![[us]](/mwf/v12/flags/us.png "United States")
![[de]](/mwf/v12/flags/de.png "Germany")
You forgot to mention the more interesting stuff on that page: Søren's reply to Levy 
Thanks for posting that link.
Thanks for posting that link.
Brilliant article. Non-literal copying??? I agree that that statement by Levy is nonsensical, an oxymoron.
Regarding history- Levy should have recognized!
If you would understand anything, observe its beginning and its development.
Aristotle
Ah! The " Hyatt Road of Doom" once embarked upon, is a perilous road to travel down, fraught with shallows, undertow, and a pons asinorum to anyone other than the sharpest of a Machiavellian minded thinker.
If you would understand anything, observe its beginning and its development.
Aristotle
Ah! The " Hyatt Road of Doom" once embarked upon, is a perilous road to travel down, fraught with shallows, undertow, and a pons asinorum to anyone other than the sharpest of a Machiavellian minded thinker.
Sorry, Soren's reply is the least interesting thing. He just wrote the article. He did say something interesting...and all too obvious: "I withdrew as Rybka forum moderator almost immediately after the ChessBase article, as there were far to much mud slinging, and moderation became almost impossible and time consuming."
>as there were far to much mud slinging, and moderation became almost impossible and time consuming
When he decided to withdraw, Hyatt was still active. Now that Hyatt is gone things are much better
He'll probably be back...seemed like it got into his blood!
Yes, and it certainly went both ways, he and the pro-Vas crowd. It takes two sides to sling that much mud.
Yes, and it certainly went both ways, he and the pro-Vas crowd. It takes two sides to sling that much mud.
![[cn]](/mwf/v12/flags/cn.png "China")
Hyatt is gone: WTF?
I hear yah! Dude!
It's like my dog when she tangles with a skunk-the odor lingers. You can use a teaspoon of dish detergent and peroxide mix with baking soda- but that just won't cut the odor. Sometimes I panic and think that damned skunk is still around -but as time goes on the stink gets weaker.
It's like my dog when she tangles with a skunk-the odor lingers. You can use a teaspoon of dish detergent and peroxide mix with baking soda- but that just won't cut the odor. Sometimes I panic and think that damned skunk is still around -but as time goes on the stink gets weaker.
Maybe he's gone to a place (Alan might comply ) called http://www.youtube.com/watch?v=BEL3GaEER_U
) called http://www.youtube.com/watch?v=BEL3GaEER_U
![[fr]](/mwf/v12/flags/fr.png "France")
I wouldn't pay Hyatt any mind. He was our useful idiot to help in wrecking the attacker case, but now he is just noise. Fade him. Everybody else does, even his erstwhile colleagues.
Exactly! Without Hyatt's spotlight on the abysmal quality of the evidence, it would have been a lot more difficult to show that the ICGA investigation was a complete sham.
I suspect he may have been muzzled. Last thing Levy wants is the ICGA case to get any worse than it has already become.
This is quite right. I can honestly say that if it wasn't for Bob exposing his own lies here on this forum, I could easily still be part of the "VIG" camp. Of course, I'm sure that others eventually would have gone through the documents just as carefully, with the realization, "wait, that code isn't there, and nothing like it is, either", but this would have taken longer.
In all fairness a case can be made either way. The rules are just a bit too vague in my opinion and should be tightened up. Once tightened up, maybe the ban on Vas can be reassessed.
Still, "derivative" seems to be the correct term for Rybka, because it was derived from Fruit, with good sized portions still present in 2.3.2a - more so than any other derivative from what I read.
When people flat out say things like "copy" they are often just throwing up a straw man argment. The ICGA complaint read something like, We believe as an unauthorized Fruit derivative Rybka's entry into ICGA events has been contrary to the ICGA rules and the rules of fair play and "Each program must be the original work of the entering developers. Programming teams whose code is derived from or including game-playing code written by others must name all other authors, or the source of such code, in the details of their submission form. Programs which are discovered to be close derivatives of others (e.g., by playing nearly all moves the same), may be declared invalid by the Tournament Director after seeking expert advice. For this purpose a listing of all game-related code running on the system must be available on demand to the Tournament Director."
Not defending anything, just pointing out a point that often goes overlooked.
Still, "derivative" seems to be the correct term for Rybka, because it was derived from Fruit, with good sized portions still present in 2.3.2a - more so than any other derivative from what I read.
When people flat out say things like "copy" they are often just throwing up a straw man argment. The ICGA complaint read something like, We believe as an unauthorized Fruit derivative Rybka's entry into ICGA events has been contrary to the ICGA rules and the rules of fair play and "Each program must be the original work of the entering developers. Programming teams whose code is derived from or including game-playing code written by others must name all other authors, or the source of such code, in the details of their submission form. Programs which are discovered to be close derivatives of others (e.g., by playing nearly all moves the same), may be declared invalid by the Tournament Director after seeking expert advice. For this purpose a listing of all game-related code running on the system must be available on demand to the Tournament Director."
Not defending anything, just pointing out a point that often goes overlooked.
>Still, "derivative" seems to be the correct term for Rybka, because it was derived from Fruit
This never was true and it never will be true. Are you now trying to replace Hyatt?
> Still, "derivative" seems to be the correct term for Rybka, because it was derived from Fruit, with good sized portions still present in 2.3.2a - more so than any other derivative from what I read.
The so-called evidence for that has been well and truly refuted. There is no proven case against Rybka.
In all fairness a case can be made either way.
Really? Bob claimed there was page_after_page of copied Fruit code in R1B. Can you please reproduce this for us?
The rules are just a bit too vague in my opinion and should be tightened up. Once tightened up, maybe the ban on Vas can be reassessed.
Do you really think that Vas is interested in playing in future ICGA tournaments? If you can answer this easy question, can you explain why the ban needs to be reassessed?
Still, "derivative" seems to be the correct term for Rybka, because it was derived from Fruit, with good sized portions still present in 2.3.2a - more so than any other derivative from what I read.
After a long investigation by some very bright people, there isn't a shred of evidence to support this thesis. You must have read this from someone who claimed there was page_after_page of copied Fruit code in R1B.
When people flat out say things like "copy" they are often just throwing up a straw man argment.
Why is Bob throwing up straw man arguments?
The ICGA complaint read something like ...
The ICGA has failed to document their claim.
Not defending anything, just pointing out a point that often goes overlooked.
Nothing was overlooked. You are repeating baseless and libelous accusations.
Really? Bob claimed there was page_after_page of copied Fruit code in R1B. Can you please reproduce this for us?
The rules are just a bit too vague in my opinion and should be tightened up. Once tightened up, maybe the ban on Vas can be reassessed.
Do you really think that Vas is interested in playing in future ICGA tournaments? If you can answer this easy question, can you explain why the ban needs to be reassessed?
Still, "derivative" seems to be the correct term for Rybka, because it was derived from Fruit, with good sized portions still present in 2.3.2a - more so than any other derivative from what I read.
After a long investigation by some very bright people, there isn't a shred of evidence to support this thesis. You must have read this from someone who claimed there was page_after_page of copied Fruit code in R1B.
When people flat out say things like "copy" they are often just throwing up a straw man argment.
Why is Bob throwing up straw man arguments?
The ICGA complaint read something like ...
The ICGA has failed to document their claim.
Not defending anything, just pointing out a point that often goes overlooked.
Nothing was overlooked. You are repeating baseless and libelous accusations.
> In all fairness a case can be made either way. The rules are just a bit too vague in my opinion and should be tightened up. Once tightened up, maybe the ban on Vas can be reassessed.
Go back in time to the origin of the conflict, 16 programmers who signed a letter because they thought Vas COPIED Fruit. This case was never about the ICGA rule #2 in the first place. Instead programmers with the Fabien letter used the ICGA as a TOOL to punish Vasik for copying Fruit under less strict rules (originality) than copyright breach.
Nowadays (february 2012) the ongoing debate is hardly about "code-theft" any longer (as even the alleged time control code theft is unraveled by now) and centers around the question what is allowed to borrow from open-sources authors like Fabien who voluntarily put their work on the internet and then file a complaint if they feel someone has borrowed "too much" using the words of ICGA official Mark Lefler.
The stick to beat Vas was the ICGA and if the ICGA did not had mishandled this case the discussion had been closed 7 months ago and I still would be VIG.
![[pl]](/mwf/v12/flags/pl.png "Poland")
"Go back in time to the origin of the conflict, 16 programmers who signed a letter because they thought Vas COPIED Fruit. This case was never about the ICGA rule #2 in the first place."
But Ed, here's that letter you signed: http://chessprogramming.wikispaces.com/Open+letter+to+the+ICGA+about+the+Rybka-Fruit+issue It's full of the Rule 2 terminology of "derived" and "derivative" and includes e.g. "We believe as an unauthorized Fruit derivative Rybka's entry into ICGA events has been contrary to the ICGA rules and the rules of fair play."
But Ed, here's that letter you signed: http://chessprogramming.wikispaces.com/Open+letter+to+the+ICGA+about+the+Rybka-Fruit+issue It's full of the Rule 2 terminology of "derived" and "derivative" and includes e.g. "We believe as an unauthorized Fruit derivative Rybka's entry into ICGA events has been contrary to the ICGA rules and the rules of fair play."
Yeah, I was infected with Caesar virus as well, but found healing.
Modern version of a quote supposedly addressed to Julius Caesar.
[mode=trolling]
"Beware the leader who bangs the drums of war in order to whip the VIG citizenry into a fruitiness fervor, for VIG is indeed a double-edged sword. It both emboldens the blood, just as it narrows the mind. And when the drums of war have reached a fever pitch and the blood boils with hate and the mind has closed, the leader will have no need in seizing the rights of the VIG citizenry. Rather, the VIG citizenry, infused with fear and blinded by fruitiness, will offer up all of their rights unto the leader and gladly so. How do I know? For this is what I have done. And I am the ICGA."
[/mode]
Modern version of a quote supposedly addressed to Julius Caesar.
[mode=trolling]
"Beware the leader who bangs the drums of war in order to whip the VIG citizenry into a fruitiness fervor, for VIG is indeed a double-edged sword. It both emboldens the blood, just as it narrows the mind. And when the drums of war have reached a fever pitch and the blood boils with hate and the mind has closed, the leader will have no need in seizing the rights of the VIG citizenry. Rather, the VIG citizenry, infused with fear and blinded by fruitiness, will offer up all of their rights unto the leader and gladly so. How do I know? For this is what I have done. And I am the ICGA."
[/mode]
actually, what I find astonishing, is how the cc-ers, having created a more or less democratic system via the forums, albeit with aspects of mob rule from time to time, then handed Levy dictatorial power with respect to Vas and Rybka. Even thought eh consensus now is that the penalty was way too harsh, only Levy has the power to change it; and, as the consensus moves and will move more to VII, again only Levy has power to change that. Insane. Why throw away the democracy to the whim of a dictator?
I have to meet the first VIG who agreed with the punishment, all I heard it was out of proportion, even Bob said so. But none of them takes the initiative to write a formal complaint with signatures to Levy and force him to it. So I would say it's not only Levy.
HW has posted about fifty tiles now that Vas should appeal the sentence. I think "they" and that includes Levy are desperate that Vas appeals, then they can reduce the sentence or something and hope the whole thing goes away. But, two points, it is not the sentence (ridiculous though that is) but the conviction that is at fault. Personally, I think Vas is better leaving icga/Hyatt dangling at the end of the rope of their own making - they are the ones who acted stupid, let them work out the way out.
> I think "they" and that includes Levy are desperate that Vas appeals, then they can reduce the sentence or something and hope the whole thing goes away.
Interesting, I hadn't come to that conclusion myself, but certainly possible . Who knows what they might be thinking
Harvey is begging.
But Ed, I'm in no way criticising you for signing the letter, I'm just noting that despite what you said above the letter WAS about the ICGA rules, and used the same vocabulary that's being used now - as did Vasik's comments about Strelka, as did Yury Osipov's response, as did Letouzey's letter about Strelka/Fruit/Rybka. It's never been a question of verbatim copying but of what's acceptable to take from another program when entering your program in ICGA events - and that's inevitably a somewhat grey area. You're trying to pretend there's been a huge shift in the ICGA position, but simply looking at the documents that's not visible.
> You're trying to pretend there's been a huge shift in the ICGA position,
The ICGA was not the topic, the role of the programmers is.
I remember the "sentiments" at the time of signing the Fabien letter, "let's get Vas", and at that time was OK with that. You are wrong about verbatim copying, the used term "We believe Rybka is a Fruit derivative albeit an advanced one" has only one meaning, take an open source as a base, make changes and call it your own. In other words, copyright breach. Ask Hyatt, ask Fabien, he took the case to the FSF and not because of rule #2. This case was always about copying and so was the sentiment.
And the programmers found ICGA rule #2 was the stick to beat Vas for copying Fruit.
If the programmers just believed Vas only took ideas from Fruit there never would have been a case because they are doing it themselves. So now that the debate is no longer about "copying" I direct my poisoned arrows (for a change) not at the ICGA but at my fellow VIG peers with the question if they anno 2012 still would file the same complaint to the ICGA.
You know the reason I alone never accepted the charge of copying, was because of the Hyatt (and others) claim that Vas took Fruit via cut and paste and transformed it line by line.
First, Vas was obviously an accomplished chess programmer and very bright guy, where else did the +400 come from?
Second, the individual elements of a chess program are perfectly easy to create, not perhaps for a beginner, but certainly easy and straightforward for a young expert.
Third, to laboriously alter, line by line, somebody else's program, converting from one data structure to another radically different one, would be programmer hell in my opinion. Bugs and problems everywhere, with massive traces left behind by force in ordering and design. Really hard work, fraught with difficulties and also really stupid from a legal/commercial point of view, and huge amounts of problematic effort compared to writing it out oneself.
So, I figured the attacker case of full-on copying and translation that Hyatt was alleging, was going to be crud, Vas certainly would have written the sub components by himself with his own ideas, and the subsequent failure to show equivalence at code level bears this out. The only question would be (a) did he use the Fruit weights with a multiplier, and (b) did he put the sub components together according to a Fruity recipe?
We now know (a) was not true, the weights are all over the place and bear no inter-relationship to each other and (b) is still waiting to be argued out. I'm confident Vas is going to be exonerated on (b) also, but the "discussion" awaits and the attackers are not here arguing anymore.
First, Vas was obviously an accomplished chess programmer and very bright guy, where else did the +400 come from?
Second, the individual elements of a chess program are perfectly easy to create, not perhaps for a beginner, but certainly easy and straightforward for a young expert.
Third, to laboriously alter, line by line, somebody else's program, converting from one data structure to another radically different one, would be programmer hell in my opinion. Bugs and problems everywhere, with massive traces left behind by force in ordering and design. Really hard work, fraught with difficulties and also really stupid from a legal/commercial point of view, and huge amounts of problematic effort compared to writing it out oneself.
So, I figured the attacker case of full-on copying and translation that Hyatt was alleging, was going to be crud, Vas certainly would have written the sub components by himself with his own ideas, and the subsequent failure to show equivalence at code level bears this out. The only question would be (a) did he use the Fruit weights with a multiplier, and (b) did he put the sub components together according to a Fruity recipe?
We now know (a) was not true, the weights are all over the place and bear no inter-relationship to each other and (b) is still waiting to be argued out. I'm confident Vas is going to be exonerated on (b) also, but the "discussion" awaits and the attackers are not here arguing anymore.
> Third, to laboriously alter, line by line, somebody else's program, converting from one data structure to another radically different one, would be programmer hell in my opinion. Bugs and problems everywhere, with massive traces left behind by force in ordering and design. Really hard work, fraught with difficulties and also really stupid from a legal/commercial point of view, and huge amounts of problematic effort compared to writing it out oneself.
Exactly. It's hell and you will be caught, you leave traces no matter how good you are. It's a fact that the history of the real clones were an easy snack for investigators.
The "Mec" guy (who is he?) wrote several passioned postings about this issue, pity he is gone.
http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=353891#pid353891
http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=353746#pid353746
![[dk]](/mwf/v12/flags/dk.png "Denmark")
Ed,
Why not try and PM him, maybe he will respond ?
Why not try and PM him, maybe he will respond ?
So why did all those experienced programmers who KNOW nobody would ever translate another program line by line like Hyatt insisted not speak out? Why when hyatt was lying through his teeth did only a few of us disagree? Why were the sixteen or fifteen or fourteen or whatever silent in the face of hyatts bullshit?
Why did those silent ones vote VIG and then stay silent?
Why did those silent ones vote VIG and then stay silent?
Leaving aside the question of line-by-line, what about if someone took Fruit and, using whatever methods, produced a bit-board evaluation function that perfectly mimicked Fruit - i.e. it would give the same evaluation of any position. Then he went to work on making some relatively small changes to the evaluation function, but bigger changes to search etc.
Would you consider that original evaluation function (with identical output) derived code? Would you consider the changed version of the program to be derived from Fruit? (not rhetorical questions, by the way - I'm really not sure what your answer will be)
Would you consider that original evaluation function (with identical output) derived code? Would you consider the changed version of the program to be derived from Fruit? (not rhetorical questions, by the way - I'm really not sure what your answer will be)
Whenever a programmer takes an open source, makes considerable changes it's still a derivative work. As such the description in the Fabien letter is correct. And that's what the Panel had to proof. And usually that's an easy job. Pre-Rybka's were unmasked quite easily, it took me a couple of hours, Houdini same story, commercial LOOP (second in Amsterdam 2007) is the next target of the ICGA, piece of cake. History is full of quick and easy snacks, Strelka, commercial Naum, Onno etc. etc.
And why is it easy to detect clones? Because they leave traces. The DNA traces of the open-source they started from. However in Rybka they found NONE. Not even after a painful 2 year decompile. They can not proof Rybka started its life as Fruit, the meaning of the word derivative.
Would I agree with you Vasik modelled his EVAL to Fruit's? Not unlikely. But there is nothing against that. As long as you write your own code. And don't buy the statement of the Rybka investigators that "Rybka's evaluation is virtually identical to Fruit's", it's simply not true. There are too many differences, worse they did not name them.
And why is it easy to detect clones? Because they leave traces. The DNA traces of the open-source they started from. However in Rybka they found NONE. Not even after a painful 2 year decompile. They can not proof Rybka started its life as Fruit, the meaning of the word derivative.
Would I agree with you Vasik modelled his EVAL to Fruit's? Not unlikely. But there is nothing against that. As long as you write your own code. And don't buy the statement of the Rybka investigators that "Rybka's evaluation is virtually identical to Fruit's", it's simply not true. There are too many differences, worse they did not name them.
Well, I guess the ICGA would claim they've found plenty of traces. And of course the mailbox/bitboard situation and not having available source code for Rybka always meant things would be more complex. You'd agree, however, that the fact it's not trivial to prove Rybka is a clone doesn't mean it's not derived from Fruit?
On identical - I'm talking about a hypothetical original conversion from Fruit that might have been identical (that would be the obvious way to work - and was the way e.g. Osipov worked). Then of course changes would be made - the ICGA mentioned some of them, you've come up with some more (some heavily disputed), but the relevant question would still be at what point the program would stop being derivative.
On identical - I'm talking about a hypothetical original conversion from Fruit that might have been identical (that would be the obvious way to work - and was the way e.g. Osipov worked). Then of course changes would be made - the ICGA mentioned some of them, you've come up with some more (some heavily disputed), but the relevant question would still be at what point the program would stop being derivative.
> Well, I guess the ICGA would claim they've found plenty of traces.
You are misunderstanding "traces" in the sense I used it. We are not talking about the implementation of common chess knowledge or algorithms but the unique DNA traces every programmer leaves in his work. A couple of examples,
1. My work can be easily traced by the use of multiple of "x" such as "xxxxxxx" being a sentinel that tells me I am at the end of a string table. As such this becomes an unique signature of my work. I have more of such unique programming things and as such my work can be easily identified.
2. Crafty has a similar sentinel, Bob uses 999999 and -999999. When you find such a number in another engine it's not unlikely you are dealing with a derivative.
3. Rybka uses the unique "0x7fffffff" sentinel. It's not in Fruit.
Looking for DNA traces you first have to define what makes an engine unique to the rest of the pack and for Fruit to name a few:
1. The evaluation technique.
2. The Quad function
Both are unique for Fruit. It's the Fabien signature.
And both are not in Rybka. If they would have been then BINGO.
Fair enough - I'm not sure if there are or aren't traces in that sense and I'm not the person to debate it with. What interests me, I guess, is the hypothetical question (the "by any method" question I asked Trotsky) of whether if, let's say, someone starts with the Fruit source code on one monitor and then writes an identical evaluation function on another monitor by recoding all the sub-functions to work with bitboards (but let's say it can be programmed in someone else's "own style", as far as that's possible for code that does exactly the same). Is that evaluation function derivative? From what Trotsky wrote I got the impression he would claim it isn't, while my understanding is that it clearly would be under the ICGA rules, even if the "recoding" might suffice for copyright purposes. And then as I understand it if you then worked on that evaluation function (but didn't come close to completely overhauling it), it would still be derivative - again, in the ICGA not copyright sense. Yes, I agree then proving something like that happened is another issue!
Ridiculous. You say "by any method" and then negate it with "start with the Fruit source" ... "on a monitor" ..... "recoding all the sub-functions".
Vas wrote a chess program. Period. You have no idea how he wrote it. Period. Speculating on the basis of "what is possible" and "what if he did so and so" is just speculation. Either there are significant traces of Fruit in Rybka or there are not. Your minions have shown no source code parallels. The 0.0 case is now debunked a few days ago by Ed. End of story. Rybka is clean.
Vas wrote a chess program. Period. You have no idea how he wrote it. Period. Speculating on the basis of "what is possible" and "what if he did so and so" is just speculation. Either there are significant traces of Fruit in Rybka or there are not. Your minions have shown no source code parallels. The 0.0 case is now debunked a few days ago by Ed. End of story. Rybka is clean.
So you refuse to answer the "by any method" question (and I meant the "by any method"), and when I instead ask a more specific question specifically so you might answer that you complain about it being specific and not allowing any method. Fair enough. I like the idea that I have minions, though
You have not asked a "by any method" question. But I already pointed that out. You're just playing games.
"You have not asked a "by any method" question. But I already pointed that out. You're just playing games."
Well, "using whatever methods", so I suppose literally you're correct...: http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=404631#pid404631
I'm not playing games, just curious. Do you claim that "Rybka is clean" if no verbatim copying of source code can be demonstrated?
Well, "using whatever methods", so I suppose literally you're correct...: http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=404631#pid404631
I'm not playing games, just curious. Do you claim that "Rybka is clean" if no verbatim copying of source code can be demonstrated?
That would be part of the clean requirement. And is probably enough to deal with copyright issues. It gets grey if there is recipe equivalence, but such equivalence needs to have prior filtered well-known sub-functions and re-introduced the abstracted weights. The weights will need to be in range of possible tuning variance and the sub-functions need to be novel and non-unique. Also to be considered is unburden evidence, for example a recipe also with substantial non-equivalent sub-functions moves back to being clean. One thing EVAL_COMP failed to do, for example, is give any benefit for use of non-equivalence and new ideas, it searched for positive match only and ignored the negative.
> if, let's say, someone starts with the Fruit source code on one monitor
I signed that Fabien letter because I believed Vas started with Fruit and thus competitive wise took an unfair advantage, the crux of the matter, see Fabien's letter.
The whole point is that it can't be proven, what was strong evidence one year ago is now debunked. The ICGA was quick to respond on the Chessbase article but they don't dare to touch my technical pages. Months ago I have challenged Bob to write a point-to-point refutation, he accepted, where is it ?
> On identical - I'm talking about a hypothetical original conversion from Fruit that might have been identical (that would be the obvious way to work - and was the way e.g. Osipov worked). Then of course changes would be made - the ICGA mentioned some of them, you've come up with some more (some heavily disputed), but the relevant question would still be at what point the program would stop being derivative.
It's a good question. And hardly debated.
![[gb]](/mwf/v12/flags/gb.png "United Kingdom")
I posted this ages ago, and it still seems to me to be relevant. http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=323151;hl=data%20program#pid323151
I deliberately left out of that post the issue of evaluation table values, but of course they are another point that could and should be debated.
I know for a fact that reverse engineering of video games to get ideas has been rife since the start of that industry, so I assume (but certainly cannot prove) that many chess programmers pre-Vas/Rybka engaged in reverse engineering to take as a minimum ideas. It is widely accepted that Shredder was the first to introduce various techniques, and I don't suppose Stefan Mayer willingly gave away commercially valuable techniques in conversations, (though I cannot prove that he didn't).
So the question of where to draw the line has always seemed to me to be the crux of the issue and rather unclear.
I deliberately left out of that post the issue of evaluation table values, but of course they are another point that could and should be debated.
I know for a fact that reverse engineering of video games to get ideas has been rife since the start of that industry, so I assume (but certainly cannot prove) that many chess programmers pre-Vas/Rybka engaged in reverse engineering to take as a minimum ideas. It is widely accepted that Shredder was the first to introduce various techniques, and I don't suppose Stefan Mayer willingly gave away commercially valuable techniques in conversations, (though I cannot prove that he didn't).
So the question of where to draw the line has always seemed to me to be the crux of the issue and rather unclear.
> So the question of where to draw the line has always seemed to me to be the crux of the issue and rather unclear.
+1
And the interpretation of rule #2 is good for any verdict.
Your question is badly thought out and self-contradictory.
It depends what you mean by "took Fruit", as in took what? The source code and changed/modified it? The executable and tracked all the outputs into a giant lookup table? Wrote down the ideas and re-coded? Because you also said "using whatever methods". Used a clean room to write out a spec and then someone else to code it?
It depends on the "methods". It's hypothetical, complex and requires a lawyer. The simple case we currently deal with is complex enough.
It depends what you mean by "took Fruit", as in took what? The source code and changed/modified it? The executable and tracked all the outputs into a giant lookup table? Wrote down the ideas and re-coded? Because you also said "using whatever methods". Used a clean room to write out a spec and then someone else to code it?
It depends on the "methods". It's hypothetical, complex and requires a lawyer. The simple case we currently deal with is complex enough.
I guess what I mean about verbatim copying is that e.g. Letouzey wrote this in the message about Strelka and Fruit that started the process: http://www.talkchess.com/forum/viewtopic.php?t=37762
"The short answer was "no", it was not a verbatim copy of the source code. All the code had been typed (can't say "designed" though, see below) by an individual. So legally there was no issue that I knew of. It was however a whole re-write (copy with different words if you like, similar to a translation) of the algorithms. Not just an extraction of a couple of ideas as is common, and normal.
That being said, some original changes and ideas were also included in the program. So it was, as has since been stated many times in fora I suppose, a bitboard re-write of Fruit with some personal (or otherwise) ideas."
That seems pretty clear to me - he thinks Strelka isn't a verbatim copy and is probably ok in copyright terms, but it is a copy ("copy with different words") in the sense of being clearly derivative of Fruit. Then the letter you signed about Fruit and Rybka to the ICGA seems to be written in similar terms. Of course the ICGA focuses on its own rules, and yes there are wider implications of having broken the ICGA rules, but I don't see anything wrong with that - and it strikes me that the ICGA and the programmers have been consistent in their public statements (I can't judge what's going on in their heads).
"The short answer was "no", it was not a verbatim copy of the source code. All the code had been typed (can't say "designed" though, see below) by an individual. So legally there was no issue that I knew of. It was however a whole re-write (copy with different words if you like, similar to a translation) of the algorithms. Not just an extraction of a couple of ideas as is common, and normal.
That being said, some original changes and ideas were also included in the program. So it was, as has since been stated many times in fora I suppose, a bitboard re-write of Fruit with some personal (or otherwise) ideas."
That seems pretty clear to me - he thinks Strelka isn't a verbatim copy and is probably ok in copyright terms, but it is a copy ("copy with different words") in the sense of being clearly derivative of Fruit. Then the letter you signed about Fruit and Rybka to the ICGA seems to be written in similar terms. Of course the ICGA focuses on its own rules, and yes there are wider implications of having broken the ICGA rules, but I don't see anything wrong with that - and it strikes me that the ICGA and the programmers have been consistent in their public statements (I can't judge what's going on in their heads).
That being said, some original changes and ideas were also included in the program. So it was, as has since been stated many times in fora I suppose, a bitboard re-write of Fruit with some personal (or otherwise) ideas.
I guess you're talking about Strelka here, but I take major issue with the statement, if it were applied to Rybka:
"So it was .... a bitboard re-write of Fruit with some personal (or otherwise) ideas"
Basically, how do you know, let alone prove it?
Say you have Fruit, then two programs written later, A and B, where A is your "bitboard rewrite of Fruit ideas plus other ideas" and B is "selfwritten with some parallel Fruit ideas plus other ideas".
How do you tell the difference between case A and case B from looking at Fruit source and A and B executables?
It is just a BELIEF is'nt it? That one is rewrite and the other is selfwritten. An asserted belief, no more and no less. And, you could just as easily believe the inverse, B the rewrite and A the selfwrite.
I guess you're talking about Strelka here, but I take major issue with the statement, if it were applied to Rybka:
"So it was .... a bitboard re-write of Fruit with some personal (or otherwise) ideas"
Basically, how do you know, let alone prove it?
Say you have Fruit, then two programs written later, A and B, where A is your "bitboard rewrite of Fruit ideas plus other ideas" and B is "selfwritten with some parallel Fruit ideas plus other ideas".
How do you tell the difference between case A and case B from looking at Fruit source and A and B executables?
It is just a BELIEF is'nt it? That one is rewrite and the other is selfwritten. An asserted belief, no more and no less. And, you could just as easily believe the inverse, B the rewrite and A the selfwrite.
I guess it comes down to probabilities - e.g. even if Marcel van Kervinck now has some problems with the verdict for other reasons he stated at ChessBase: "The most likely reason for the evaluation overlap in my opinion is that VR designed Rybka's evaluation first-most as an as faithful as possible replica of Fruit's".
Of course in the case of Yuri Osipov and Strelka we have his testimony that he did exactly that - rewrote Fruit for bitboards before making changes - and he claimed Vas had followed the same path. You can argue how much faith to put in his comments, but if he's telling the truth he should be the best authority around to judge the issue except for Vas himself.
On what you said about the difficulty of copying mailbox-bitboard - sure, I can believe you on that, but if you're going to undertake that tricky task surely first making the copy as exact as possible would be the easiest and most logical approach. Otherwise you're simultaneously trying to do another tricky task (adding improvements and obfuscating things) before you even know if the whole thing will work and be worth your effort. The copying of Crafty and Vas' comments on what he'd do if he was starting out now (and the speed with which he produced Rybka) suggest he favours the faster approach, even if that makes it more likely to leave significant traces of what you've done. Perhaps he wasn't really thinking of any future cloning issues. Perhaps he thought that legally and morally recoding for a different data structure and making improvements was sufficient change to avoid problems. Or perhaps he thought he'd get away with it.
I agree proving that's what happened is a different matter, though, and one I'm not well-placed to judge. Still, something like that applies to all cases of plagiarism, unless you e.g. film the process happening. You have to go on a balance of probabilities.
Of course in the case of Yuri Osipov and Strelka we have his testimony that he did exactly that - rewrote Fruit for bitboards before making changes - and he claimed Vas had followed the same path. You can argue how much faith to put in his comments, but if he's telling the truth he should be the best authority around to judge the issue except for Vas himself.
On what you said about the difficulty of copying mailbox-bitboard - sure, I can believe you on that, but if you're going to undertake that tricky task surely first making the copy as exact as possible would be the easiest and most logical approach. Otherwise you're simultaneously trying to do another tricky task (adding improvements and obfuscating things) before you even know if the whole thing will work and be worth your effort. The copying of Crafty and Vas' comments on what he'd do if he was starting out now (and the speed with which he produced Rybka) suggest he favours the faster approach, even if that makes it more likely to leave significant traces of what you've done. Perhaps he wasn't really thinking of any future cloning issues. Perhaps he thought that legally and morally recoding for a different data structure and making improvements was sufficient change to avoid problems. Or perhaps he thought he'd get away with it.
I agree proving that's what happened is a different matter, though, and one I'm not well-placed to judge. Still, something like that applies to all cases of plagiarism, unless you e.g. film the process happening. You have to go on a balance of probabilities.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill