![[us]](/mwf/v12/flags/us.png "United States")
At the present time there is no video contemplated, and honestly, I wouldn't mind a bit if someone else stepped up and put on a whiz-bang production. A half-hour interview can involve several days of frantic work. The only compensation is knowing that over 2000 people downloaded the interview, which I find to be pretty cool. Can you imagine addressing a crowd that large, consisting of such eclectic and way above-average people? I just wish I could have been better at it!
![[mx]](/mwf/v12/flags/mx.png "Mexico")
Since no transcript of the interview exists, can someone please post a summary of the questions asked and Vas's answer to them? It doesn't need to be textual, I'm just dying to really know what the interview was about.
You did a fantastic job, Nelson, with the interview and we do appreciate all of the hard work. It would be great to hear Vas speak on the subject again. If there is anyone that could get another Vas interview, that person is you.
Tell you what, I am videoed out right now. The only person who I might make an exception for is the mysterious Houdart. A lot of people would like to take his measure, I'm sure. I don't know if such a video could be posted here but I'm sure someone else would pick it up. Of course, I suspect it would be difficult to persuade Robert to talk to me. The natural instinct would be to think I was a Rajlich disciple intent on trying to ambush him or something. This is a paranoid hobby!
But before I even think about doing that I need professional equipment, for goodness sake. You have no idea; it was such an ordeal to make it work! This last time there were a three separate times I thought the whole thing was going to be a failure. The first I've already described: the disastrous, hasty first try. The second two situations were both technical failures, one during filming and another in the editing phase. It's enough to drive a man crazy when you have a scoop that nobody else has and time is of the essence! Whole days were lost trying desperate workarounds.
But before I even think about doing that I need professional equipment, for goodness sake. You have no idea; it was such an ordeal to make it work! This last time there were a three separate times I thought the whole thing was going to be a failure. The first I've already described: the disastrous, hasty first try. The second two situations were both technical failures, one during filming and another in the editing phase. It's enough to drive a man crazy when you have a scoop that nobody else has and time is of the essence! Whole days were lost trying desperate workarounds.
![[no]](/mwf/v12/flags/no.png "Norway")
You mean like Nalimov TB code? 
Eugene's code is _anything_ but public domain. To use it requires his express written permission, as well as Andrew's since he wrote the compression/decompression code. This code was first used in Crafty, as that is where Eugene developed it.
there is no public domain code in Crafty or Fruit to "take".
I dont get this nitpicking, since I could read that Vas probably asked Nalimov, it was a communication problem. I am a lay, therefore I see it as if Vas took it from you because it was easier sine he knew your program. But from there it's miles and miles to the conclusion that he stole something. It's all a war about nitpicking word twisting. You are a strange trademark, without any human touch. You appear as the sour loser that was presumed by Jeroen. Sorry, Bob, you just are a terrible splitter. Instead of a warmhearted uniter. Not realizing that they only instrumentalized you for their bad feelings about the better programmer Vas. You will never convince me that Vas became only the best because he took something out of Crafty or Fruit, what all other closed sourcers also could have done without any scapegoating.
You are a terrible formalist as if the standards of your classes would mean a thing in commercial competition. Vas will survive you but then you will have destroyed that part of computerchess.
Vas is back in chess with the online features based on cluster.
I must also confess something evil. If I were Vas I would never again wanting to have something to do with these biased lynch justice fiddlers, I say, even, if say David Levy considers a sort of amnesty. No, I would say go away and remember to never again treat others with this sort of illegal prejudging and character assassinating.
But this is only me. Vas is a better human being than me.
You are a terrible formalist as if the standards of your classes would mean a thing in commercial competition. Vas will survive you but then you will have destroyed that part of computerchess.
Vas is back in chess with the online features based on cluster.
I must also confess something evil. If I were Vas I would never again wanting to have something to do with these biased lynch justice fiddlers, I say, even, if say David Levy considers a sort of amnesty. No, I would say go away and remember to never again treat others with this sort of illegal prejudging and character assassinating.
But this is only me. Vas is a better human being than me.
I think that if any actual code were taken, we would have seen evidence of this by now. First, Vas isn't going to introduce something new that isn't already well-known, and second, Vas isn't going to say anything that hurts himself.
No, you just stop at the PST stuff. Did you go down any further. For just one example, go to the passed pawn section. That is not about comparing PSTs.
If you trust Vas more than me (or anyone else for that matter) here, that is simply your problem, not mine...
If you trust Vas more than me (or anyone else for that matter) here, that is simply your problem, not mine...
Can you tell us precisely where in that section is supposedly copied code that is also exactly as listed in Rybka 1.0 Beta?
Easiest example is to go down to what I _think_ is "page 11". The section on "flags". Then when you have finished that, which is pretty easy to follow, go down to the section on "phases". That's bigger and easier to see, and much easier to realize "this is too much code that is identical, for the two pieces to have been written by two different people."
I think if you search for "flags" and then "phase" you will get down to that pretty easily...
I think if you search for "flags" and then "phase" you will get down to that pretty easily...
Again, Zach says, "the code to set the flags is not there". Thus, we again have an idea, not code, that was taken. It is well-known that a formula cannot be copyrighted.
If this is the "easiest example", then obviously, the ICGA should be talking to its lawyers to try to come up with a defense against a forthcoming slander lawsuit (assuming that Vas actually cares...).
If this is the "easiest example", then obviously, the ICGA should be talking to its lawyers to try to come up with a defense against a forthcoming slander lawsuit (assuming that Vas actually cares...).
I agree. There seems to be a huge emphasis on piece values and tables. That is not code, not copyrightable in my opinion, and perfectly OK to use. And as for Rybka 1.61. or whatever it is, it never entered into any ICGA tournaments, was a private version as I understand, and is totally irrelevant. Vas was just probably playing around to see what works before re-writing.
The whole ICGA thing is a sham, but you're right turbo, Vas just doesn't seem to care, and also the costs of any action would be huge. If I had won Euromillions the other week I would certainly have deposited a very large sum of money with a firm of lawyers, for them to hire independent experts or whatever they felt they would need, at Vas's disposal, and whatever not used given to charity
The whole ICGA thing is a sham, but you're right turbo, Vas just doesn't seem to care, and also the costs of any action would be huge. If I had won Euromillions the other week I would certainly have deposited a very large sum of money with a firm of lawyers, for them to hire independent experts or whatever they felt they would need, at Vas's disposal, and whatever not used given to charity
This is not a "formula". It is an implementation detail.
See the rybka/crafty data dealing with "EvaluateWinners()". How you set the flags is not important. Where you set them is not important. It is the very unique way they are used that makes this a "craftyism"...
If you want to take that single statement "the code to set the flags is not there" and use that to discount the entire flag section, that's ok. Then move down to the "phase" section and see what you can find there that you don't like...
See the rybka/crafty data dealing with "EvaluateWinners()". How you set the flags is not important. Where you set them is not important. It is the very unique way they are used that makes this a "craftyism"...
If you want to take that single statement "the code to set the flags is not there" and use that to discount the entire flag section, that's ok. Then move down to the "phase" section and see what you can find there that you don't like...
We will see what the court says. I have a feeling that legal experts will view it in the same way as a formula.
I guess we will see.
Ed, you're a pretty reasonable guy. You know what you're talking about. Is it really necessary to start threads like this? The only outcome is going to be more people yelling incomprehensibly about a subject that they do not understand.
I will give you another secret: this is a technical document. It's not really designed to be read by an audience that isn't (at the very least) programmers. Perhaps I could try explaining everything in great detail to convey the meaning of each piece of evidence in layman's terms. But most people would still be unable to understand it, and their opinion is going to be shaped by so many things other than evidence that it's really pointless.
I will give you another secret: this is a technical document. It's not really designed to be read by an audience that isn't (at the very least) programmers. Perhaps I could try explaining everything in great detail to convey the meaning of each piece of evidence in layman's terms. But most people would still be unable to understand it, and their opinion is going to be shaped by so many things other than evidence that it's really pointless.
Zach, I understand that this is not a multi-million dollar government project and you invested loads of your spare time into the document for which only praise.
But don't you think that by just showing the PST's from Fruit and Rybka would have been a lot more clear for everybody ?
But don't you think that by just showing the PST's from Fruit and Rybka would have been a lot more clear for everybody ?
![[fr]](/mwf/v12/flags/fr.png "France")
Apologies for jumping in, but ......
But don't you think that by just showing the PST's from Fruit and Rybka would have been a lot more clear for everybody ?
well, if You, Ed were doing this, you would be seeking to find TRUTH, so your suggestion would be more than sensible.
But Zach was trying to prove guilt, which ain't the same thing at all. He NEEDS side by side visual matching to wave around the wiki panel, the icga and then the public forums. Idea is that nobody can argue with two columns of "identical" stuff.
If he puts side by side PST pairs, there's no visual match.
If he puts side by side PST pairs, and next to them their division matrix, trying to show multiplicative matching through the division matrix showing one constant, it will work a bit but nowhere enough, the bulk of the tables would be too different.
So that leaves the apparently most damning side by side possibility consisting of the Fruit source alongside Zachs entirely fantasy source. So that was chosen. Objective: to prove guilt.
And this "proof" worked amongst the esteemed panel and among the icga board and was released (as links) in the public document, presumably also available to the world's press.
It was only when it got to the REAL "court", ie open debate, that the "proof" fell apart, but it had done it's job by then. So, as Zach tried to say: "move along now, nothing to see here".
But don't you think that by just showing the PST's from Fruit and Rybka would have been a lot more clear for everybody ?
well, if You, Ed were doing this, you would be seeking to find TRUTH, so your suggestion would be more than sensible.
But Zach was trying to prove guilt, which ain't the same thing at all. He NEEDS side by side visual matching to wave around the wiki panel, the icga and then the public forums. Idea is that nobody can argue with two columns of "identical" stuff.
If he puts side by side PST pairs, there's no visual match.
If he puts side by side PST pairs, and next to them their division matrix, trying to show multiplicative matching through the division matrix showing one constant, it will work a bit but nowhere enough, the bulk of the tables would be too different.
So that leaves the apparently most damning side by side possibility consisting of the Fruit source alongside Zachs entirely fantasy source. So that was chosen. Objective: to prove guilt.
And this "proof" worked amongst the esteemed panel and among the icga board and was released (as links) in the public document, presumably also available to the world's press.
It was only when it got to the REAL "court", ie open debate, that the "proof" fell apart, but it had done it's job by then. So, as Zach tried to say: "move along now, nothing to see here".
Trotsky, just a hug for the fluency of your nice speech which explains to me what I had wringled up in my mind long before. Thanks so much for your assistance.
I asked this elsewhere, but again here.
Fruit initializes PSTs at run-time. Rybka has them hard-coded with no run-time initialization at all.
How to best show that the values are the same?
(1) run the fruit code, then just dump the PSTs after they are initialized. Now you have (for a single piece type and color) an array of 64 values. The values are not the same because of the scaling where fruit and rybka differ greatly on "what is the numeric value of a single pawn, material only?" The average person is not going to be able to visualize the two PST's and mentally divide one set of values by a large constant and see that they are the same.
(2) determine if the Fruit PST initialization code works to produce Rybka's numbers, by using the same PST initialization code in Fruit, but tuning all the values to adjust for the material difference in the value of a pawn. Do you get the same numbers?
I personally find it much easier to notice that the code to produce Fruit's PSTs will also produce Rybka's PSTs, if you just adjust the values to account for the difference in the pawn material value. Either way would work just as effectively, since you end up with the same set of values either way. But with code, you can compare line-by-line to see the difference (if any) where with two 8x8 arrays, it is a good bit of "eye-work" to compare each element individually and then do the math for the conversion factor.
When a DNA test is done in court, what is generally presented is "this DNA sample came from that test donor with a 1 in 7,500,000,000 chance of an error. They don't explain how the tests are done, how DNA is "cleaved", how the gels work. Average layman wouldn't begin to understand that. So they just present the result. Our report tried to be much more careful, showing _everything_ that was found and how it was found. It was _never_ intended for a non-programmer or a non-chess-programmer. Who would give C code examples to people that could not program at all?
The audience it was directed to, the ICGA board, were all competent programmers, and understood computer chess programming as well. We tried to make things crystal clear to them. The report is less "technical" but refers to some pretty technical stuff in the stuff Mark, Zach and others of us provided. But when you jump into Zach's report, that is not going to be readable by the non-technical person. Nor should it be.
Fruit initializes PSTs at run-time. Rybka has them hard-coded with no run-time initialization at all.
How to best show that the values are the same?
(1) run the fruit code, then just dump the PSTs after they are initialized. Now you have (for a single piece type and color) an array of 64 values. The values are not the same because of the scaling where fruit and rybka differ greatly on "what is the numeric value of a single pawn, material only?" The average person is not going to be able to visualize the two PST's and mentally divide one set of values by a large constant and see that they are the same.
(2) determine if the Fruit PST initialization code works to produce Rybka's numbers, by using the same PST initialization code in Fruit, but tuning all the values to adjust for the material difference in the value of a pawn. Do you get the same numbers?
I personally find it much easier to notice that the code to produce Fruit's PSTs will also produce Rybka's PSTs, if you just adjust the values to account for the difference in the pawn material value. Either way would work just as effectively, since you end up with the same set of values either way. But with code, you can compare line-by-line to see the difference (if any) where with two 8x8 arrays, it is a good bit of "eye-work" to compare each element individually and then do the math for the conversion factor.
When a DNA test is done in court, what is generally presented is "this DNA sample came from that test donor with a 1 in 7,500,000,000 chance of an error. They don't explain how the tests are done, how DNA is "cleaved", how the gels work. Average layman wouldn't begin to understand that. So they just present the result. Our report tried to be much more careful, showing _everything_ that was found and how it was found. It was _never_ intended for a non-programmer or a non-chess-programmer. Who would give C code examples to people that could not program at all?
The audience it was directed to, the ICGA board, were all competent programmers, and understood computer chess programming as well. We tried to make things crystal clear to them. The report is less "technical" but refers to some pretty technical stuff in the stuff Mark, Zach and others of us provided. But when you jump into Zach's report, that is not going to be readable by the non-technical person. Nor should it be.
What amazing tosh!
If Zach's report "should not be readable by non-technical persons" what the hell do you think you were doing running around public forums waving the Zach paper and telling all and sundry to "look at the side by side source code, all identical"?
If Zach's report was "designed to be read by technical persons" why didn't you read it with sufficient engagement such you would not have been misleading everyone with "all six PST tables were direct multiplicative of Fruits", when that statement is untrue in both quantity and quality (as proper reading of Zach's paper would have told you?)
I would suggest that in the atmosphere of group think that existed in that wiki attack group, practically NOBODY read anything with a critical and detached air. If I extrapolate from their leader and sidekick, that is. Your problem was that you defined Vas as guilty before you started, everything you did was designed to fit that and you had filtered and excluded the potential critical thinkers who would have corrected you. Fatal error.
If Zach's report "should not be readable by non-technical persons" what the hell do you think you were doing running around public forums waving the Zach paper and telling all and sundry to "look at the side by side source code, all identical"?
If Zach's report was "designed to be read by technical persons" why didn't you read it with sufficient engagement such you would not have been misleading everyone with "all six PST tables were direct multiplicative of Fruits", when that statement is untrue in both quantity and quality (as proper reading of Zach's paper would have told you?)
I would suggest that in the atmosphere of group think that existed in that wiki attack group, practically NOBODY read anything with a critical and detached air. If I extrapolate from their leader and sidekick, that is. Your problem was that you defined Vas as guilty before you started, everything you did was designed to fit that and you had filtered and excluded the potential critical thinkers who would have corrected you. Fatal error.
![[nl]](/mwf/v12/flags/nl.png "Netherlands")
> If Zach's report "should not be readable by non-technical persons" what the hell do you think you were doing running around public forums waving the Zach paper and telling all and sundry to "look at the side by side source code, all identical"?
+1
I have said it before and say it again: Bob's 'watertight' case and undisputable 'facts' are falling apart.
No wonder he is spending most of his time here, pleading his 'case'.
Literal copying of code has not and can not be proven since the original source code is not available, simple as that.
All the 'evidence' is deducted, while the charge is literal copying. It's a catch 22. It can never be proven without the original source.
The reason I am here is in an (apparently) vain attempt to stamp out ignorance. Either my boots are not large enough, or the ignorance here is overwhelming. Hard to say which.
Your cause is lost and attempts are in vain.
No panel puppets here.
No panel puppets here.
Nope. Just a flock of ostriches that keep their heads buried in the sand so that they don't have to see the evidence...
> The reason I am here is in an (apparently) vain attempt to stamp out ignorance.
And we are here to help you come to your senses- and recognize you own ignorance and closed minded biases-which has no place in science-which you should realize being a classroom prof.
![[gb]](/mwf/v12/flags/gb.png "United Kingdom")
Bob,
I do not believe you are wasting your time; some are listening - the two polls I have seen on this site one I think relating to a question around whether Vas 'plagiarised' Crafty code {I think the result so far is approx. around 50:50} and one in respect to the lifetime ban imposed by the ICGA {and on this one we have had around 200 votes and 4 out of 10 people believe the punishment to be fitting}.
I do not believe you are wasting your time; some are listening - the two polls I have seen on this site one I think relating to a question around whether Vas 'plagiarised' Crafty code {I think the result so far is approx. around 50:50} and one in respect to the lifetime ban imposed by the ICGA {and on this one we have had around 200 votes and 4 out of 10 people believe the punishment to be fitting}.
Look, that woke me up. The 2nd one- 4/10 for lifetime ban. I have no idea, so I will let you choose. You will have a choice of, I believe it is 4 options. But you are entitled to only 1 of them.
1. Someone lied to you and you just passed the lie along.
2. You created the lie, which I would hope is not the right pick.
3. Your figures are a typo.
4. You need to get a sample group of humans to poll, and not use monkeys.
gts
1. Someone lied to you and you just passed the lie along.
2. You created the lie, which I would hope is not the right pick.
3. Your figures are a typo.
4. You need to get a sample group of humans to poll, and not use monkeys.
gts
I really don't understand this response George; if you want me to understand then you will need to try again otherwise, no worries.
> The reason I am here is in an (apparently) vain attempt to stamp out ignorance. Either my boots are not large enough, or the ignorance here is overwhelming. Hard to say which
+1
![[at]](/mwf/v12/flags/at.png "Austria")
>> Literal copying of code has not and can not be proven since the original source code is not available, simple as that.
>> All the 'evidence' is deducted, while the charge is literal copying. It's a catch 22. It can never be proven without the original source.
That's exactly what I'm talking about all the time.
Really, I'd prefer to draw a line under this whole case, because nothing will change ICGA's decision, nor does VR wish to ever compete in ICGA events again, anyways.
VR major problem or major chance of repairing his reputation in the near future could be a major lawsuit (brought by FSF & F. Letouzy) due to GPL-Copyright violation.
If it's true, that VR doesn't have the R1 sources any longer (burn the ships) and the FSF succeeds in proving that VR copied Fruit's source code verbatim when he programmed R1 (six years ago, and using a RE disassembly for the purpose of proving their claim)
VR still couldn't comply to the GPL (publish the source code along with the program), because he doesn't have the source code of R1 any longer.
Hence he can't open R1's source code.
However, considering how powerful FSF's lobby is in some countries, a judge (who's maybe a naive FSF-fan) could nevertheless find VR guilty of having broken Fruit's GPL.
If this should happen, then only due to circumstantial evidence, because disassemblies usually aren't convincing at all, neither for the expert witness and even less so for the layman/judge.
Again, this doesn't mean that the FSF couldn't win such a lawsuit at all (due to insufficient evidence), because it's lobby in jurisdiction is quite influential.
So whatever may happen, whether FSF sues VR due to GPL-violation in R1, or whether they back away from that, it doesn't really matter from VR's perspective.
Maybe he'll need to pay a few bucks to FSF (should he really lose, which isn't very likely, due to FSF having nothing else but circumstantial evidence) in order to get them off his back,
but because he doesn't even have R1's sources any longer he still won't be able to publish them (so bad luck for Prof. Hyatt & Wegner who wanted to study them in order to steal R1 cutting-edge pruning techniques for use in their own programs).
So it seems that - if FSF was really going to sue VR (can the FSF do anything else aside from ruining the lifework of a poor little programmer?) - they sure would have foreseen that they can't punish or harm VR, if they chose to accuse him of
having violated GPL in Rybka 1.
So my guess is, FSF will most likely try to charge VR of having violated GPL in later Rybka versions (2-4), and they'll hope that the judge will find VR guilty of whatever FSF chooses to impend on him, and that the judge forces VR to make
his later Rybka versions (2-4) Open Source (GPL compliant).
That way FSF could really ruin VR's lifework substantially, after all - R2-R4 are/were the versions for which VR has spent the last 6 years of his life and which yielded him his (meager?) means of subsitence.
My magic marble somehow tells me that things have a high chance to work out like that in the near future, mainly because I've never trusted FSF anyways - and of course that's a really dangerous situation for VR's and Rybka's future now.
It's not easy to 'survive' a court batte if your opponent is FSF (no matter how wrong their allegations may be, their lobby in our world's jurisdiction is just too influential).
Well ... IF my magic marble is right on all this then sure Rybka, any further versions of Rybka, and
(most important) VR as programmer will certainly be obliterated from computer chess' history & future.
It's only computer chess, after all, right?
Doesn't really matter to any computer chess enthusiast at all, who has enjoyed Rybka since 2005, right?
Hmm, Trotzky?
Do you think that I've just described the ideal run of events in the near future - of course from Hyatt's & his ICGA inquistor's points of view?
Ruining VR's reputation clearly wasn't enough damage, no, they keep on spreading their lies in almost every CC forum there is, they sure won't ever put by this Rybka issue before FSF will have sued VR successfully
until he's ruined completely and also as a person.
I guess now I've clearly DEBUNKED the malicious intentions of the ICGA-inquisitors, all they ever wanted is the total destruction of VR as a person! Those ICGA-inquisitors are nothing else but abyssally vitriolic buggers who have nothing else to do in their spare time but finding evidence which they can use in order to ruin VR.
Spending 6 years for finding some dubious evidence testifies how obsessed they were with damaging VR's reputation & VR as a person. Even now they still can't get enough and keep on spreading their lies all over the internet until VR's lifework Rybka is completely destroyed by the FSF.
DEBUNKED.
I have never seen him say he does not have the R1 source code. He says he does not have the R3 source.
Right, but if Vas "happens" to lose the source code for Rybka 1 then I assume that that convenience could not possibly escape even the most blind Vas follower. At least I hope ;)
My only comment is that you do _not_ need source to prove a copyright infringement. You need a good assembly language programmer / compiler person. I don't know why you think that can't be done reliably, but you are simply 100% wrong. YOU might not be able to take assembly language and compare it to C, but there are a number of us that can certainly do this. And that number is growing by at least 25-30 per semester thanks to my CS330 class. And there are others that teach this at other universities as well. So the actual number is far higher. But keep saying it can't be done, maybe _somebody_ will believe it here and there if you say it often enough.
But it is a crock. A BIG crock.
But it is a crock. A BIG crock.
>> My only comment is that you do _not_ need source to prove a copyright infringement. You need a good assembly language programmer / compiler person
Well then, so please remember that a compiler will often replace certain terms of the C-source code by equivalent (but faster) instructions before it maps them to Opcodes.
Don't be deluded and think you or your students are -the-only- guys out there who know assembly language, in any case the transformation C-Source-Code -> Assembly Language certainly is NOT bijective at all, hence it's never been possible to reconstruct the
exact C-Source Code.
However, the latter would be a requirement to prove verbatim code copying, so everything a (good) defender needs to demonstrate to a layman (any expert witness should know about the difficulties of reconstructing
the-one-and-only-possible line of C-Source Code which may have caused the compiler to choose a certain Opcode/Register combination anyways) are several examples when totally different lines of C-source code are compiled to identical Machine Instructions.
I really don't doubt you know exactly what I'm talking about, after all you're certainly an expert of assembly language, and that's why I'm wondering why you keep on saying that a compiler's transformation C-source code -> Assembly Language is bijective and
(as a consequence) doubtlessly reversible. That's simply NOT true.
E.g. modern day compilers (if you tell them to optimize your program for speed) will often replace terms like a *= 4 with the equivalent (but faster) line of code a <<= 2 before they translate it to machine instructions.
Hence it's demonstrated that two (syntactically) different lines of C-Source Code can be compiled to identical Opcodes.
On many occasions where the C-programmer uses loops with only a few iterations you won't find a matching JE instruction at all in the assembly because the compiler unrolled it (if it was told to optimize your source code for speed, I guess that's certainly a compiler
option a chess programmer should choose because speed is critical), the compiler isn't really constrained to follow the programmer's intentions of using char, short or whatever other data types either and will often store char/short variables as Integers if it thinks
it's better for performance (it does so even without the programmer knowing about this). The fact that for/while/do loops often are totally interchangeable and will be compiled to identical blocks of machine instructions (more often than not) is certainly a thing an
assembly 'expert' like you should know, too. Of course I could list several more examples in order to disprove your claim, that the translation process C-source-code -> Assembly language is bijective and 100% reversible, but - assuming you're an assembly expert -
you should have known better in the first place, anyways. So - once more - this renews the question, why an assembly-language-expert like you would keep on spreading the lie of being able to reconstruct the exact lines of Rybka 1's C-source code
by only looking at its RE disassembly? You should know better, that this is not possible. The very fundamental premise of your argument, The translation of C-source-code to Assembly Language is bijective is clearly wrong.
I seriously hope you don't teach this wrong premise of your argument to your 25-30 CS students ...
Jury Osipov has mentioned this in some other post already, that reconstructing the exact lines of C-source code from only looking at a RE disassembly is impossible.
He's certainly an expert programmer (that is, if he's the same guy who programmed Strelka) and not a layman, so maybe that's more convincing for you if you read his remarks about this matter ...
Whether you wanna reconsider your (wrong) idea of being able to reconstruct the exact lines of Rybka 1's C-source code (did VR use C-source for R1 at all? It's not even possible to say THAT for sure - from only looking at a disassembly ...), whether FSF will sue
VR and even whether they succeed with such a destructive case doesn't really matter from VR's point of view anyways. If he doesn't have a copy of R1's source code any longer he can't comply to GPL either way, so at most he'll need to pay a few bucks to FSF
and F. Letouzy for having sold Rybka 1 (0$ !! Really outrageous, isn't it? I doubt that you had given away Crafty for 0$ ever - in case Crafty had been the clearly strongest chess program at that time ...), and that's it.
However this FSF case should turn out, VR's way of dealing with old and out-dated source code (burn the ships) just doesn't allow for Rybka's sources to become Open Source/GPL-compliant (so don't be deluded, you won't ever get a chance to study Rybka's
source code and learn about it's superior pruning-techniques in a more comfortable way than studying its RE disassemblies ... Maybe this whole unworthy Rybka case is really only indebted to your academic interest of studying VR's secrets he used throughout his
Rybka career ...).
You can't be serious that a major lawsuit of FSF against VR is fruitful (
) for computer chess, after all - the likely outcome of such a case would be the retirement of a highly competent chess programmer like VR from computer chess, and sure VR can stillcontribute so many precious ideas to chess programming (also from an academic point of view) that this would certainly be a great loss to computer chess as a whole.
So if this should really be the run of events in the near future (just don't say you want this to happen, or do you?) the great loser of this whole Rybka inquisition would be computer chess as a whole and not only VR ...
What is even worse is that another great programmer like Robert Houdart doesn't even get a single chance to present his super-strong chess engine Houdini in an ICGA world championship.
Again, this is only due to a few self-appointed chess programming experts who have spread the rumours that Houdini isn't Robert Houdart's original work, no matter how often Robert Houdart reinforced his claim that Houdini really is his work.
Again you and your talkchess fellows come up with the highly questionable argument that a super-strong chess program like Houdini couldn't appear out of the blue sky. Just the same (wrong) argument you used to 'disprove' Rybka's originality.
So VR really isn't the only very competent and highly talented programmer of a super-strong chess program whose reputation has fallen victim to your & your ICGA-inquisitors' efforts of defaming who-ever is capable of programming a stronger chess program than
your own. This argument that great chess program couldn't appear out of nowhere just doesn't hold. It's been disproved several times before Rybka 1.0, and Houdini is only one more evidence to the contrary ...
It's hard to say how many great chess programmers have never dared to publish the results of their hard work on a super-strong chess programs, because they've witnessed how unworthily the computer chess community treats a 'sudden' newcomer whose programs
are too strong for their own ... But I'd guess there are a lot of them. And that's certainly annoying from a user's point of view.
regards
Mec
The crux of your argument seems to be that one would need the exact source to prove infringement. This depends on your definition of “prove”, of course, but usually what you need for a court is just something that is beyond reasonable doubt. If so, you don't need an exact bijection.
/* Steinar */
/* Steinar */
Yeah but certainly it can be shown beyond reasonable doubt (in case VR lost R1's sources and only R1's disassembly is available) that due to the fact VR used a different board representation, a different set of search routines and also several different pruning techniques (some of them weren't even known before VR & Rybka came along), that R1's source code must have looked substantially different from Wegner's fantasy code or Fruit's original source code. Even in case VR really lost R1's source code.
Assuming that the type of board representation is so central & fundamental for every chess program, and that it's referenced many thousands of times throughout the whole source code, this fact certainly doesn't encourage any further considerations
whether substantial & verbatim copying occured.
There just isn't any means to determine whether VR programmed his bitboard engine Rybka from scratch (by typing the required C-source code to an empty Rybka.c), or whether he first marked major portions of Fruit's source code, copied & pasted them to his rybka.c, and only then began to replace those thousands of lines where Fruit used a reference to its substantially different (mailbox) board, called (substantially) different search & pruning routines etc. ... by means of overwriting them with a semantically equivalent or at least similar bitboard code.
Both ways of how VR might have programmed Rybka.c are at least equally reasonable to any objective third party, however - the second version (which Bob & his fellows believe) requires the premise that VR is-guilty-no-matter-what
in order to believe that Hyatt's & Wegner's version is also reasonable. Their whole 'proof' is basically an exemplary way of circular argumentation, because their premises (VR-is-guilty-no-matter-what) and conclusions (VR-is-guilty-no-matter-what)
are actually equivalent. The whole chain of arguments in between doesn't really increase the entropy of their whole theory (VR copied Fruit) at all.
In a regular proceeding however VR is innocent until proven otherwise, hence the accuser has the burden of proof (beyond reasonable doubt).
And well ... in order to prove their point of view the FSF would at least need to refute Version 1 (VR is a reasonable chess programmer, hence in case he decided to program a chess engine which uses a different board representation, a different set of search
routines plus several other & original pruning techniques (compared to Fruit) he certainly would've found it impractical to derive Rybka from Fruit (due to those substantial differences);
And thus he simply programmed Rybka from scratch instead (like every other honorable chess programmer would've done it, if they had seen how extraordinary different Fruit was from the kind of program they wanted to realize ...)
Even in case VR or any other reasonable programmer really wanted to derive Rybka from some other chess engine (by means of copy & paste) it's very unlikely that he'd choose a program like Fruit for this purpose, which does so many things in such a profoundly different manner than what VR wanted (Bitboard, better search routines, better pruning) when he decided to program Rybka 1.
Such a programmer certainly who really had the will to derive his chess program from another one would've certainly chosen any other chess engine (not Fruit), which would've had all those things (Bitboards etc.) already built-in and hence would've been way
more suitable for deriving Rybka by means of copy & paste.
It's really very counter-intuitive to derive a program from Fruit if the derivative is supposed to do so many things in a totally different way than Fruit, that you'll hardly ever find a reasonable programmer out there who'd be so stupid.
You certainly couldn't copy anything more than very, very few lines of source code in case you'd really wanted to derive Rybka from Fruit, using copy & paste in such an eccessive manner like the ICGA-inquistors imputed it on VR.
The ICGA inquistors (but certainly not FSF) also forgot, that in order to prove Copyright Infridgement they'd also need to show substantial portions of source code which were verbatim (not semantically!) equivalent in Rybka's and Fruit's sources.
A few lines of identical (fantasy) code here and another few lines (of fantasy code) there certainly aren't enough to prove substantial code copying, as these very few lines of (allegedly) 'identical' lines of source code can certainly be the result of
coincidence (if you consider that competitive chess programs contain 50.000 lines of C-Source Code there's really a very high chance of finding 'a few lines' of identical source code in both of them. Btw. this Nr. of 50.000 was mentioned by chess programming
expert Prof. Hyatt himself and certainly not 'invented' by me ...)
On top of that the ICGA-inquisitors even forgot to do some kind of AFC test in the first place (of course that's somewhat difficult if the author of R1 seems to have lost R1's age-old sources, however - things like harddisk crashes can also happen to VR's computer, so this shouldn't be blamed on VR either (the loss of six-years old source codes due to vis major (harddisk failure) certainly isn't unusual at all, if you consider the low durability of today's hardware and its predetermined breaking points!)
Well, there may be several other reasons why VR didn't want to open R1's sources, you know - some programmers of commercial software have a really messy style of programming (which certainly doesn't exclude the possibility that their sources are written in a very efficient manner!),
hence some of them are ashamed if they need to show 'their' source codes to other programmers (especially if they're dogmatic & intolerant source code aesthetes like Bob Hyatt & his ICGA fellows).
But guess what?
VR simply isn't a student at Bob Hyatt's school, so actually he really hasn't any good reason at all of showing Rybka's (probably really 'messy' and 'unaesthetic') sources to anyone ...
He even has the 'freedom' ('free' as in 'freedom') to drag his messy R1 sources over Windows' dustbin (perhaps unintentionally), if only for the reason he suddenly came to worry about source code aesthetics who want to dismiss his messy style of programming ...
hmmm, all I'm doing here is presenting all kinds of good counter-evidence & counter-arguments which could inspire VR's trial lawyer in case the FSF really chooses to sue VR for whatever damages he may have caused
(Remember that R1 was free 0$, so any demand of profit-sharing by the FSF would be ridiculous anyways ...)
regards
Assuming that the type of board representation is so central & fundamental for every chess program, and that it's referenced many thousands of times throughout the whole source code, this fact certainly doesn't encourage any further considerations
whether substantial & verbatim copying occured.
There just isn't any means to determine whether VR programmed his bitboard engine Rybka from scratch (by typing the required C-source code to an empty Rybka.c), or whether he first marked major portions of Fruit's source code, copied & pasted them to his rybka.c, and only then began to replace those thousands of lines where Fruit used a reference to its substantially different (mailbox) board, called (substantially) different search & pruning routines etc. ... by means of overwriting them with a semantically equivalent or at least similar bitboard code.
Both ways of how VR might have programmed Rybka.c are at least equally reasonable to any objective third party, however - the second version (which Bob & his fellows believe) requires the premise that VR is-guilty-no-matter-what
in order to believe that Hyatt's & Wegner's version is also reasonable. Their whole 'proof' is basically an exemplary way of circular argumentation, because their premises (VR-is-guilty-no-matter-what) and conclusions (VR-is-guilty-no-matter-what)
are actually equivalent. The whole chain of arguments in between doesn't really increase the entropy of their whole theory (VR copied Fruit) at all.
In a regular proceeding however VR is innocent until proven otherwise, hence the accuser has the burden of proof (beyond reasonable doubt).
And well ... in order to prove their point of view the FSF would at least need to refute Version 1 (VR is a reasonable chess programmer, hence in case he decided to program a chess engine which uses a different board representation, a different set of search
routines plus several other & original pruning techniques (compared to Fruit) he certainly would've found it impractical to derive Rybka from Fruit (due to those substantial differences);
And thus he simply programmed Rybka from scratch instead (like every other honorable chess programmer would've done it, if they had seen how extraordinary different Fruit was from the kind of program they wanted to realize ...)
Even in case VR or any other reasonable programmer really wanted to derive Rybka from some other chess engine (by means of copy & paste) it's very unlikely that he'd choose a program like Fruit for this purpose, which does so many things in such a profoundly different manner than what VR wanted (Bitboard, better search routines, better pruning) when he decided to program Rybka 1.
Such a programmer certainly who really had the will to derive his chess program from another one would've certainly chosen any other chess engine (not Fruit), which would've had all those things (Bitboards etc.) already built-in and hence would've been way
more suitable for deriving Rybka by means of copy & paste.
It's really very counter-intuitive to derive a program from Fruit if the derivative is supposed to do so many things in a totally different way than Fruit, that you'll hardly ever find a reasonable programmer out there who'd be so stupid.
You certainly couldn't copy anything more than very, very few lines of source code in case you'd really wanted to derive Rybka from Fruit, using copy & paste in such an eccessive manner like the ICGA-inquistors imputed it on VR.
The ICGA inquistors (but certainly not FSF) also forgot, that in order to prove Copyright Infridgement they'd also need to show substantial portions of source code which were verbatim (not semantically!) equivalent in Rybka's and Fruit's sources.
A few lines of identical (fantasy) code here and another few lines (of fantasy code) there certainly aren't enough to prove substantial code copying, as these very few lines of (allegedly) 'identical' lines of source code can certainly be the result of
coincidence (if you consider that competitive chess programs contain 50.000 lines of C-Source Code there's really a very high chance of finding 'a few lines' of identical source code in both of them. Btw. this Nr. of 50.000 was mentioned by chess programming
expert Prof. Hyatt himself and certainly not 'invented' by me ...)
On top of that the ICGA-inquisitors even forgot to do some kind of AFC test in the first place (of course that's somewhat difficult if the author of R1 seems to have lost R1's age-old sources, however - things like harddisk crashes can also happen to VR's computer, so this shouldn't be blamed on VR either (the loss of six-years old source codes due to vis major (harddisk failure) certainly isn't unusual at all, if you consider the low durability of today's hardware and its predetermined breaking points!)
Well, there may be several other reasons why VR didn't want to open R1's sources, you know - some programmers of commercial software have a really messy style of programming (which certainly doesn't exclude the possibility that their sources are written in a very efficient manner!),
hence some of them are ashamed if they need to show 'their' source codes to other programmers (especially if they're dogmatic & intolerant source code aesthetes like Bob Hyatt & his ICGA fellows).
But guess what?
VR simply isn't a student at Bob Hyatt's school, so actually he really hasn't any good reason at all of showing Rybka's (probably really 'messy' and 'unaesthetic') sources to anyone ...
He even has the 'freedom' ('free' as in 'freedom'
) to drag his messy R1 sources over Windows' dustbin (perhaps unintentionally), if only for the reason he suddenly came to worry about source code aesthetics who want to dismiss his messy style of programming ...hmmm, all I'm doing here is presenting all kinds of good counter-evidence & counter-arguments which could inspire VR's trial lawyer in case the FSF really chooses to sue VR for whatever damages
he may have caused(Remember that R1 was free 0$, so any demand of profit-sharing by the FSF would be ridiculous anyways ...)
regards
I didn't read all of your post, since most of it was not relevant to what I said at all. I have not taken any stance here; I was just pointing out where you seem to talk past each other.
/* Steinar */
/* Steinar */
![[ch]](/mwf/v12/flags/ch.png "Switzerland")
> Well then, so please remember that a compiler will often replace certain terms of the C-source code by equivalent (but faster) instructions before it maps them to Opcodes.
>
> Don't be deluded and think you or your students are -the-only- guys out there who know assembly language, in any case the transformation C-Source-Code -> Assembly Language certainly is NOT bijective at all, hence it's never been possible to reconstruct the
>
> exact C-Source Code.
>
This has been stated many times, including Bob himself.
> However, the latter would be a requirement to prove verbatim code copying, so everything a (good) defender needs to demonstrate to a layman (any expert witness should know about the difficulties of reconstructing
>
> the-one-and-only-possible line of C-Source Code which may have caused the compiler to choose a certain Opcode/Register combination anyways) are several examples when totally different lines of C-source code are compiled to identical Machine Instructions.
>
There are two types of copying in relation to copyright. Literal copying and non-literal copying. Literal copying (or verbatim copying) is what you are refering to, and, in the case of Fruit/Rybka, can not be proven 100% without the Rybka source code. Also, given that Fruit is mailbox and Rybka is bitboard, literal copying did not occur in regards to the evaluation. Said by more than one person, and again that includes Bob.
Non-literal copying is paraphrasing or translating a work without changing the overall structure and organization (in order to avoid copyright violation I must provide this link
http://euro.ecom.cmu.edu/program/law/08-732/Copyright/CopyrightInfringementOfSoftware.pdf ). That is the charge here against Rybka. As noted in the article, the determination of non-literal copyright violations has not been completely determined in U.S. courts (I have no idea about other countries, though Mark Watkins seems to have done some reading about Polish law). However, the ICGA was not concerned about copyright violations. Rather, the non-originality of works entered into its competions is its concern. Despite the flaws in the investigation process and in the presentation of the final report, the non-originality of earlier versions of Rybka seems to have been shown. It appears that non-literal copying of Fruit took place. The amount of non-literal copying is still open to debate.> I really don't doubt you know exactly what I'm talking about, after all you're certainly an expert of assembly language, and that's why I'm wondering why you keep on saying that a compiler's transformation C-source code -> Assembly Language is bijective and
>
> (as a consequence) doubtlessly reversible. That's simply NOT true.
>
>
The closest thing that I have seen Bob say in relation to this is that an expert can infer the nature of the unseen source code from looking at the disassembled object code. Tedious but straight-forward, to paraphrase Bob. Seeing as how there are several people who have demonstrated the ability to ascertain facts from closed-source engines (Jury Osiprov, Richard Vida, the mastermind behind Ippolit, Mark Watkins, Zach Wegner, etcetera), there is no way I can't believe him. Besides, he has explicitly stated that it is a surjective mapping.
Adam
Literal copying (or verbatim copying) is what you are refering to, and, in the case of Fruit/Rybka, can not be proven 100% without the Rybka source code.
In the case of Fruit/Rybka there CANNOT be any verbatim copying and nobody makes that claim. According to the ICGA report the
similarity between the evaluation in Fruit and Rybka is the most damaging. To quote fro the report on evaluation:
"Simply put, Rybka's evaluation is virtually identical to Fruit's. There are a few important changes though, that should be kept in mind when viewing this analysis.
Most obviously, the translation to Rybka's bitboard data structures. In some instances, such as in the pawn evaluation, the bitboard version will behave slightly differently than the original. But the high-level functionality is always equivalent in these cases; the changes are brought about because of a more natural representation in bitboards, or for a slight speed gain. In other cases the code has been reorganized a bit; this should be seen more as an optimization than as a real change, since the end result is the same. All of the endgame and draw recognition logic in Fruit has been replaced by a large material table in Rybka. This serves mostly the same purpose as the material hash table in Fruit, since it has an evaluation and a flags field. All of the weights have been tuned. Due to the unnatural values of Rybka's evaluation parameters, they were mostly likely tuned in some automated fashion. However, there are a few places where the origin of the values in Fruit is still apparent: piece square tables, passed pawn scores, and the flags in the material table."
From this report its clear that the claim is NOT that the evaluation tables in Fruit were copied in the verbatim way. Even though there were no verbatim copying (at least this is not claimed in the reports) the ICGA rules could still be broken. The case with "great lion" shows that the ICGA rules are interpreted in a quite restrictive way.
To me its important to establish whether its conceivable that Vas after having had studied Fruit simply implemented something akin to Fruits evaluation tables (without having the Fruit code in front of him). If this is a likely scenario the methodology used in the ICGA investigation does not distinguish the difference between reusing ideas (which is OK) and non-literal copying (=translating). I think this key point needs be investigated further.
In the case of Fruit/Rybka there CANNOT be any verbatim copying and nobody makes that claim. According to the ICGA report the
similarity between the evaluation in Fruit and Rybka is the most damaging. To quote fro the report on evaluation:
"Simply put, Rybka's evaluation is virtually identical to Fruit's. There are a few important changes though, that should be kept in mind when viewing this analysis.
Most obviously, the translation to Rybka's bitboard data structures. In some instances, such as in the pawn evaluation, the bitboard version will behave slightly differently than the original. But the high-level functionality is always equivalent in these cases; the changes are brought about because of a more natural representation in bitboards, or for a slight speed gain. In other cases the code has been reorganized a bit; this should be seen more as an optimization than as a real change, since the end result is the same. All of the endgame and draw recognition logic in Fruit has been replaced by a large material table in Rybka. This serves mostly the same purpose as the material hash table in Fruit, since it has an evaluation and a flags field. All of the weights have been tuned. Due to the unnatural values of Rybka's evaluation parameters, they were mostly likely tuned in some automated fashion. However, there are a few places where the origin of the values in Fruit is still apparent: piece square tables, passed pawn scores, and the flags in the material table."
From this report its clear that the claim is NOT that the evaluation tables in Fruit were copied in the verbatim way. Even though there were no verbatim copying (at least this is not claimed in the reports) the ICGA rules could still be broken. The case with "great lion" shows that the ICGA rules are interpreted in a quite restrictive way.
To me its important to establish whether its conceivable that Vas after having had studied Fruit simply implemented something akin to Fruits evaluation tables (without having the Fruit code in front of him). If this is a likely scenario the methodology used in the ICGA investigation does not distinguish the difference between reusing ideas (which is OK) and non-literal copying (=translating). I think this key point needs be investigated further.
So by accounts it is not that the data gathered in the ICGA report is conclusively evidential -it is far from being definitively concrete- but gives way to being problematic by way of interpretation. [edit] Which is what I believe michiguel was also pointing out.
>> Literal copying (or verbatim copying) is what you are refering to, and, in the case of Fruit/Rybka, can not be proven 100% without the Rybka source code.
>
> In the case of Fruit/Rybka there CANNOT be any verbatim copying and nobody makes that claim.
>
From what I understand, a mailbox to bitboard translation does not involve the search function. So I was leaving open the possibility of copying parts of Fruit's search. Though, the investigation dealt primarily with evaluation and so someone like me has to assume that none of the search was copied. In other words, I was trying make no assumptions and make no charges.
> From this report its clear that the claim is NOT that the evaluation tables in Fruit were copied in the verbatim way. Even though there were no verbatim copying (at least this is not claimed in the reports) the ICGA rules could still be broken. The case with "great lion" shows that the ICGA rules are interpreted in a quite restrictive way.
>
Your statement is exactly what some people tend to forget. As far as restrictive interpretation of ICGA rules go, that is certainly open to criticism but ultimately is up to the ICGA board.
> To me its important to establish whether its conceivable that Vas after having had studied Fruit simply implemented something akin to Fruits evaluation tables (without having the Fruit code in front of him). If this is a likely scenario the methodology used in the ICGA investigation does not distinguish the difference between reusing ideas (which is OK) and non-literal copying (=translating). I think this key point needs be investigated further.
Yet the investigation seems to distinguish between reusing ideas (OK) and the plagiarism of the expression of those ideas (not OK). Plagiarizing somebody else's expression of ideas is also non-literal copying.
To quote Mark Watkins:
"From the standpoint of originality, the principal element of evidence in the Fruit/Rybka controversy seems to be the strong overlap in choice of evaluation features. To be clear, the argument is not that there is a general commonality, but nearly an exact match, both in choice of features and (in the majority of cases) specific conditions in their implementation. Nor is it particularly relevant that one can find everything in Fruit via a patchwork of prior engines --- it is the specific Fruit/Rybka realisation of "feature ideas" that is of import."
This conveys to me that the overall form of Fruit's evaluation was copied, that it was the blueprint for Rybka's evaluation. That falls under the realm of plagiarism.
>> There are two types of copying in relation to copyright. Literal copying and non-literal copying. Literal copying (or verbatim copying) is what you are refering to, and, in the case of Fruit/Rybka, can not be proven 100% without the Rybka source code
Good to read a more differentiated view of things from someone like you :)
Although you also wanna see Rybka, VR & his legacy go down the tubes - at least you haven't (actively) un-learnt to tell the differences of literal and non-literal copying.
It's a pity that some of your fellows like Hyatt, Watchman & Wegner still plague the CC forums all over the internet and spread their lie, that VR has copied major parts of Fruit's source code verbatim, and that the evidence for their claim mustn't be questioned ever and who ever does is a layman.
Well, as a cc enthusiast I certainly don't wanna see Rybka, VR & his legacy go down the tubes because I firmly believe that a genius like VR is still able to come up with several more & very precious contributions to computer chess and especially chess programming.
He has certainly proven that he's actually one of the best chess programmers and thus I'm sure he's still perfectly capable of accomplishing several more chess-programming miracles in the future (e.g. Rybka 5, although I'm afraid it will - again - be disassembled immediately and then be cloned by VR's Ippo-spongers ...). Guess why nobody will disassemble & clone Hiarcs, Rondo etc. ? They just aren't worth the effort.
If somebody's intention is to 'produce' a super-strong chess engine today he'll certainly choose to RE Rybka 3 or Rybka 4 for his filthy intentions.
So if VR really would be forced to retire from chess programming (due to that upcoming FSF lawsuit) it won't be any good for computer chess as a whole and even less so for those who've enjoyed using the clearly best chess engine
(or some dubious clone of it) for the purpose of analyzing their blunders (which they've played in their recent games of chess).
So no, it's not like a little, buggy & utterly redundant chess engine is - (don't get me wrong on this! I really appreciate everyone who's willing to take on the tedious work of programming any chess program!) - is about to go down the tubes if VR retires, it's more like the major carthorse of computer chess going down the tubes, which certainly has attracted a lot of people by itself - people, who would've never been interested in computer chess at all otherwise, in case Rybka & VR hadn't ever shown up.
It's just Rybka's very human-like way of playing chess (clearly the result of VR being an IM) which sets Rybka apart from Fruit & most other chess engines.
When watching Rybka's play I sometimes get the impression that it's possessed by the spirits of Capablanca & Bobby Fischer, due to it's positional way of playing chess (and compared to most other chess programs Rybka seems to follow long-term and very logical
strategic plans - hence watching its play looks like watching the play of a human chess master.)
Until 2005 the overall play of most chess engines seemed to be way more aimless (although they were nevertheless 'monsters' in tactical positions), and this has really only changed since an IM decided to program Rybka 1 in 2005.
The times of 'mechanical' and 'uninspired' computer play in chess are certainly over by now, and that's definitely owed greatly to VR and his efforts of programming Rybka.
Well, at least we Rybka supporters here on the Rybka Forum should certainly have a right to hope for the best, we'll hope that FSF won't sue VR and ruin his precious contributions to computer chess all over,
that we will get & enjoy Rybka 5 someday (2012 maybe? But hurry up, Vas! End of days is said to be Dec 21st, 2012! I sure wanna have experienced Rybka 5 before everything goes down the tube !!
),and that VR may still have a prosperous future and won't retire from chess programming - with or without ICGA, this shouldn't really prevent him from releasing several more Rybka 'miracles' in the future ...
And as long as FSF doesn't choose to charge Vas until he's a broken, impoverished & old guy I sure won't give up my hope that Rybka will 'survive' this whole, malicous scenario (which was caused by his envious and former fellow programmers at talkchess,
especially Prof. Hyatt, Wegner, Watchman & [YOU? Really? Do you really wanna be remembered someday as a guy who assisted the ICGA-inquisitors in their efforts of ruining the lifework of a very talented chess programmer like VR? Are you a computer chess
enthusiast at all? Do you think that ICGA's life-time ban on Vas is enough of a punishment now, or do you secretly hope that FSF will sue Vas until he & Rybka are tilted from history? Because that's basically what Hyatt, Wegner & Watchman want to happen.
Or so it seems to me whenever I read their posts here on the Rybka Forum. They sure can't do anything else aside from affirming their profound hate & scorn for VR & Rybka whenever they choose to post something here on the Rybka forum ... )
>>> There are two types of copying in relation to copyright. Literal copying and non-literal copying. Literal copying (or verbatim copying) is what you are refering to, and, in the case of Fruit/Rybka, can not be proven 100% without the Rybka source code
> Good to read a more differentiated view of things from someone like you :)
Thank you
>
> Although you also wanna see Rybka, VR & his legacy go down the tubes - at least you haven't (actively) un-learnt to tell the differences of literal and non-literal copying.
>
I have no desire to see Rybka, Vas, nor his legacy destroyed. I agree whole-heartedly with Jeremy Bernstein (aka sockmonkey). I believe that he said that there was much to admire about Vas' work with Rybka. But that does not excuse not acknowledging the apparent basis of Rybka and following the relevant GPL licensing (merely saying that Fruit was responsible for 20 Elo is not enough).
> It's a pity that some of your fellows like Hyatt, Watchman & Wegner still plague the CC forums all over the internet and spread their lie, that VR has copied major parts of Fruit's source code verbatim, and that the evidence for their claim mustn't be questioned ever and who ever does is a layman.
>
I have not seen the claim that Fruit's source was copied verbatim. Though, I do believe there is much confusion over how Bob has phrased certain things. And over what is a copyright violation. And actually the irrelevance of copyright to the ICGA's investigation concerning a contestant in its competitions.
I have to stop right now. I will be able to respond to the rest of your post a little later.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill