On discussing clones

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

Moderators: hgm, Rebel, chrisw

bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: On discussing clones

Post by bob »

Volker Pittlik wrote:
bob wrote:...That is my point. How could I have read it? :) It was deleted within hours of posting...
As Tord, I and the other moderators in charge have told you: there was a side remark in that post that there are rumors that program A is a clone of program B. No more or less. Cause you are also massively complain when posts are edited we deleted the whole post.
I simply state that the best action is "no action". It was not insulting or personal to make that kind of statement. It might start a dialog that leads to proof of cloning or not cloning. There's absolutely nothing in the charter that suggests that deleting the post was the proper action. not even the "legal" clause since this is not a "legal issue" at all...


Without discussion with the other mods I hereby invite you to come to to the mod forum to inspect our discussion about the issue. I'm sure the other will agree and Sam will quickly enable you to do that.

So far I didn't notice any item in the charter that the decision of the mods need your approval or the acceptance of a group unknown to me you are referring to as "we" or "us". If so I strongly suggest to make a clear statement about that.

vp
Quite simple. "we" are the "programmers" that chose to leave r.g.c.c years ago due to Rolf's ultra-high noise level there which made technical discussions impossible, as they got lost in the noise. "we" wanted a place where programmers could discuss computer chess issues, and others that are interested could also come and participate. "we" did not intend for this to become a place where "not-we" types were going to dominate the conversations and control what is acceptable and what is not. The charter is quite simple to read and understand. And current actions with respect to deleting posts are _far_ outside of that. "we" have generally tolerated much off-topic stuff in the past so long as it remains remotely related to chess, whether it be human or computer has not been a particularly strict issue. "we" have moderated personal attacks when they went too far.

Now it seems that (a) any comment that could be interpreted by any single person in the world as "persona' gets removed. Any comment about potential clone issues, particularly when the target is commercial, and _very_ particularly when the topic is Rybka is quickly removed. And then moderators quote the "legal issue" which they apparently have no grasp of (legal problems are criminal problems and there is zero criminal activity with respect to clone discussions). It is a _real_ stretch to imagine any of these discussions could lead to civil actions since we have found and exposed so many true clones over the years that it would be impossible to prove any sort of "intent to do harm" in a libel case.

The best moderation is "no moderatoin". For the past 18 months things have gone to hell in a handbasket. Just look around to see which programmers have left, and which are just too old to want to start over again somewhere else. But even I have my limits... And I don't intend to continue wasting time trying to figure out what was quashed today, what will be quashed tomorrow. When it reaches a point, which is not that far away with respect to me, yet another programmer is going to disappear. And before long, there will be none. If that is the goal, it is easily achievable... It is almost done, in fact...

The moderators should stick to dealing with personal attacks when they get out of hand, and stay the heck out of technical debates that they have no business interfering in, and which they may well be unqualified to deal with in any case...

We have had clone accusations and discussions since this place was started. I have served as a moderator more than any other person that is still here. And I have _never_ deleted a single thread or post dealing with clone issues. Personal attacks? yes. Porn/SPAM? Yes. But that has been _it_.

We did not elect "thought police". We elected people we hoped would simply keep the personal stuff limited to an managable level. Perhaps we once again erred. Or perhaps the moderators simply do not "get it" as to what they are supposed to do and not do...
trojanfoe

Re: On discussing clones

Post by trojanfoe »

The last I heard Zack was finding more evidence to discuss about this particular cloning issue. Without this evidence there is nothing to discuss, or do you think enough has been presented to make a "decision"?

Cheers,
Andy
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: On discussing clones

Post by bob »

trojanfoe wrote:The last I heard Zack was finding more evidence to discuss about this particular cloning issue. Without this evidence there is nothing to discuss, or do you think enough has been presented to make a "decision"?

Cheers,
Andy
A couple of issues:

(1) verifying cloning is a time-consuming process. I've done it several times now and it can waste a day or a month, depending on how clever the cloner was. Some failed to even change output strings in Crafty. Some went much farther. The key question becomes "when does the primary investigator become convinced, and then what will be needed to convince the many naysayers that abound here.

In my case, there is convincing evidence that the original Rybka 1 came in significant parts, directly from Fruit. Too many structural similarities, function name similarities, and then scoring numbers like piece/square tables. There is nothing to gain, IMHO, by looking further. I'm sure, beyond a reasonable doubt, that this happened. That's all I wanted to know. For someone to continue to work to convince _others_ is a more judgemental decision. Is the time spent for a useful undertaking? Will it affect anything in computer chess today? Hard to say. But once I look at enough to reach a conclusion (with respect to the many past clones of Crafty specifically) I stop. There are only so many hours in a day.

2. As previously mentioned, it is possible to write a short C program that will, given enough time, produce Fruit's binary exactly. It might take a few billion years to enumerate enough cases to reach a perfect match, but that shows it is possible to repeat this with some "luck". And this means that those that don't want to be convinced, never will be convinced. Programmers can't even agree exactly on material values for pieces and pawns, so how likely is it two random samples would have the same piece/square table scores? Same odd longjmp() construct, same program structure in the places that were compared? Depends on how convinced you are that you really can win the lottery I suppose.

So, given (2), continuing (1) is a waste of time for Zach and others. IMHO. Nothing will change no matter how many similarities are uncovered. So what would be the point?

Shoot, people are still quoting Rybka NPS and Depth numbers here. :)
terminator

Re: On discussing clones

Post by terminator »

Robert Hyatt, you would not be "singing the same tune" if some internet character like Sam Sloan came to this forum and began attacks on you/Crafty. Neither would you if people began saying, "Crafty has the code in program Y, I haven't seen it but 3 un-named people have done so" in some un-related discussion.
User avatar
michiguel
Posts: 6401
Joined: Thu Mar 09, 2006 8:30 pm
Location: Chicago, Illinois, USA

Re: On discussing clones

Post by michiguel »

Tord Romstad wrote:
Matthias Gemuh wrote:
AdminX wrote:
Tord Romstad wrote: The post presented no evidence or interesting data whatsoever. I don't remember the exact words, but it basically just said that "Program X is a clone of Program Y, according to some people I have spoken to", and nothing more. The two programs in question were named, but not the people who were supposed to have claimed that X was a clone of Y. I don't see any value in such a post. Presenting or discussing evidence that a program is based on another is fine, but starting apparently unfounded rumors about a program is not, particularly when the author of the program is not active on this message board.

Tord
I agree, however the only purpose I can see in such a post was to start another debate.

Ted,

you agree that Tord summerized it adequately ?

Code: Select all

and nothing more.
?
No mention of assembler source code ?
Did Vincent say only the 3 other guys saw assembler source code and he himself not ?
People should be careful when using

Code: Select all

and nothing more.
, as that can be very misleading.
I should probably have added an "if I recall correctly" somewhere: I don't remember any reference to assembler source code, but perhaps there was such a reference.

Nevertheless, even if I do remember wrong, it does not change my opinion in this case. When you don't show the assembly language, or even mention the names of the three guys who are supposed to have seen it, it doesn't count as evidence to me. It is still only spreading baseless rumors, and not a fruitful way to start a discussion. The post contained neither any data nor a place to go looking for data.

I also see the distinction between programmers and non-programmers being discussed in this thread: In my opinion, whether someone is a programmer or a non-programmer has no relevance here. We are talking about a deleted post without any technical content. The opinions of non-programmers are worth just as much as the opinions of programmers.

Tord
Exactly. Besides, when did the programmers become first class citizens? I am asking because I did not receive my passport.

Miguel
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: On discussing clones

Post by bob »

terminator wrote:Robert Hyatt, you would not be "singing the same tune" if some internet character like Sam Sloan came to this forum and began attacks on you/Crafty. Neither would you if people began saying, "Crafty has the code in program Y, I haven't seen it but 3 un-named people have done so" in some un-related discussion.
Been there. Done that. Got the T-shirt. And I survived. I personally doubt that would happen, since Crafty has been "open source" for almost 15 years now and anyone can look at it without the necessity of the reverse-engineering from a binary executable.

But if someone wants to make such a claim, it deserves to be heard and (hopefully) put to rest quickly..
Christopher Conkie
Posts: 6073
Joined: Sat Apr 01, 2006 9:34 pm
Location: Scotland

Re: On discussing clones

Post by Christopher Conkie »

bob wrote:
terminator wrote:Robert Hyatt, you would not be "singing the same tune" if some internet character like Sam Sloan came to this forum and began attacks on you/Crafty. Neither would you if people began saying, "Crafty has the code in program Y, I haven't seen it but 3 un-named people have done so" in some un-related discussion.
Been there. Done that. Got the T-shirt. And I survived. I personally doubt that would happen, since Crafty has been "open source" for almost 15 years now and anyone can look at it without the necessity of the reverse-engineering from a binary executable.

But if someone wants to make such a claim, it deserves to be heard and (hopefully) put to rest quickly..
Sam Sloan? Now there's a lunatic to conjure with.......

:lol:

Funny how Mr Tomato seems to know about that. Now all we must do is determine exactly who HE is, for he is the real clone Bob.

I'd have liked to participate by the way, but not as this thread stands and the way it is developing.

I will just say this one thing.

I agree with you on your initial posting in this thread.
Alexander Schmidt
Posts: 1203
Joined: Thu May 10, 2007 2:49 pm

Re: On discussing clones

Post by Alexander Schmidt »

Volker Pittlik wrote:Deine Niedertracht ist wirklich kaum noch zu überbieten.
Insults in german language are allowed here for all, or only for moderators? Just curious...
User avatar
mclane
Posts: 18754
Joined: Thu Mar 09, 2006 6:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub

Re: On discussing clones

Post by mclane »

how many stopped computerchess because cloned engines using fruit as starting-point cut the market ?
User avatar
Graham Banks
Posts: 41455
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: On discussing clones

Post by Graham Banks »

mclane wrote:how many stopped computerchess because cloned engines using fruit as starting-point cut the market ?
Perhaps they could post here to let us know. :)
gbanksnz at gmail.com