Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Move Ordering and Transpostion Tables -- A Question

Author: Erik Evenson

Date: 10:43:25 08/10/01

Go up one level in this thread


Interesting.

This and another reply seems to indicate that I need to develop some additional
logic in my search routine to look at a less than full movelist first (for
cutoffs, principal vars, etc.) and then, if necessary, look at a full movelist.
It also implies a refinement to my move data structure -- something to store the
evaluation type (actual, fail high, fail low).

Makes sense, and keeps me from storing _every_ move possible from a given
position.  Adds a loop to my code, but nothing overly onerous.

As I continue to refine my code, I continually seem to find myself putting
refinements into all of my code to improve a single part of it.  In retrospect,
this is not surprising, but I shudder to think what this would mean if I were
using a language other than Java (I know, too slow).  With Java, I can quickly
refactor the code, but with something else, I gotta believe this would be
painful.

In fact, since I know that I will eventually want to port my code (or critical
portions of it) to a native code set, I'm thinking getting the logic working in
Java first may be a dominant strategy for someone like me -- short of time and
certainly short of non-Java programming skill....

Anyone know of a Java-to C or C++ translator?  :)

Thanks a lot!

Erik



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.