Complete.Org: Mailing Lists: Archives: freeciv-dev: July 1999:
Re: [Freeciv-Dev] pollution and ai (PR#82)
Home

Re: [Freeciv-Dev] pollution and ai (PR#82)

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: Reinier Post <rp@xxxxxxxxxx>
Cc: neff0003@xxxxxxxxxxxxxxxxxxx, freeciv-dev@xxxxxxxxxxx, bugs@xxxxxxxxxxxxxxxxxxx
Subject: Re: [Freeciv-Dev] pollution and ai (PR#82)
From: Tony Stuckey <stuckey@xxxxxxxxxxxxxxxxx>
Date: Thu, 29 Jul 1999 12:57:41 -0500

On Thu, Jul 29, 1999 at 07:04:23PM +0200, Reinier Post wrote:
> > I'm not sure if this is a bug or "feature"
> > But when I play against the ai, they don't seem to clean up their pollution
> > problems.
> 
> This is a known problem.

        They clean up some of them.

> > be coming from the ai.  Is there anything I am missing or any patch?
> 
> Not that I know of.  The AI code should value pollution cleaning more.
> Probably a matter of tweaking the heuristics in auto_settler_findwork():
> 
>   http://www.freeciv.org/lxr/ident?i=auto_settler_findwork&v=cvs

        The AI values pollution based on the value of the polluted square.
If an unproductive square, like a Tundra, Desert, etc becomes polluted, the
AI will ignore it.
        We need a valuation which is less subjective.  Pollution in PayCiv
has been -10 points per polluted square, so I have been testing a formula
based on that.  It works much better.

        The real trouble is that the relative weights of trade, production,
etc, come from hard-coded constants.  Thus the AI values squares that
produce more shields over everything else when choosing squares to utilize
and terraform.  This is quite brain-dead.  The weights need to be adaptive.
-- 
Anthony J. Stuckey                              stuckey@xxxxxxxxxxxxxxxxx
"When I was young, the sky was filled with stars.
 I watched them burn out one by one."  -Warren Zevon

[Prev in Thread] Current Thread [Next in Thread]