like you mentioned, the accusation that something is a copy of something else is a factual statement, easily verifiable by anyone.

Sven
Posts: 3695
Joined: Thu May 15, 2008 7:57 pm
Location: Berlin, Germany
Full name: Sven Schüle
Contact:

Re: Fabien's open letter to the community

Post by Sven » Tue Jan 25, 2011 1:28 pm

Laskos wrote:
Sven Schüle wrote:
Laskos wrote:
Sven Schüle wrote:If B is derived from A, and C is derived from B, then C is also derived from A.

But if B (Strelka) is derived from A (Rybka 1.0) and B is also derived from C (Fruit 2.1) then there is no "is-derived-from" relationship between A and C.

To make it simple, let's use "A --> B" for "B is derived from A".

This is correct:
(A --> B and B --> C) implies (A --> C)


But this is wrong:
(A --> B and C --> B) implies (A --> C)


Therefore your last sentence, if you would have finished it in the way most readers would expect, would be lacking some logical foundation.
That is wrong, I agree:

"But this is wrong:
(A --> B and C --> B) implies (A --> C)
"

Yes. But it has nothing to do with my post.

What I was saying, and Vasik said half of that, Fabien said another half is the correct one:

(A = B and C --> B) implies (C --> A)

I confirmed what was Vasik saying in my post with similarity graph in this thread. In fact this similarity graph probably confirms what Fabien is saying too (Rybka 1.0 is shown closer to Fruit 2.1 than Houdini 1.5 is to Rybka 3).
Your "A = B", intended as "Rybka = Strelka", is clearly wrong. It is a known fact that Strelka was based on both Fruit and Rybka.

http://www.talkchess.com/forum/viewtopi ... 815#347815

Therefore any connection between Fruit and Rybka can't be concluded based on Strelka but needs a _direct_ comparison.

Sven
Vasik claimed Strelka 2.0 sources as his own. My similarity test shows Strelka 1.8 as being identical to Rybka 1.0 within error margins. I don't think the fact that Strelka 1.8 is arguably almost an identical clone of Rybka 1.0 or it is exactly 100% Rybka 1.0 changes something. This is an argument ad absurdum just to defend a preferred soccer team. Take Vasik words that Strelka's code is his as granted.

The real issue is what Fabien is saying, confirms the findings or else, that second half of the of assumptions.

Kai
I gave a link to the statement of the Strelka author where he explained exactly what he did to create Strelka. That tells you that A != B. Yuri Osipov also explained that he did not succeed in creating a compilable codebase by only using disassembled R1 code.

So I have to reject your view of my argument being "an argument ad absurdum just to defend a preferred soccer team".

The same post of Yuri Osipov which I linked to above also contains his statement "[...] I gradually added the parts of code and tables from Rybka. It was easy, because Rybka was made on the same way." I do not say this is wrong, I can only say he cannot know exactly that Rybka was made like this. He thinks so. But he can only know for sure what he did by himself, not what someone else did.

Using Don's similarity test is not valid to prove that two programs are "identical". It was not intended for this purpose, I may assume that you know that. Surely you can show that two programs _behave_ very similar. And in case of Strelka vs. R1 this is no big surprise, since Strelka was explicitly designed for exact that purpose.

My point remains that Strelka can't be used to conclude anything about Fruit-R1 connections.

Sven

Gian-Carlo Pascutto
Posts: 1107
Joined: Sat Dec 13, 2008 6:00 pm
Contact:

Re: Fabien's open letter to the community

Post by Gian-Carlo Pascutto » Tue Jan 25, 2011 1:31 pm

hgm wrote: We might have a different view on this. I think the FSF would prosecute any violation of the GPL, once someone took the trouble to point it out to them.
If the violation is obvious, I would hope so. If the violation is not obvious, I guess they would ask someone knowledged about the issue for an opinion. The original author would fit.

But this is just a hunch about how a third party would behave.
Like you mentioned, the accusation that something is a copy of something else is a factual statement, easily verifiable by anyone.
I'm not so sure about the easily myself, but if you think it's easy to determine, I have to ask:

Is Rybka a derivative work of Fruit in your opinion?
Is Ippolit a derivative work of Rybka in your opinion?

Maybe you've made clear statements about this before. In that case, I might have missed them in all that discussion.
Last edited by Gian-Carlo Pascutto on Tue Jan 25, 2011 1:33 pm, edited 2 times in total.

Robert Flesher
Posts: 1207
Joined: Tue Aug 18, 2009 1:06 am

Re: Fabien's open letter to the community

Post by Robert Flesher » Tue Jan 25, 2011 1:31 pm

hgm wrote:
wims wrote:There is no way of knowing wether or not the code has its origin in a decompile or if its a rewrite. Its possible to write code that looks exactly like a decompile.
Indeed. Like it is possible to write code that looks exactly like the original. But it would be a bit stupid to do either of that, as it would be a way to make sure the original copyrights still apply to it. The mechanism by which you actually produce the copy is irrelevant. What is relevant is how much it looks like the original.
If decompiling produces the exact same output as rewriting it then you need to explain why decompiling is cloning while rewriting is not. Both are clearly cloning in my opinion, both are at least derivatives
By 'output' you mean here 'moves it makes', as opposed to the binary output from the compiler? This has been a well-fought battle in the software industry. If I write an operating system that requires exactly the same input as Unix, and gives us the same output on it, would it be a violation of the AT&T copyrights on it? The answer is an unequivocal 'no'. Hence we have Linux...
HGM, I believe you always present a compelling argument even if I believe on this issue it may be a tad myopic. Of course this is not regarding you obvious great understanding of the code copying/cloning/licenses, but rather the use of which to further complicate an already controversial issue. That being, did Rybka illegally/unethically use a GPL program (Fruit), yes, or no? I guess time will tell, but....

If in the next few days/weeks Fabien states Vas did use Fruit, the proverbial "shit that hit the fan", will be on alot of people's faces. IMHO

User avatar
hgm
Posts: 22581
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: Fabien's open letter to the community

Post by hgm » Tue Jan 25, 2011 1:54 pm

Gian-Carlo Pascutto wrote:I'm not so sure about the easily myself, but if you think it's easy to determine, I have to ask:

Is Rybka a derivative work of Fruit in your opinion?
Is Ippolit a derivative work of Rybka in your opinion?

Maybe you've made clear statements about this before. In that case, I might have missed them in all that discussion.
I cannot possibly know, as I don't have Rybka. When I say 'easy', I mean 'easy' on the scale of engaging in a lawsuit. So I assume that before suing they would be willing to go to the trouble of buying a copy of the involved Rybka version to submit it as evidence, and do the comparison on that. It would look a bit bad, I suppose, if they were making this claim based on an illegal copy of Rybka... I am sure people at the FSF are pretty adept at recognizng code from binaries. It is their business, after all.

Once the case goes to court, things become even simpler. To defend against a credible allegation, the accused part would be forced to produce source code for his defense, and anyone could compare that.

For Strelka vs. Fruit it is indeed very easy for all of us, as they are both released as source code.

Which of course does not mean that I have actually done it, or even that I am prepared to do it. I am not really interested in the answer, as none of the inolved engines holds any appeal to me, and I have more important things to do.

But from what I heard from people that seemed to have made a serious effort to dig up the truth, my current opinions on this are:

* Rybka is a legal derivative of Fruit
* Ippolit is an illegal derivative of some later Rybka

Gian-Carlo Pascutto
Posts: 1107
Joined: Sat Dec 13, 2008 6:00 pm
Contact:

Re: Fabien's open letter to the community

Post by Gian-Carlo Pascutto » Tue Jan 25, 2011 1:59 pm

hgm wrote: So no matter how eager people are for Fabien to say anything else than he did, even to the point where they accuse me of misunderstanding him, the bare fact remains that Fabien states:

Strelka would be a legal derivative of Fruit, even if it had remained closed source.

Now you can disagree with that opinion, and you can even show facts to prove that Fabien is completely wrong about this, or claim that I misunderstood him completely. Or people could start calling me names and could call me a Rybka fanboy just because I point out something they do not like. But none of that alters the fact that this is what Fabian actually wrote.
Putting something in a big font doesn't make it true. I can gather no such thing from the email, Fabien certainly did not make that exact statement, and some statements at the end hint at the opposite.

For example:

Fabien says "Strelka did not look like a problem because I assumed it was free.". Now this is an exact quote, unlike yours. If Strelka is a legal derivative, why would it matter if it was free?

My interpretation would be that Fabien would care less about the copyright being violated if that was not done to make a profit of.

User avatar
hgm
Posts: 22581
Joined: Fri Mar 10, 2006 9:06 am
Location: Amsterdam
Full name: H G Muller
Contact:

Re: Fabien's open letter to the community

Post by hgm » Tue Jan 25, 2011 2:03 pm

Robert Flesher wrote:That being, did Rybka illegally/unethically use a GPL program (Fruit), yes, or no? I guess time will tell, but....
Well, from what Fabien writes now (which is similar to other investigations I saw posted on this subject, so I have no reason to disbelief him), it seems that Strelka wold not be illegal, even as closed source. Going on the assumption that Rybka is basically a closed-source Strelka that would make Rybka legal to.

I cannot say anything about 'ethical'. There are as many different kind of 'ethics' as there are persons. This is a subective issue, and what is ethical to some, will always remain unethical to others.

Gian-Carlo Pascutto
Posts: 1107
Joined: Sat Dec 13, 2008 6:00 pm
Contact:

Re: Fabien's open letter to the community

Post by Gian-Carlo Pascutto » Tue Jan 25, 2011 2:05 pm

hgm wrote: I cannot possibly know, as I don't have Rybka.
But you said:
Like you mentioned, the accusation that something is a copy of something else is a factual statement, easily verifiable by anyone.
If one has to go through court to find out, that isn't easy, and since not everyone can bring this to court, it's hardly verifiable by anyone.

Post Reply