[computer-go] Silver Star 2006 anyone?

[Michał Bażyński]

there was an article where author showed kcc (or one of its earlier 
versions) had hunderds of identical assembly code at times, and showed 
some assembly functions from both programs that did close to nothing and 
were simply unused. to me it was quite convincing.  IIRC he also showed 
how new versions tried to obfuscate the new code to make it about the 
same yet harder to find such long identical parts. i couldn't find it 
right now, but probably people still have the link somewhere...

Chrilly napisał(a):
> ----- Original Message ----- From: "Nick Wedd" <nick at maproom.co.uk>
> To: "computer-go" <computer-go at computer-go.org>
> Sent: Friday, September 08, 2006 1:56 PM
> Subject: Re: [computer-go] Silver Star 2006 anyone?
> 
> 
>> In message <f06110400c1270051f2fe@[193.57.136.156]>, Francois Grieu 
>> <fgrieu at francenet.fr> writes
>>> Any info on the go program "Silver Star 2006"?
>>> It was recently mentionned at
>>> http://uk.theinquirer.net/?article=34230
>>
>> I assume it is a version of the program also known as "KCC Igo", 
>> originally created by disassembling HandTalk.
>>
>> Nick
>> -- 
> 
> How does one know that they have disassembled Handtalk?
> 
> There was a case in computer-chess at the WC in Portoroz (somewhere in 
> the 80ths). The german Mr.Langer directly used a chip of R.Lang and 
> build around his own chess-computer. The R.Lang team proved the cloning 
> by showing that Mr.Langers machine had exactly the same bugs.
> 
> In case of  reverse engineering it is usually not possible to reproduce 
> the original code by 100%. Even with the "best" attempts of the reverser 
> the programm is not bug-compatible. If one transfers. the 
> reverse-engineered programm from Assembly to C its also by 
> reverse-engineering the clone difficult to prove the clonining. I assume 
> nobody has ever dissassembler KCC Igo, because its a terrible work to do 
> so.
> I have my doubts that reverse-engineering pays off. Personally I have 
> reverse-engineered the Shredder and the Rybka search. But nothing worked 
> in Hydra. Same happened for Fruit were the nicely documented code is 
> available. I have also tried to improve Fruit with Hydra ideas. Did not 
> work either. I have learned much more by playing against Shredder and 
> Rybka, reverse engineering was just a nice intellectual challenge. Only 
> the Rybka-team profited from my attempts. I found a bug in their 
> mating-code and sended them the fix.
> Another rule of reverse engineering is: One can only can 
> reverse-engineer something where one knowns how it works. If e.g. a 
> function is called recursively and the arguments are negated and 
> flipped, its alpha and beta, the argument which is decremented by 1 is 
> the search depth......
> 
> Chrilly
> 
> _______________________________________________
> computer-go mailing list
> computer-go at computer-go.org
> http://www.computer-go.org/mailman/listinfo/computer-go/
>