Re: A Crime by a Board of Old Imposters

darrz <[EMAIL PROTECTED]> wrote:
> On Tue, 02 Dec 2003 00:27:35 GMT, "Znarf" <[EMAIL PROTECTED]> wrote:
>>The ICGA could only access and review the source code upon an accusation of
>>"cheating."  The Accuser must provide a factual basis for the accusation
>>(e.g., it must be more than "The accused program performed the same two
>>moves as program x.").  This would protect entrants from unfounded
>>accusations.

> Not really. Read your statement again. Where are the facts in the
> accusation? What facts can you have as a cc tournament entrant, about
> another program?

> You have no source code, no performance of the suspect program on any
> critical positions or test suites. Nothing!

> All you have is a suspicion, based on the play you see the suspect
> program make at the tournament WHICH IS ALREADY IN PROGRESS.

> Are you going to stop the whole tournament and examine the suspect
> program for a few hours? Based on someone's _suspicions_ ??

> I don't think there is a good answer anywhere in this rat hole!

> You either run the risk of someone cheating and winning a tournament
> with a program like Crafty, or you disrupt the tournament and make
> unfounded accusations.

> I have another idea, which may not be popular, but here goes:

> In an open tournament, any program competes. Bring your open source
> program, modified or not. You're in like a porch climber!

> Open source authors, like Dr. Hyatt, would just need to keep their
> strongest version under wraps until after that year's championship.

> It seem's odd that someone would make bikes, give them away freely,
> and then say "but you can't race with this bike, against me".

the problem with that logic is that the next event will have 1000+
participants, and a Swiss big enough to find the best "program"
will take a long while.  I don't see the use in allowing 100 modified
versions of a single program to participate.  How would you host an
event with 1000 programs?  Where do you find 1000 computers to run it
on?  Where do you put 'em?

just having 40 is a big enough problem.


> Phooey! Every "bike" can "race" in an open tournament. Every program
> can compete with different settings or "personalities" that Joe or
> Jill Consumer have discovered, and care to enter into the open
> tournament.

> If you (as a chess programmer) make your program open source, you need
> to make one version stronger than the source you released, and keep it
> confidential, to compete with an edge in the next open cc tournament.

> Result?

> Lots of people make lots of little modifications to Crafty and other
> open source chess programs. Most will be bad, but some will be good. 
> More competitors, more interest in cc, and any good modifications that
> prove valuable may expand our knowledge of what makes a stronger chess
> program.

> Note that I'm not saying anyone has the right to take open source code
> and sell it commercially. I'm only saying for an open tournament,
> every program, of every type, should be allowed to compete.

> That clears the air of any claims of cheating, and is in the true
> spirit of an open tournament. IMO


What about pure copies?  No changes?  What would the point be in
that participating?


> It also makes computer chess more interesting, to more people. 

> If current open source chess program authors are offended by that, I
> believe they should not make their program open source, anymore, and
> really ask themselves "why did I make this program open source?".


> Your thoughts?


> Darrz



-- 
Robert M. Hyatt, Ph.D.          Computer and Information Sciences
[EMAIL PROTECTED]                   University of Alabama at Birmingham
(205) 934-2213                  136A Campbell Hall
(205) 934-5473 FAX              Birmingham, AL 35294-1170