Complete.Org: Mailing Lists: Archives: freeciv-dev: April 2004:
[Freeciv-Dev] Re: (PR#8474) clean up road_bonus()
Home

[Freeciv-Dev] Re: (PR#8474) clean up road_bonus()

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: use_less@xxxxxxxxxxx
Subject: [Freeciv-Dev] Re: (PR#8474) clean up road_bonus()
From: "rwetmore@xxxxxxxxxxxx" <rwetmore@xxxxxxxxxxxx>
Date: Thu, 8 Apr 2004 07:26:15 -0700
Reply-to: rt@xxxxxxxxxxx

<URL: http://rt.freeciv.org/Ticket/Display.html?id=8474 >


Per I. Mathisen wrote:
> <URL: http://rt.freeciv.org/Ticket/Display.html?id=8474 >
> 
> On Thu, 8 Apr 2004, James Canete wrote:
[...]
>>should the AI think about connecting up its cities directly?
> 
> The argument against is that it makes it easier for enemies to conquer the
> AI. Especially if we extend it to rail. We should at least look for ways
> to protect the AI against quick conquest along our road/railways, since it
> tends to keep cities unprotected when there are no enemies nearby.
> 
>   - Per

The AI needs to develop a concept of "regional closeness" that is an 
extension of adjacency or distance in terms of movement points. The 
PathFinding maps allow one to compute such distances by lookup.

What might help is to build and cache such a map with "optimistic" 
non-dynamic assumptions. One then finds and moves units to "choke points" 
which cause the move cost to change significantly.

Alternatively, these are the points where pillaging a railroad or building a 
city/manned fort would break the network into regional subsystems.


The algorithm for emptying cities can then easily understand the level or 
depth of regional layering of the empire, i.e. what is central vs peripheral 
on the basis of unit movement vs pure geographic distance. This will also be 
a better system for the AI when one moves to maps where adjacency is 
determined by linkage (i.e. through portals), and not strict physical position.


Cheers,
RossW
=====




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