[Home] [Groups] - Message: [Prev in Group] [Next in Group]

nu.kanga.list.mud-dev

9387: [MUD-Dev] Re: mobile movement

[Full Header] [Plain Text]
From: Oliver Jowett <icecube@ihug.co.nz>
Newsgroups: nu.kanga.list.mud-dev
Date: Sat, 16 Jan 1999 03:39:14 +1300 (NZDT)
Organization: Kanga.Nu
On Mon, 11 Jan 1999, J C Lawrence wrote:

> Given the fact that many MUD worlds are topologically a single long
> road with a variety of self contained (ie not or barely
> interconnected) blobs of "areas" hung off it, this approach has some
> really good real-world justifications.  The processing loop then
> becomes a simple (simplified):
> 
>   Is target in same area as me?
>   Yes?  Plot and follow path to target.
>   No?  Plot and follow path to highway.
>   Is target on highway?
>   Yes?  Follow highway in that direction.
>   No?  Plot and follow path to target area.
>   Plot and follow path in target area to target.

It's worth noting that there's no requirement that the highway be linear.
All that is needed is a region of the world that:

 - has easy-to-find connection to the start and end areas
 - allows paths between points within the region to be found more easily
   than paths involving other points, for whatever reason

AR has a 2D map covering the game world, with areas (the diku-like graph
type, with associated pathfinding ugliness) linked in at specific
locations.  Since the map uses a consistent coordinate system, it's
relatively cheap to find a path between two points on the map. The
pathfinding approach we use looks very like the one outlined above,
treating the map as the "highway". 

-O