Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Any idea's on how long it takes to learn C++, then create a chessmonster

Author: Gareth McCaughan

Date: 18:08:40 12/11/01

Go up one level in this thread


On December 11, 2001 at 18:22:27, Dann Corbit wrote:

>On December 11, 2001 at 18:01:27, Gareth McCaughan wrote:
>
>>On December 11, 2001 at 17:07:17, David Hanley wrote:
>>
>>> Curious : how many NPS are you getting out of your python program?
>>> My lisp program is getting 3,500NPS on a pentium 200 with zero
>>> optimizations.  I think a faster computer and optimizations would
>>> increase that tenfold, at least.
>>
>>How many NPS? Oh, *several*. :-) Seriously: between about 30 and
>>about 100, on a 1GHz Athlon. It would probably be possible to get
>>another factor of 10 without much pain; the implementation is
>>very inefficient at the moment. (Remember it's intended to give
>>a fairly level game to a 5-year-old.) If I count all nodes
>>encountered in the qsearch, the level is more consistent; usually
>>80-100NPS.
>>
>>I realise that I did the program a disservice when I described
>>its features. It does a few simple search extensions too.
>
>I'd like to see the source code, if you allow that.  I'm pretty curious as to
>what it looks like.

It looks like a nasty weekend hack. There's a reason for this. :-)

I'm not keen on distributing the source any further than necessary,
for fear of embarrassment. Some day I might clean it up, but I'm
more likely to write something entirely from scratch, in a language
better suited to chess programs.

However: yes, you can see the source code. Is the e-mail address
at solutionsiq.com a good place to send it?

--
g



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.