Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fafis 2.6

Author: Mario Antonio F.

Date: 09:58:50 01/17/06

Go up one level in this thread


Hello,

I'd like to write this note not as a pure response of your comment, but as a
starting point of thought (not necessarily discussion).

I have also read the rumour that Fafis is a clone, but I have not read any proof
of this.

Stating that a program is a clone should be considered as a very serious matter.

First, we need to define clone. A clone chess engine is a copy of an existing
engine with minor or no modifications.

I do not consider an engine to be a clone if it uses the common Nalimov end game
tables, or whether it uses Null Move (Feist?) or whether it uses MTD(f) (Plaat),
bitboards, parallel search (Hyatt), etc. Actually a lot of algorithms have been
kindly provided by authors (Schroeder and many others).

It will be interesting how it was determined that Fafis or any others are
clones. If someone states the engine is a clone, it is necessary to probe the
point.

Now, the point is not easy to do since all we have is a binary file. It may be
possible to extract function signatures from the code and state that some
functions have the same signature than another engine.

It may also be possible to write a very complex program that checks for similar
'chunks' of memory between programs and cross fingers that both authors used the
same compiler [including version and flags] and libraries.

Like others, I do not agree that an engine clones another without stating it
somehow. On the other hand, it is not possible to reject an engine because of a
rumour, which affects a good intentioned author.

If anyone has any solid information on this cloning statement, please publish
it.

Thank you,

Mario Antonio


On January 17, 2006 at 03:47:11, Graham Banks wrote:

>On January 17, 2006 at 02:39:39, Petr Dobes wrote:
>
>>Hello,
>>does anyone know  informations about engine Fafis 2.6? Download link, if is it
>>free...?
>>   Thanks Peter
>
>A
>s far as I'm aware, Fafis was exposed as a clone. The Exactachess website seems
>to confirm this.
>
>Regards, Graham.



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.