Computer Chess Club Archives




Subject: Re: Fascinating Software Tools Idea!

Author: jefkaan

Date: 09:31:31 11/20/02

Go up one level in this thread

On November 20, 2002 at 10:36:36, Bob Durrett wrote:

>Would you care to elaborate about that higher level programming language?

sure, these would be 'scripting languages, see eg:
besides Tcl (also used for graphical interface for
the freeware chess database SCID) there's Python or
Ruby and so on ; but to program something like a chess search
algorithm would make a terribly slow exec,  see:
However maybe it would be possible to set up such
higher level programs in a modular way, using 'lower level'
modules for search etc.; ideally you then could use move
generators etc. as compiled in C, and expand chess knowledge
(evaluation) on a higher level; problem imho is that the
eval comes at the end of the qsearch which as a result
would again make the whole program incredibly slow..
>Would the compiler for that language produce machine instructions, or some
>intermediate product?

machine intructions as well, but in an object oriented way
so it could use/interact with other modules; with Ruby for simple
programs its also possible to convert the source back to C,
after which you could have a look to it and compile it yourself
with a C compiler; quite frankly i don't have experience with
such higher level languages but i prefer doing a proper feasibility
study before re-inventing the weel with eg. bitboards etc :)
See for example:
Another question to make things real confusing is whether you
really like or need object oriented programming or the
more oldfashioned  procedural  or table oriented languages :)
(personally i think i still prefer procedural languages,
so i dont know what tool is the best for computerchess and
whether its available, except things like Delphi or BCB).
So probably for chess its ok to stick to MS and use
Visual Studio C++ or so, as there is an excellent new
graphical interfaces for chess engines namely Arena.
But it also depend on just what you tend to like, or
what other applications you build in future, even
when you would prefer to stick to goodoldfashioned
Fortran :) See for example:
But i'm curious what others, and you,  have to say
about this issue, and hopefully with some
brainstorming we both at least could see some
lights at the end of some of these tunnels.. :)

This page took 0.03 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

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