Rybka 1.0 vs. Strelka

[trojanfoe]

Andy, you cant seriously mean it. Apart from me and partially Uri there is nobody here in defense of Rybka. Didnt you see this?

Oh yes I mean it, and it's unfortunate for Rybka that you are the only two 'defending' it - you put up a poor fight. The fact that you are the only two speaks volumes

If you would change into Rybka forum you would see many people who think different. But why should they argue in this climate. The reason for me to make some arguments is NOT that I know for sure what is going on but when the anti-Vas side is making shallow messages. Of course I cant judge on the quoted code at all. But I have experience how people are doing a witch-hunt.

Well the Rybka forum is, as would be expected, full of people who want to validate their purchase in any way possible. This is not new behaviour - we all do it in some form or other. It's called being a Zealout, but in more modern times, a 'fanboy'. You see it particularly amoungst console owners (mainly Sony against Microsoft). I don't like it there - too much like being in church or something where you feel like an alien among the 'converted'. It's not a place to have such a discussion.

But of course having such little support here, you would love to have the discussion there wouldn't you - kinda like having peace talks in one of the countries involved. This forum is Switzerland in comparison and I'd rather discuss it here, thanks.

This is actually what is happening here. Of course the triumph will be China. Otherwise these people here cant hurt Vas and Rybka.

Trimuph or not, it will cheapened if it's proved that Rybka was created from GPL'd code. The bigger they are...

Since you are a new name for me, just for me, and I still take you on a high level and ethical grounds would you please comment on what Corbit has reported here, namely that as he showed the analyses to Fabien, he reacted so, that he didnt care. And although some have explained that this meant that others couldt neither do something. But we still have the campaign here. Could you explain to a lay what tghis is all about? Some wrote "envie" but why?

You are funny Rolf - I will give you that - and one of life's personalities, that's for sure. I think we have 'met' before, in the 90's on rec.games.chess.computer, but I think we may have been ships that passed in the night. As for your question, I am not putting myself forward as an expert, however I have a great deal of programming experience and some of it within the chess domain, but I wouldn't say that this is a campaign (which is very negative) or that anyone is doing it out of 'envy'. I would say there is an element by which people would feel 'peeved' that the no.1 engine might have started life as an open source program; if it had been no.12 for example, I'm pretty sure this discussion wouldn't be so large. It's like watching someone being crowned king and knowing that they wouldn't have got there without the help of someone who has not been given the rewards they deserve. The difference is that the 'king', in this case, cannot stop the opposition from protesting - which is a very democratic IMHO.

And then we have a strange situation that the main attackers have themselves a special standing. I wait on a short exchange wih Bob and then I will decide how to handle that topic.

(shrug)

If you could help with some thoughtful reflections then it would be positive for the community. You know I can only oppose as a democrat and humanistic guy all kind of doctored annihilation of other peple and the products of their talents.

My fantasy often tells me that Vas will retaliate at his chosen moment and then people should not be surprised if they are taken to task. In case they are Europeans like CT. I fully agree with what Graham has written.

I think I have bollocked on enough, so I will stop now..

Cheers,
Andy

[geots]

Note that this is the discussion about Free version of Rybka (Rybka 1.0)

Zach gave me a permission to edit the thread title and and insert "1.0" next to the engine name wherever he mentioned the engine, just to avoid confusion.

I agree with ChrisW that whenever you mention Rybka, It'd help if you include the version number along with it.

I am not personally convinced that it matters. What is the probability that Rybka 3 is vastly different from Rybka2, and R2 vastly different from V1? Most do not do _complete_ rewrites, which means much GPL code, if it was present in R1 will also be present in R3. So this realistically applies to all versions, if it applies to any.

"if it applies to any"......... Extremely clever way you phrase that, when one considers that anyone with one eye and half a brain already knows your true feelings about this whole issue.

Bob has the right to express his professional opinion in this matter. Whether you agree or not.

Terry, from Christophe's thread and yours also- it is easy to see that neither of you caught on to what i was trying to get across. As Bob is certainly entitled to his opinion- i as well am entitled to mine.

Yes you are but you're suggesting Bob's assessment is defamatory, IE the carry over part....


Christophe and myself can read George.

Terry, i never once said you could not read. Your problem here seems to be with interpretation.

[Terry McCracken]

Note that this is the discussion about Free version of Rybka (Rybka 1.0)

Zach gave me a permission to edit the thread title and and insert "1.0" next to the engine name wherever he mentioned the engine, just to avoid confusion.

I agree with ChrisW that whenever you mention Rybka, It'd help if you include the version number along with it.

I am not personally convinced that it matters. What is the probability that Rybka 3 is vastly different from Rybka2, and R2 vastly different from V1? Most do not do _complete_ rewrites, which means much GPL code, if it was present in R1 will also be present in R3. So this realistically applies to all versions, if it applies to any.

"if it applies to any"......... Extremely clever way you phrase that, when one considers that anyone with one eye and half a brain already knows your true feelings about this whole issue.

Bob has the right to express his professional opinion in this matter. Whether you agree or not.

Terry, from Christophe's thread and yours also- it is easy to see that neither of you caught on to what i was trying to get across. As Bob is certainly entitled to his opinion- i as well am entitled to mine.

Yes you are but you're suggesting Bob's assessment is defamatory, IE the carry over part....


Christophe and myself can read George.

Terry, i never once said you could not read. Your problem here seems to be with interpretation.

Is it?

And what does Bob wish George.....Spell it Out!

So we won't misinterpret you.....

[bnemias]

3. If Vas copied Fruit, he ought to simply say so. Then he could rewrite every remaining line and be free of the GPL requirements and this would settle down.

Is that really true? I mean, the binaries are out already. Unfortunately, you can't fix retroactively, the binaries. Thus, to fix it would require releasing the source, no?

[Rolf]

Here is a summary of my thoughts, given in bits and pieces in previous posts.

1. Vas is obviously talented. Whether he originally copied fruit or not. He either wrote the currently strongest chess program from scratch, or else he copied fruit and then modified it into the stongest program around. Either task is not simple.

2. I did not ever like the "amateur turns professional" approach when the amateur lurks and asks lots of questions, and learns lots of things that would normally take years, and then when he finds a new idea, mum is the word. But I have lived with that for years and it doesn't cause me to lose sleep.

3. If Vas copied Fruit, he ought to simply say so. Then he could rewrite every remaining line and be free of the GPL requirements and this would settle down.

4. If he didn't copy fruit, then he ought to explain the glaring similarities between strelka (which he claims as his own) and Fruit, as presented by various posters here.

As far as I am concerned, either 3 or 4 would close this book and everyone would move on. But right now, the claims and questions are being answered by a deafening silence. Silence is a form of fertilizer for these discussions and they will continue to grow.

So at your age you dont prefer peace and friendship but you prefer dreaming about rising processes. Fine. Let it rise. But you have supported the mess, while I with Terry was for peace. George Speight too. Bob, let's continue this for 20 years and then we bury you with the expression "he looked into the rise of the war evidence instead of watching the daffodiles grow".

[bob]

The only legal problem I see looming is for me, because one day I am going to come to your house and kill all those damned monkeys you have typing over there.

(2) the source code snippets posted here provide a direct link between fruit and strelka/rybka, when examined closely. I've not done the data / evidence gathering, I simply commented on what was being shown.

So there is no "credo" I follow, other than I would personally like to see the truth, the whole truth, and nothing but the truth at some point in time. One can be convicted without ever taking the witness stand, if the evidence piles up until it becomes overwhelming. We are not there yet, but it is piling up with no contest from anyone at all.

ad 1) correct if you wanted to then you would do this. Fine. But the reverse is wrong. If you do it then the only reason could be you wanted to "cheat". Bob, this is just nonsense. Logically. But perhaps programmers have always both directions in mind.[/quote[



I have made no accusations. I have had two consistent comments that I have made multiple times:

(1) one would choose to obfuscate nodes, depth and PV if one wanted to attempt to hide/conceal internal details of how the search works.

The issue is that there is _no_ reason to doctor up the data to mislead. If it is not intended to mislead, then anyone could jump in and explain some other reason that will stand up under scrutiny. There are potential search algorithms where one could quite justifiably adjust the node count. It would not be technically correct to do so, but you could say "look, before I search a move, I make them each, one at a time, to look at the resulting positional scores, so that I can use that for ordering, extending, reducing and even pruning. And since I actually make the moves, and do that extra work, I am adding those to my node count, in addition to the moves I actually search with alpha/beta." I would respond "good for you" and never give it a second thought, other than to remember that your NPS number is going to be a bit different from others. Any sensible explanation would have put this to rest. But, apparently, there was no sensible way to explain what others were seeing with respect to nodes displayed and NPS.

Then there is depth. One could always say "OK, I search to depth D, but with reductions using R=3, my actual depth is much less in many positions so I am going to arbitrarily report depth-3 as my depth value. Some would probably point out that you also extend other lines, so maybe this is not a good idea, but it would be an explanation.

Then there is the PV. We know that Deep Blue PVs were incomplete, because they had no way to obtain the PVS from the chess processors, as they had no way to back up a PV, nor a hash table to allow later reconstruction. So we know their PVs are short by 6-7-8 plies depending on the depth the hardware processors were searching. Makes their PV less informative, but the reason is legitimate. just not showing the last few moves so that no one knows exactly what terminal position you are evaluating is another way to hide things. Give me several positions, and your scores, and I can begin to piece together what positional ideas you are evaluating.

So good explanations end discussion. Nonsense or no explanations simple cause it to continue and pop back up in a cyclical pattern until the issue is resolved.

ad 2) are you really happy to rely anything on the Osipov activities? And how do you know who he was and what motives had held him? What is the difference in your eyes between a reverse doctored Rybka 1 and Tiger 14? What sense does it give you ff you must work with illegally doctored data? Would you help to doctor Tiger for balance reasons? Or let's take Fritz 9. Would you help? If not why you are here in it against Rybka 1? Out of cooperation with CT in a revival of Ruffian times?

What does "help doctor xxx" mean? I have not extracted source. I have looked at what has been presented and drawn conclusions. And I have read lots of nonsense about how this could happen quite by chance...

Why not dream with me on peace? Making a picture in our brains and then let it become real? Peace is better than watching something piling up. Let' get ethical, Bob! Please! Peace for all!

Peace would be a natural consequence of answering the questions others are asking. I can't bring about peace in the middle east, I have no control. Ditto with this discussion. Only the "principals" can settle the issue, and one seems unwilling to discuss anything at all. I don't know how I could deal with that to help...

[trojanfoe]

So at your age you dont prefer peace and friendship but you prefer dreaming about rising processes. Fine. Let it rise. But you have supported the mess, while I with Terry was for peace. George Speight too. Bob, let's continue this for 20 years and then we bury you with the expression "he looked into the rise of the war evidence instead of watching the daffodiles grow".

Could you please make constructive counter arguments to each of Bob's comments? All you are saying is that Bob should keep his opinions to himself else he will stress himself into an early grave. Please make an effort Rolf.

Cheers,
Andy

[geots]

Note that this is the discussion about Free version of Rybka (Rybka 1.0)

Zach gave me a permission to edit the thread title and and insert "1.0" next to the engine name wherever he mentioned the engine, just to avoid confusion.

I agree with ChrisW that whenever you mention Rybka, It'd help if you include the version number along with it.

I am not personally convinced that it matters. What is the probability that Rybka 3 is vastly different from Rybka2, and R2 vastly different from V1? Most do not do _complete_ rewrites, which means much GPL code, if it was present in R1 will also be present in R3. So this realistically applies to all versions, if it applies to any.

"if it applies to any"......... Extremely clever way you phrase that, when one considers that anyone with one eye and half a brain already knows your true feelings about this whole issue.

Bob has the right to express his professional opinion in this matter. Whether you agree or not.

Terry, from Christophe's thread and yours also- it is easy to see that neither of you caught on to what i was trying to get across. As Bob is certainly entitled to his opinion- i as well am entitled to mine.

Yes you are but you're suggesting Bob's assessment is defamatory, IE the carry over part....


Christophe and myself can read George.

Terry, i never once said you could not read. Your problem here seems to be with interpretation.

Is it?

And what does Bob wish George.....Spell it Out!

So we won't misinterpret you.....

Terry, with all due respect to you- as i like you- this started off as a discussion between Bob and myself. And i think it should stay that way- i have no idea how it pertains to you.

[bob]

Note that this is the discussion about Free version of Rybka (Rybka 1.0)

Zach gave me a permission to edit the thread title and and insert "1.0" next to the engine name wherever he mentioned the engine, just to avoid confusion.

I agree with ChrisW that whenever you mention Rybka, It'd help if you include the version number along with it.

I am not personally convinced that it matters. What is the probability that Rybka 3 is vastly different from Rybka2, and R2 vastly different from V1? Most do not do _complete_ rewrites, which means much GPL code, if it was present in R1 will also be present in R3. So this realistically applies to all versions, if it applies to any.

"if it applies to any"......... Extremely clever way you phrase that, when one considers that anyone with one eye and half a brain already knows your true feelings about this whole issue.

So you are also a mind-reader as well? I am, and always have been "anti-clone". Other than that I have little interest in what goes on in computer chess misbehavior. But "clones" and "copying" indirectly affects everyone that is active... But I'd be interested in what you think "I think about the whole issue". I'm only following the GPL discussion and that is an important issue to consider.

But I'd be interested in what you think "I think about the whole issue". Ah, i never thought i would hear those words coming from you. Nice that you added "whole issue" - because what i am referring to has nothing to do with GPL licenses, legality in the computerchess field, or anything so mundane as that. Please lets dont argue about the appropriatness of the word "mundane" here. Each to his own. Past that- it's not a matter of what "I think you think"- but a matter of what "I know you wish".

And what do you believe that to be???

[geots]

Note that this is the discussion about Free version of Rybka (Rybka 1.0)

Zach gave me a permission to edit the thread title and and insert "1.0" next to the engine name wherever he mentioned the engine, just to avoid confusion.

I agree with ChrisW that whenever you mention Rybka, It'd help if you include the version number along with it.

I am not personally convinced that it matters. What is the probability that Rybka 3 is vastly different from Rybka2, and R2 vastly different from V1? Most do not do _complete_ rewrites, which means much GPL code, if it was present in R1 will also be present in R3. So this realistically applies to all versions, if it applies to any.

"if it applies to any"......... Extremely clever way you phrase that, when one considers that anyone with one eye and half a brain already knows your true feelings about this whole issue.

So you are also a mind-reader as well? I am, and always have been "anti-clone". Other than that I have little interest in what goes on in computer chess misbehavior. But "clones" and "copying" indirectly affects everyone that is active... But I'd be interested in what you think "I think about the whole issue". I'm only following the GPL discussion and that is an important issue to consider.

But I'd be interested in what you think "I think about the whole issue". Ah, i never thought i would hear those words coming from you. Nice that you added "whole issue" - because what i am referring to has nothing to do with GPL licenses, legality in the computerchess field, or anything so mundane as that. Please lets dont argue about the appropriatness of the word "mundane" here. Each to his own. Past that- it's not a matter of what "I think you think"- but a matter of what "I know you wish".

And what do you believe that to be???

I do really think that a point has been reached where it would be to no one's benefit to carry this further on CCC. I would be more than happy to discuss this with you further in private, if you like. I actually think it would be a good idea, because i have some preconceived ideas about your motives that i think you should have the chance to defend. And everyone deserves that chance.

Best,