Progess (often) is ugly, a summary try

General discussion about computer chess...
User avatar
Rebel
Posts: 515
Joined: Wed Jun 09, 2010 7:45 pm
Real Name: Ed Schroder

Re: Progess (often) is ugly, a summary try

Post by Rebel » Mon Jul 04, 2011 7:39 am

hyatt wrote:
Rebel wrote:
hyatt wrote:There is no "legacy" when you steal something, use the money for good, then get caught.
So now you are going to remove all the stolen Rybka idea's in Crafty ?
Please feel free to point out _one_ block of stolen code from Rybka that is in Crafty. Then we will talk...
You should not twist my words, the issue here is the stolen legacy of Vas, the stolen ideas good for 300-400 elo points. If you are going to argue the big elo jumps Crafty made had nothing to do with the stolen idea's from Vas then it's more likely the Earth is flat after all.
Any others you want to mention???
Oh absolutely. Any programmer who spoke the verdict "guilty" at the tribunal and at the same time has profited from the stolen ideas of Rybka is a fence.

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: Progess (often) is ugly, a summary try

Post by Jeremy Bernstein » Mon Jul 04, 2011 7:47 am

Rebel wrote:
hyatt wrote:
Rebel wrote:
hyatt wrote:There is no "legacy" when you steal something, use the money for good, then get caught.
So now you are going to remove all the stolen Rybka idea's in Crafty ?
Please feel free to point out _one_ block of stolen code from Rybka that is in Crafty. Then we will talk...
You should not twist my words, the issue here is the stolen legacy of Vas, the stolen ideas good for 300-400 elo points. If you are going to argue the big elo jumps Crafty made had nothing to do with the stolen idea's from Vas then it's more likely the Earth is flat after all.
Any others you want to mention???
Oh absolutely. Any programmer who spoke the verdict "guilty" at the tribunal and at the same time has profited from the stolen ideas of Rybka is a fence.
There is no such thing as a stolen idea, just stolen code. You know that, as well, so I'm not sure what your point is. That other chess authors learned from Vas' improvements and implemented similar in their programs? That's the normal state of affairs. "Guilty of plagiarism" doesn't mean "free of originality". So what is your point?

jb

Gerd Isenberg
Posts: 37
Joined: Wed Jul 07, 2010 11:11 pm
Real Name: Gerd Isenberg

Re: Progess (often) is ugly, a summary try

Post by Gerd Isenberg » Mon Jul 04, 2011 7:56 am

I share your critique concerning the life-time ban, the media coverage, and that Chris W. was banned from the panel.

Gerd

User avatar
Matthias Gemuh
Posts: 295
Joined: Wed Jun 09, 2010 2:48 pm
Contact:

Re: Progess (often) is ugly, a summary try

Post by Matthias Gemuh » Mon Jul 04, 2011 7:59 am

Rebel wrote: You should not twist my words, the issue here is the stolen legacy of Vas, the stolen ideas good for 300-400 elo points. If you are going to argue the big elo jumps Crafty made had nothing to do with the stolen idea's from Vas then it's more likely the Earth is flat after all.
Has _anybody_ been condemned for stealing ideas ?
Please, name the Rybka ideas that boosted Crafty (though nobody condemns the "stealing" of ideas).

Matthias.
Aided by engines, GMs can be very strong.
http://www.hylogic.de

hyatt
Posts: 1242
Joined: Thu Jun 10, 2010 2:13 am
Real Name: Bob Hyatt (Robert M. Hyatt)
Location: University of Alabama at Birmingham
Contact:

Re: Progess (often) is ugly, a summary try

Post by hyatt » Mon Jul 04, 2011 8:03 am

Apparently that there would be some difference if I used an idea that Vas proposed, even though Vas used many of my ideas (rotated bitboards, other pieces of source code, etc.) not to mention other ideas from other authors, going all the way back to alpha/beta and minimax.

But in any case, I stand by my statement that we have yet to look at _any_ ideas from the various ip*/robo* programs. The only idea I even looked at was the TT-singular thing. I found it didn't work for me. I removed it from stockfish and it seemed to make no difference there. At much longer time controls, it did eventually seem to help stockfish a tiny amount. But it is not in Crafty anywhere. We do a check evasion extension and nothing else.

hyatt
Posts: 1242
Joined: Thu Jun 10, 2010 2:13 am
Real Name: Bob Hyatt (Robert M. Hyatt)
Location: University of Alabama at Birmingham
Contact:

Re: Progess (often) is ugly, a summary try

Post by hyatt » Mon Jul 04, 2011 8:07 am

Gerd Isenberg wrote:I share your critique concerning the life-time ban, the media coverage, and that Chris W. was banned from the panel.

Gerd

Just for the record, Chris applied, there was a 1-2 day lag as each applicant was screened, we did not want anonymous people, nor people with no technical background. During that lag in admission, Chris did his usual and started his complaining. The group generally decided that they could do without his noise. I did not vote to deny him admission, but then I did not object due to the usual flare-up while he was waiting. And in retrospect, it was a + for the panel, Chris had already made a zillion bogus arguments about why vas "might not have actually copied" fruit code, even though they are identical. Those discussions are still available. Might have been here, am not sure, or at CCC perhaps...

It would have been about as productive to include Rolf. Wonder if some of you guys would have criticized us had he applied and been turned down???

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: Progess (often) is ugly, a summary try

Post by Jeremy Bernstein » Mon Jul 04, 2011 8:17 am

hyatt wrote:
Gerd Isenberg wrote:I share your critique concerning the life-time ban, the media coverage, and that Chris W. was banned from the panel.

Gerd

Just for the record, Chris applied, there was a 1-2 day lag as each applicant was screened, we did not want anonymous people, nor people with no technical background. During that lag in admission, Chris did his usual and started his complaining. The group generally decided that they could do without his noise. I did not vote to deny him admission, but then I did not object due to the usual flare-up while he was waiting. And in retrospect, it was a + for the panel, Chris had already made a zillion bogus arguments about why vas "might not have actually copied" fruit code, even though they are identical. Those discussions are still available. Might have been here, am not sure, or at CCC perhaps...

It would have been about as productive to include Rolf. Wonder if some of you guys would have criticized us had he applied and been turned down???
Except that Rolf is a troll with no technical background whatsoever, whereas Chris is a respected chess engine programmer with years of experience and plenty of technical background. No matter how annoying he is. As a secretariat member, your role was, in fact, to determine which of the panel members had "bogus" arguments before passing along your recommendation to the Board, not to filter the panel ahead of time and prevent unwanted, "unproductive" input.

EDIT: the result, in this case, is unlikely to have been affected, given the weight of the technical evidence. But in a process like this, it's kind of important to demonstrate a willingness to hear all points of view.

Jeremy

User avatar
Rebel
Posts: 515
Joined: Wed Jun 09, 2010 7:45 pm
Real Name: Ed Schroder

Re: Progess (often) is ugly, a summary try

Post by Rebel » Mon Jul 04, 2011 8:23 am

Jeremy Bernstein wrote:There is no such thing as a stolen idea, just stolen code. You know that, as well, so I'm not sure what your point is. That other chess authors learned from Vas' improvements and implemented similar in their programs? That's the normal state of affairs. "Guilty of plagiarism" doesn't mean "free of originality". So what is your point?
Rybka was re-engineered into readable source code, with that it was easy to extract the basic idea's of Rybka's superiority and thus the implementation in other engines began. We have seen all the spectacular elo jumps. The hacker being hacked. So in one hand Vas has been rightfully convicted, on the other hand there are programmers doing exactly the same, using hacked stuff that is not theirs and never was intended to be theirs.

If that is not a moral dilemma then my mother did not raise me well.

PS, notice the subject of the thread.

Gerd Isenberg
Posts: 37
Joined: Wed Jul 07, 2010 11:11 pm
Real Name: Gerd Isenberg

Re: Progess (often) is ugly, a summary try

Post by Gerd Isenberg » Mon Jul 04, 2011 8:26 am

hyatt wrote:
Gerd Isenberg wrote:I share your critique concerning the life-time ban, the media coverage, and that Chris W. was banned from the panel.

Gerd

Just for the record, Chris applied, there was a 1-2 day lag as each applicant was screened, we did not want anonymous people, nor people with no technical background. During that lag in admission, Chris did his usual and started his complaining. The group generally decided that they could do without his noise. I did not vote to deny him admission, but then I did not object due to the usual flare-up while he was waiting. And in retrospect, it was a + for the panel, Chris had already made a zillion bogus arguments about why vas "might not have actually copied" fruit code, even though they are identical. Those discussions are still available. Might have been here, am not sure, or at CCC perhaps...

It would have been about as productive to include Rolf. Wonder if some of you guys would have criticized us had he applied and been turned down???
You can not compare Chris with Rolf on that matter. Chris was commercial chess programmer, and played multiple ICGA tournaments. I am more or less aware of Chris' agenda with the ICGA. Nevertheless, he would be a valuable "devils advocate", which was otherwise appreciated by Mark Lefter directed to other panel members.

For the media coverage, the ICGA should have published the disqualification first on their own side or the investigation wiki rather than on ChessVibes.

Beside these points and the life-time ban, which is too hard for my taste, I am in agreement with the ICGA decision.

BB+
Posts: 1484
Joined: Thu Jun 10, 2010 4:26 am

Re: Progess (often) is ugly, a summary try

Post by BB+ » Mon Jul 04, 2011 8:27 am

Briefly, I agree the "lifetime ban" is a bit of salesmanship, as the Rajlich-ICGA feelings are likely mutual. And presumably Rajlich could ask for reconsideration in any event [which as likely as ...]. The main aspect of it might be to make other organising bodies somewhat less likely to accept a Rajlich entry.

Regarding the mainstream media, having been involved in various attempts to popularise maths research, I can say it is almost impossible to guess what direction they will go, once a ball starts rolling [in whatever direction]. I have no idea if the ICGA was a major instigator or not (beyond ChessVibes, which seems decent to me). If so, I agree that it has mushroomed beyond what I think to be the due proportion.

Concerning Chris Whittington, I exchanged a few long emails with him, and also intervened with the Secretariat on his behalf at one point. I can discuss the issue more if desired, but the main points are: he was overly rude regarding a request for identity check [his email was not the one the ICGA on record, he doesn't have a fixed landline (IIRC), and most notably he used the phrase "little Hitlers" regarding the identity verification process], and it was unclear whether he would be civil in the discussions (there was also some issue that Bob replied erroneously to him regarding an internal UAB email [with a "groan" type comment], and CW felt the "groan" was actually directed at him, and Bob was making a cover-up excuse).

Secondly, I received some assurance from Levy and the Secretariat that if Rajlich was going to join in the discussions of the Investigation Panel, then everyone (in particular CW) who was "known" to be favouring his side of things would be admitted. OTOH, if the Panel was going to proceed without input from VR [and so, he would speak directly Board, if at all], then I had to agree that there didn't seem any specific reason to admit CW under the immediate circumstances. Finally, the "ban" by the Secretariat on CW was only for a fixed amount of time (either 2 weeks or a month), at which point he could apply again. I communicated this temporal aspect to him at the start, but by the time it rolled around he seems to have lost interest (he was more concerned about whether some sort of deal could be worked out with FL and VR, so as to avoid tearing the already-broken CC community apart).

As I say, the process was imperfect in many regards, but I can't say that it was particularly unfair.

Post Reply