Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Help Elucidating Fruit 2.1 Source Code

Author: Steve Maughan

Date: 08:25:32 11/17/05

Go up one level in this thread


Michael,

I've had a look at Fruit's code particulalry version 1.5, but also v 2.0 and
2.1.  Here's my conclusions as to why it's strong:

1.  Few bugs.  Fabian uses many "asserts" and is clearly focusing on reducing
the number of bugs.  This is probably the #1 reason that Fabian has been able to
create a strong program so quickly

2.  Solid search.  The search is really quite straight-forward with nothing
startlingly special.  It does contain all of the basics e.g. null move, ETC,
iterative deepening, history pruning.  Fabian hasn't tried to be too fancy with
the implementations of these techniques and his programming style is ultra clean
which also helps tremendously.

3. Evaluation function that covers all of the basics terms.  Again it doesn't
contain anything too special just the basics.

I think overall you could say that Fabian has a knack of implementing the most
important aspects of a chess program without over complicating the code.

Regards,

Steve



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.