Christophe, Zach and Norman? Made a Promise- I Want It

[geots]

With all due respect, i waited till the elections were over. A good while back you guys dragged Vas' program and reputation thru the mud with innuendos and theories. When questioned, you promised you would show me (and the rest of us) the evidence- but that it just took time to put it all together. Well, enough time has passed- more than enough- and i would like the promise kept. Just think of it as a Christmas present to all of us.

Best,

[Zach Wegner]

Funny, I was just working on this...

[kranium]

With all due respect, i waited till the elections were over. A good while back you guys dragged Vas' program and reputation thru the mud with innuendos and theories. When questioned, you promised you would show me (and the rest of us) the evidence- but that it just took time to put it all together. Well, enough time has passed- more than enough- and i would like the promise kept. Just think of it as a Christmas present to all of us.

Best,

George...
i prefer to not re-open a can of worms, why stir up trouble?

i had hoped that your post below was sincere, and you wanted to move on...
http://www.talkchess.com/forum/viewtopi ... highlight=

for the record, here is my last post concerning this issue:

"I do want to point out that it all started as a simple discussion motivated by Rick Faddens post, and Vas's declaration that Strelka was his own software. the issue started as a simple discussion, not as a smear campaign.

it was a question begging to be asked:
if Stelka is known to be a clone of Rybka, (stated as such in Wikipedia, confirmed by Vas himself, w/ apparent agreement by CEGT and CCRL)
and since the Strelka source code was released with abundant Fruit code in it..., isn't it logical, fair, and correct to question the legitimacy of both Strelka and Rybka?

the chess community has said no:
to paraphrase: "we want 100% proof. what has been disasembled and shown is only UCI parser stuff, non-critical engine AI code, etc. and besides, Vas is respected and liked, and has stated it is clean."

I can and have accepted the chess communities decision, and have moved on 100%, and urge everyone to do the same."

for me, i'm 100% done with this discussion. i moved on months ago, and never promised anything to anybody, so i have no clue what you're talking about.
please let sleeping dogs lie...
(or is there some special need to re-open it with you taking the lead now that Vas is a member?)

PS- this is my last response on the issue...i'm not interested in discussing it any further.

Norm

[Rolf]

George...
i prefer to not re-open a can of worms, why stir up trouble?

i had hoped that your post below was sincere, and you wanted to move on...
http://www.talkchess.com/forum/viewtopi ... highlight=

for the record, here is my last post concerning this issue:

"I do want to point out that it all started as a simple discussion motivated by Rick Faddens post, and Vas's declaration that Strelka was his own software. the issue started as a simple discussion, not as a smear campaign.

it was a question begging to be asked:
if Stelka is known to be a clone of Rybka, (stated as such in Wikipedia, confirmed by Vas himself, w/ apparent agreement by CEGT and CCRL)
and since the Strelka source code was released with abundant Fruit code in it..., isn't it logical, fair, and correct to question the legitimacy of both Strelka and Rybka?

the chess community has said no:
to paraphrase: "we want 100% proof. what has been disasembled and shown is only UCI parser stuff, non-critical engine AI code, etc. and besides, Vas is respected and liked, and has stated it is clean."

I can and have accepted the chess communities decision, and have moved on 100%, and urge everyone to do the same."

for me, i'm 100% done with this discussion. i moved on months ago, and never promised anything to anybody, so i have no clue what you're talking about.
please let sleeping dogs lie...
(or is there some special need to re-open it with you taking the lead now that Vas is a member?)

PS- this is my last response on the issue...i'm not interested in discussing it any further.

Norm

It all began as a question and a debate, not a smear campaign, well, I have data out of this debate and I see therefore also a smear campaign in special for the following reasons. But let me make clear, you began and continued a debate in your own perception but people who remind you of what you three or four have announced and promissed, they want to cause trouble and open a can of worms. That is sort of foul play.

But now the main reason for my personal problem with our debate and I made this clear from the beginning.

As a matter of fact debates about computerchess programs move in different areas. There are legal aspects, technical aspects and also traditions in a field. And then we also have, excuse me, a psychological problem how these topics are discussed and how we can define "data" and "facts" and human made declarations.

To this latter complex the stick note above said "libel is not appreciated". Someone like Bob however said, it's already fact that the actually found data, months ago, at least justified further inquiry, and that was therefore going on, but he said, from all what he knew, the hints are so telling that also with the admission of Vas himself, that this first free version of Rybka is in parts Fruit taken. He went much farther and claimed that also the next versions therefore have the similar problem because he wouldnt believe that all the code could have been "re-written".

So, what do you expect us spectators to think about your "group". Bob said you are heavily working, but that it took time etc. man had already told you that this job is without meaning for several reasons. I dont want to repeat all of them but you gave one example yourself. The UCI "parser stuff". We then had already a little private court case where one side claimed that it's already a proven wrong if that "stuff" was taken and NOT mentioned, also if it's only <5% of the whole code.

The pressure on your group or team grew when Isi explained that probably >80% of the whole codes of all programs in chess is taken from past authors without mentioning every detail.

So, psychologically you already sat in a little sand hole. You are wrong if you always argued like in the quote above that no matter what might have happened the computerchess audience at least had "decided" that this were no problem for further mistrust etc.

Already on the basis of the mentioned aspects until here you really have no case. But your situation deteriorates rapidly if we analyse the incredible gaps in strength of the new Rybkas. If it was just 80% like others and some rest from Fruit, woudnt it therefore be reasonable to assume that there could be a "force" in the programmer Vas himself that placed him always almost 100 points ahead of all the competitors?? And could this be expected by Fruit alone?? That is the reflection normal people are making against your "research hypotheses".

Let me say it this way against you and Bob, in science, I mean real science, nobody would get enough money to research something on the base of your expectations because already with my little knowledge it's possible to judge it as unrealistical.

In that respect I agree with you that it's a wise decision that you have stopped any further "debating", because when you are already in a sand hole it's fatal to dig deeper. Because then a legal aspect comes back to you. What is with the damage against the original programmer.

The complete silence of some and the wise distancing of some more is therefore the best reaction if you want to get second place in the Table of Honor. ut numer one would have required that you admit that you have probaly made a mistake and that Vas has totlally confused you with his claim that Strelka is his program. It's really funny. Because for some of you this was really the basic source for your research expectation that then you could easily find something where Vas had really acted not exactly in the generally tolerated paths.

Ok, one can never know you might say ut people like my thread opener mainly asked you to apologize as a group and also each one of you for any possile damage you might have caused.

You know, it doesnt look kosher if you, without now the expected proof, elaorate about the end of debates, about opening a can of worms, how many, IMO not so many, and aout trouble. For whom? For Vas or for you?

BTW I'm still waiting for the proof that someone like Christophe Theron could ALSO increase the strength of his program if he did the "same".

[Nimzovik]

No offense.......but MOOOOOOOAAAAAAAAAANNNNNNNNNNNNNNNNNNNNNNNNNNNNNnnnnnnnnnnnnnnn................................ As a consumer I have what I want in Rybka 3. How it was hatched does not enter my realm of meditation. Just my Humble opinion.

[Gerd Isenberg]

The pressure on your group or team grew when Isi explained that probably >80% of the whole codes of all programs in chess is taken from past authors without mentioning every detail.

Do you mean me? Iirr I asked some questions, whether things may be explained elsewhere. I was and I am still torned. There are lot of symptoms about similarities to say at least, the uci-go parser stuff as well as a lot of eval constants and piece square tables. For the uci-go parser, I would vote to make that stuff with hindsight public domain rather than GPL. Some other stuff is more serious.

I may understand something about chess programming and assembly language, but I don't claim to have the competence to judge the findings so far as evidence and I have no time and energy to study the stuff by myself - because it is no fun and may be because I finally fear the truth.

The base algorithms are well known and public. Alpha-beta and even PVS are only a few lines of code. But the concrete implementation is usually many pages and zillions of implementation details may be recombined, and implemented in other ways. Otoh reading open source code may bias a lot of programmers, to write things from scratch in similar ways - recently when I re-read Ken Thompson's Reflections on Trusting Trust, I found following paragraph remarkable:

That brings me to Dennis Ritchie. Our collaboration has been a thing of beauty. In the ten years that we have worked together, I can recall only one case of miscoordination of work. On that occasion, I discovered that we both had written the same 20-line assembly language program. I compared the sources and was astounded to find that they matched character-for-character. The result of our work together has been far greater than the work that we each contributed.

Of course the same rule should be applied for everyone, no matter how much the original framework was improved. Whether it is Vasik Rajlich, Eugenio Castillo or me.

Cheers,
Gerd

[Rolf]

I was and I am still torned. There are lot of symptoms about similarities to say at least, the uci-go parser stuff as well as a lot of eval constants and piece square tables. For the uci-go parser, I would vote to make that stuff with hindsight public domain rather than GPL. Some other stuff is more serious.

Would you say that you could tolerate if collegues take a free prog but that is not open source and begin to make research with reveiling at first that RYBKA 1 beta is a clone or practically FRUIT? Let me make a lay comparison: take two different cars. The first is well known through Do-it-yourself advices , now you take the second, research it and go public with the same tech as the first model. How similar is my example with the CC debate RYBKA1 and FRUIT? Isnt it a stupid activity but not only this, isnt it a smear campaign against company of the second car?

I forgot to mention this: the second car is allegedly similar but it is 20% faster, same gazoline. Do you have explanation for that if the second is only built after the Do-it-yourself model? I dont buy that. And IF there is some totally new code/tech in the second, then how would you claim stealing from the model??

Where is the beef?

I may understand something about chess programming and assembly language, but I don't claim to have the competence to judge the findings so far as evidence and I have no time and energy to study the stuff by myself - because it is no fun and may be because I finally fear the truth.

Do you fear being guilty of unallowed activities? Do you want to help Vas by NOT starting further analyses? Do you need money?

The base algorithms are well known and public. Alpha-beta and even PVS are only a few lines of code. But the concrete implementation is usually many pages and zillions of implementation details may be recombined, and implemented in other ways. Otoh reading open source code may bias a lot of programmers, to write things from scratch in similar ways - recently when I re-read Ken Thompson's Reflections on Trusting Trust, I found following paragraph remarkable:

That brings me to Dennis Ritchie. Our collaboration has been a thing of beauty. In the ten years that we have worked together, I can recall only one case of miscoordination of work. On that occasion, I discovered that we both had written the same 20-line assembly language program. I compared the sources and was astounded to find that they matched character-for-character. The result of our work together has been far greater than the work that we each contributed.

Of course the same rule should be applied for everyone, no matter how much the original framework was improved. Whether it is Vasik Rajlich, Eugenio Castillo or me.

Cheers,
Gerd

Question: Why dont you or any of your collegues, nor the group of hunters, "research" FRITZ, SHREDDER?

Question: Why only RYBKA1 was taken? Because it was for free? But FRITZ LIGHT is also for free, why isnt it researched?

This is not at all a reproach against you, but I conclude in general: certain people might have motives against Vas because he was so revolutionary with his new program and much better than all others although it was only the first Beta. Therefore they took the free given version and ripped it apart which is only allowed if you have open source.

IMO Vasik is simply more clever than all others because if RYBKA 1 were only a copy then why nobody could faire the same copy? I see the similar motivation with project TOGA. Also that one claims similarity of FRUIT but it cant be compared with RYBKA and its performance. So TOGA is proving to me that RYBKA is the creation of a genius and not just a copy freak.

The topic reminds me of Einstein's alleged theories having been thought out by his wife. Perhaps Vask's brothers have helped him too. SUch familiar help is better than what Bob does with strangers in his Crafty group. Also Bob is more a statistic freak where Vasik is the one with new ideas...

[Edsel Apostol]

Apologies for interjecting into the topic but I just want to give an opinion on this issue that I've been following for quite some time now.

My humble opinion is that I think the reason why some people are enraged at Vas is when he says that without Fruit, Rybka 1 will only be 20 elo point weaker, but then the examination of the Strelka source code, that they say is a clone of Rybka 1, proves otherwise, that most of the ideas there are from Fruit.

This are some of the reasons why a lot of questions and issues was being raised. One of the issues concerned was about using ideas from open source engines. In my humble opinion again I think it is okay. But what about the issue of giving a false statement that the idea that was borrowed from an open source engine is not that effective wherein it was later demonstrated that the opposite is true?

I'm not attacking anyone here. All I can say is that most people here believe that the end justify the means.

[Michael Sherwin]

What Vas did with Rybka is perfectly fine. He made a better wiget--that is what us humans do--we make better wigets! Okay, he may have used a very good wiget as a base model, but then he used a better material for his wiget (bitboards) and made many improvements. His wiget is a far superior widget and he deserves the credit for his improved wiget!

My gripe is that Fruits source code was released in the first place. I wish that it hadn't been released. This is purely selfish as I believe that my program would have a substantially higher position in the rating list if Fruits sources hadn't been released and would be more important to people today than it is. The number of new/improved programs just skyrocketed after Fruit was released. My program though has not benifited from Fruits release as I have taken nothing from it or any other program except that I learned the basics from TSCP!

But, now my originality does not count for much and to even get close to the top, I must consider doing what Vas may have done!

[bob]

I was and I am still torned. There are lot of symptoms about similarities to say at least, the uci-go parser stuff as well as a lot of eval constants and piece square tables. For the uci-go parser, I would vote to make that stuff with hindsight public domain rather than GPL. Some other stuff is more serious.

Would you say that you could tolerate if collegues take a free prog but that is not open source and begin to make research with reveiling at first that RYBKA 1 beta is a clone or practically FRUIT? Let me make a lay comparison: take two different cars. The first is well known through Do-it-yourself advices , now you take the second, research it and go public with the same tech as the first model. How similar is my example with the CC debate RYBKA1 and FRUIT? Isnt it a stupid activity but not only this, isnt it a smear campaign against company of the second car?

I forgot to mention this: the second car is allegedly similar but it is 20% faster, same gazoline. Do you have explanation for that if the second is only built after the Do-it-yourself model? I dont buy that. And IF there is some totally new code/tech in the second, then how would you claim stealing from the model??

Where is the beef?

You simply miss the point. The probability of two different people designing something that is "close" is probable. But the probability that two individuals will design the same identical engine (displacement, number of cylinders, number of valves, valve diameters, camshaft overlap, duration, lift, fuel injector design, fuel pressure, flow rate, pressure adjustment, engine management computer, its software, its air/fuel/timing ratio tables, .... you get the idea. I have said it a hundred times now, developing similar ideas is possible. Within reason. Developing identical code is beyond improbable, just as in the car example above. Both might have a similar body style, but the internals will be _completely_ different.

Unless one steals the design or parts from the other, that is...

So let's get off this bad analogy approach unless you can choose an analogy that actually supports your position rather than contradicts it.

I may understand something about chess programming and assembly language, but I don't claim to have the competence to judge the findings so far as evidence and I have no time and energy to study the stuff by myself - because it is no fun and may be because I finally fear the truth.

Do you fear being guilty of unallowed activities? Do you want to help Vas by NOT starting further analyses? Do you need money?

The base algorithms are well known and public. Alpha-beta and even PVS are only a few lines of code. But the concrete implementation is usually many pages and zillions of implementation details may be recombined, and implemented in other ways. Otoh reading open source code may bias a lot of programmers, to write things from scratch in similar ways - recently when I re-read Ken Thompson's Reflections on Trusting Trust, I found following paragraph remarkable:

That brings me to Dennis Ritchie. Our collaboration has been a thing of beauty. In the ten years that we have worked together, I can recall only one case of miscoordination of work. On that occasion, I discovered that we both had written the same 20-line assembly language program. I compared the sources and was astounded to find that they matched character-for-character. The result of our work together has been far greater than the work that we each contributed.

Of course the same rule should be applied for everyone, no matter how much the original framework was improved. Whether it is Vasik Rajlich, Eugenio Castillo or me.

Cheers,
Gerd

Question: Why dont you or any of your collegues, nor the group of hunters, "research" FRITZ, SHREDDER?

Question: Why only RYBKA1 was taken? Because it was for free? But FRITZ LIGHT is also for free, why isnt it researched?

This is not at all a reproach against you, but I conclude in general: certain people might have motives against Vas because he was so revolutionary with his new program and much better than all others although it was only the first Beta. Therefore they took the free given version and ripped it apart which is only allowed if you have open source.

IMO Vasik is simply more clever than all others because if RYBKA 1 were only a copy then why nobody could faire the same copy? I see the similar motivation with project TOGA. Also that one claims similarity of FRUIT but it cant be compared with RYBKA and its performance. So TOGA is proving to me that RYBKA is the creation of a genius and not just a copy freak.

The topic reminds me of Einstein's alleged theories having been thought out by his wife. Perhaps Vask's brothers have helped him too. SUch familiar help is better than what Bob does with strangers in his Crafty group. Also Bob is more a statistic freak where Vasik is the one with new ideas...

Again, more nonsense without knowing past history. How long has Fritz been around? How long was it developed? That is the difference here. A program that rises to the top in less than 2 years is a bit different from a program that has evolved over 10-15 years. Then some noticed similarities between early Rybka and Fruit output, analysis, etc. The list goes on. There are plenty of reasons why this particular program was singled out... It was a clear example of something that violated past experiences...