I am feeling ill

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

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Milos
Posts: 3387
Joined: Wed Nov 25, 2009 12:47 am

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by Milos » Tue Dec 01, 2009 7:06 pm

Rolf wrote:We have no murder case at all. We have Rybka, the best program. By a notable scientist from MIT. My argument is this.
...
Let's believe in honesty of people from MIT.
I really don't understand your childish fascination with MIT, seams like you have watched Good Will Hunting too many times.
Every year tons of students finish MIT, a lot of them get a PhD there, a heavy majority of them are Chinese/Indian/Russian. This doesn't make them notable scientists. Moreover, making the strongest commercial chess engine doesn't make someone a noticeable scientist either.

Noticeable scientist is Bob and ppl like him, who wrote countless great papers on the subject, who pioneered so many ideas, who taught generations of young programmers and chess enthusiast...

And here we have you, with rather limited and rudimentary chess programming knowledge arguing with Bob. These are really silly stuff.

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by Rolf » Tue Dec 01, 2009 7:16 pm

Milos wrote:
Rolf wrote:We have no murder case at all. We have Rybka, the best program. By a notable scientist from MIT. My argument is this.
...
Let's believe in honesty of people from MIT.
I really don't understand your childish fascination with MIT, seams like you have watched Good Will Hunting too many times.
Every year tons of students finish MIT, a lot of them get a PhD there, a heavy majority of them are Chinese/Indian/Russian. This doesn't make them notable scientists. Moreover, making the strongest commercial chess engine doesn't make someone a noticeable scientist either.

Noticeable scientist is Bob and ppl like him, who wrote countless great papers on the subject, who pioneered so many ideas, who taught generations of young programmers and chess enthusiast...

And here we have you, with rather limited and rudimentary chess programming knowledge arguing with Bob. These are really silly stuff.

My expertise comes from Princeton BTW and it allowed me to see that we have no murder case. Write me a postcard if you disagree.
-Popper and Lakatos are good but I'm stuck on Leibowitz

Milos
Posts: 3387
Joined: Wed Nov 25, 2009 12:47 am

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by Milos » Tue Dec 01, 2009 7:36 pm

Rolf wrote:My expertise comes from Princeton BTW and it allowed me to see that we have no murder case. Write me a postcard if you disagree.
Again, a lot of childish bragging. You could have a triple PhD from Harvard, Stanford and MIT in law, medicine and chemistry respectively and you would still probably know less about programming then any hacker kid from the neighborhood.
On the other hand you have youths like Zach who probably know more about chess programming then some mathematics PhDs from MIT.
University itself doesn't provide any scientific respect nor a commercial success does it (otherwise we could call B. Gates noticeable scientist :lol:). Papers and works in the field are the things that do that.
Last edited by Milos on Tue Dec 01, 2009 7:40 pm, edited 1 time in total.

K I Hyams
Posts: 3508
Joined: Fri Mar 31, 2006 9:21 pm

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by K I Hyams » Tue Dec 01, 2009 7:37 pm

Graham Banks wrote:
bob wrote: (1) Vas took parts of a GPL program to build a program that he then used to violate the GPL itself, because if you use parts of a GPL program, your program must also be GPL compliant. Rybka 1 was not.
Vas posted the following in the Rybka forum:

Once again: Rybka is 100% original at the source code level, not counting public-domain snippets like population cnt, etc.

Zach responded with:

Perhaps you could explain to me, then, why:

Rybka's piece square tables are generated from the same code as Fruit's (same KnightRank, etc. constants, but different KnightRankOpening weights)
Rybka's pawn evaluation is virtually identical to Fruit's (different weights again, candidate pawns and backward pawns have a very slightly different formulation)
Rybka's passed pawn evaluation is virtually identical to Fruit's (same bonuses using the quad array {0...,26,77,154,256}, only difference is weights and free_passer split into 3 separate bonuses and based on rank)
Rybka's piece evaluation is virtually identical to Fruit's (different weights only)
Rybka's king shelter evaluation is virtually identical to Fruit's (different weights again, king square generalized to C1, E1, or G1 to store in the pawn table, and a slightly different formula for shelter_file()/storm_file())
Rybka's king safety evaluation is virtually identical to Fruit's (different weights for KingAttackWeight, KingAttackUnit)
Rybka's "pattern" evaluation is virtually identical to Fruit's (different weights, TrappedBishop is not halved for A6/H6).

I have confirmed all of this from reverse engineering Rybka 1, though anyone can see for themselves by looking at Strelka. Rybka's entire evaluation is basically an optimized and tuned bitboard translation of Fruit's, with Fruit's material evaluation replaced by the infamous lookup table. EVERY single evaluation term in Rybka 1, except for the material imbalance table, appears in Fruit. If everyone wants to consider that "original", then computer chess is really dead.

And this is only the evaluation. There are many more similarities...


With so much talent and expertise, I really wish that Zach would focus on making his engine ZCT a real killer. :P
The message from Zach was posted today, December 1st. Those who have access to the Rybka website and who want to follow events can find the thread at:
http://rybkaforum.net/cgi-bin/rybkaforu ... ?tid=13846
Graham, if there is no satisfactory explanation from Vas, are there any significant ways in which your views on the issues being discussed will change?

Alexander Schmidt
Posts: 1086
Joined: Thu May 10, 2007 12:49 pm

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by Alexander Schmidt » Tue Dec 01, 2009 7:42 pm

K I Hyams wrote:Graham, if there is no satisfactory explanation from Vas, are there any significant ways in which your views on the issues being discussed will change?
Not before the United Nations sends peacekeeping forces.

bob
Posts: 20635
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by bob » Tue Dec 01, 2009 7:44 pm

Rolf wrote:
bob wrote: There is very little doubt left about this happening.
If all run with spikes, spikes cant be discriminated as foulish.

That evil doing is no question, and that anonymously evil doing is even worse, this is no question in all Western countries. You wanna disagree or claim absent? Anyway that's the bias.
You cleverly omit the _main_ issue. That you can't improperly reverse-engineer a GPL program..

tano-urayoan
Posts: 638
Joined: Thu Aug 30, 2007 6:23 pm
Location: San Juan, Puerto Rico

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by tano-urayoan » Tue Dec 01, 2009 7:45 pm

Graham Banks wrote:
With so much information gleaned from reverse engineering or disassembling other engines, it might be quicker now. :wink:
Harsh comments from your part.

You quoted his post in the Rybka forum, that most probably will go unanswered. He explicitly stated "similarities" between the two programs, yet you still judge him for showing evidences that people has been asking for.

I think you own him an apology.

bob
Posts: 20635
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by bob » Tue Dec 01, 2009 7:57 pm

Rolf wrote:
bob wrote: You continue to claim that this investigation was unfair because it only targeted one program/author, not all commercial programs and authors. I pointed out that criminals are caught and prosecuted every day, with no requirement that before one can be prosecuted, all must be prosecuted. You simply "catch as catch can" and deal with what/who you catch.
Bob, you didnt quite understand my point. My fault of course. Let me try once more. And please grant me your protection as a mod.

We have always the same debate if it goes about crime, revenge and that IMO a democratic system should do this or that, for all not execute death penalties.

I do NOT contradict because Vas as you think was taken by chance, and not also the others, but I say, again with your protection as mod I am allowed to have the opinion that one should never use undemocratic, unethical methods to "reveil" something other than blackmail, murder or such cases.

I was never your opponent in making fertilizer but I was arguing that such a thought should be published lightly and for all under no stress but in a debate in computerchess circles.

We have no murder case at all. We have Rybka, the best program. By a notable scientist from MIT. My argument is this. You think there is something. You research. All in silence. Not like Donninger. Everythink about using illegal methods are banished in public. Then you come closer to the core. You evaluate, here 10% Fruit if ever, there then the strikingly new stuff that brings the extra power. All original Rajlich. Then you consider: is this now a murder case or should it be a witchhunt after Vas? Are you God? Then you inform your collegues all silently. Guys this Vas is a talent.Rybka is years ahead. He should be told to clean a bit and case closed. NB Fabien had retired and wanted that his code lived on. All programmers could profit but only Vas understood it well enough.

You keep side-stepping the _key_ issue. I do not believe anyone doubts Vas is a gifted programmer. There are enough of them both past and present involved in computer chess. That's not the issue. The issue is simply about copying someone's source code, code that is released under the GPL, which says that if you copy any part of it, you must release _your_ modified version's source code as well, and it, too, will be constrained by the GPL as well.

This was not done. Anyone can look at my code, and look for code copied from other programs. The only code I have used is the magic move generation code that has been used by many others. And the Nalimov EGTB access code that has been used by many others. Not one line of the search, or the evaluation, or anything else has been copied from another program. And that is an easy claim to verify. Most follow that methodology. But not _all_.



So my verdict you reacted in lynch manner without any data about those who might have taken but with less genius. You was driven by obsession against Vas but not because was guilty in any real life sense. Perhaps it was a fault, but that was corrected. Case closed and not such a campaign. You absolutely overreacted.
This was not a "lynching" There was a public trial, with a jury of his peers evaluating his code and comparing it to Fruit. There's no "overreacting". There was simply a search for the truth. I find it _extremely_ unlikely that a new program comes out of the woodwork and is at the top of the heap in a year or two. Even Hsu's first attempt fell flat, and he had copied all of Belle with Ken's permission (Ken worked with him on the project, in fact).

Rybka is not the only such example of "instant stardom". There are others. There will _continue_ to be others.

But Rybka was the one that was caught because of the fruit-rybka-strelka lineage.


Now you are basing on anonymous jerks. You agree with multi-time cloners like Norm. Look, this is ok in extreme crime cases, but not in such computerchess sheneningas.
I am a "never-time cloner". DItto for Zach. He is improving his program by hard work, just as we are doing. Others that helped in the analysis did the same. None were overnight wonders that just burst onto the scene with a world-champion program. So who are the "multi-time cloners" other than norm, who helped us but was not the "primary mover" in the analysis???



You are a free mann, so you can still correct and change your mind. You are Hyatt, not McCarthy, right? Why was this noble social doctor called back and disarmed? Because he was all wrong? Nope. But because the costs for society became too high in all the mistrust.

Bob, dont feed the anonymous jerks. Dont feed blackmailers. Dont feed cloners. But show some respect for a guy like Vasik, who still has the strongest entity available.

For the best of computerchess.

Let' not be Tall Poppy infected. Let's believe in honesty of people from MIT.
If it looks like a duck, walks like a duck, flies like a duck, quacks like a duck, swims like a duck, and even tastes like a duck, it's a damned duck. I'm not willing to say "it's a chicken" and move on. Because then _I_ would be the one making false statements.




P.S. Please add the NOT into the text where I have forgotten. For me as a German it's extremely difficult to adapt to that special trick in your language. Sorry. The attentive reader will discover what was meant.

bob
Posts: 20635
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by bob » Tue Dec 01, 2009 7:59 pm

Milos wrote:
Rolf wrote:My expertise comes from Princeton BTW and it allowed me to see that we have no murder case. Write me a postcard if you disagree.
Again, a lot of childish bragging. You could have a triple PhD from Harvard, Stanford and MIT in law, medicine and chemistry respectively and you would still probably know less about programming then any hacker kid from the neighborhood.
On the other hand you have youths like Zach who probably know more about chess programming then some mathematics PhDs from MIT.
University itself doesn't provide any scientific respect nor a commercial success does it (otherwise we could call B. Gates noticeable scientist :lol:). Papers and works in the field are the things that do that.
There are major cases of plagiarism with distinguished faculty as well, so earning a degree from a prestigious university is no guarantee of ethical behaviour. It should plant the seeds for such, but there's no guarantee they will grow.

User avatar
Graham Banks
Posts: 33216
Joined: Sun Feb 26, 2006 9:52 am
Location: Auckland, NZ

Re: Why the Campaign vs RYBKA is Biased was Re: I am feeling

Post by Graham Banks » Tue Dec 01, 2009 8:02 pm

tano-urayoan wrote:
Graham Banks wrote:
With so much information gleaned from reverse engineering or disassembling other engines, it might be quicker now. :wink:
Harsh comments from your part.

You quoted his post in the Rybka forum, that most probably will go unanswered. He explicitly stated "similarities" between the two programs, yet you still judge him for showing evidences that people has been asking for.

I think you own him an apology.
I'm not accusing Zach of being a dishonest person. Anybody will learn new ideas from studying the source of other engines, just as most of us learn through reading books. That was what I was getting it. That's the whole idea behind open source.

Cheers,
Graham.
My email addresses:
gbanksnz at gmail.com
gbanksnz at yahoo.co.nz

Post Reply