[MUD-Dev] Thief skills, Dream Park and me (again)

[Adam Wiggins]

[Ling:]
> I just dreamt up a very crude way of representing an irregular container
> so I'm gonna send everyone to sleep with it.  Imagine a forest container
> (or room if you prefer).  It looks like this: 
>  
>     123456789
> 
>  1  X ****			// X = container origin/handle
>  2  ******			// * = Forest
>  3   ********   		// very small and cute forest
> 
> The above would represented as a series of horizontal lines and edges:
> 
>   Line 1: 3 - 6
>   Line 2: 1 - 6
>   Line 3: 2 - 9
> 
> Gedit?  Don't know how it'll cope with something that looks like:
> 
>    ***   ****
>    *********
> 
> Then, still in design stages...  any feedback?
> Something I borrowed from graphics.

Sounds like RLE to me, one of those too-simple-to-be-true techniques I've
always loved.  By simply having end-of-line codes in the data you can
easily have the second example you gave; the data looks like so:

1-3, 7-11;
1-10;

Where the semicolon is the end-of-line.


This would work just fine as long as you have a very fixed grid.  I
suppose you could try to define a level-of-detail if your grid is a
bit more dynamic, but this sounds like it would get nasty.
Secondly, it depends on what kind of operations you were planning to do
on the data.  The RLE method is fine for 'there is an object, here, and
it is shaped thus' - which is, of course, why it's used for bitmap
graphics.
The kind of amorphis object-shapes Nathan, JC, myself, and some others
had been chatting about recently is more like a collection of points,
likely other objects.  This means slightly more amorphis (not fixed
to any particular grid other than normal metrics) shapes which may
mutate into completely new forms as the points (objects?) move about,
or as new points are added or deleted.  Also, this handles n-dimensional
situations pretty easily; RLE extended into the third dimension means
something which looks like a voxel shape, which is neither very easy to
deal with nor represent.
I guess the real question is, what kinds of things to you want to *do*
with this data?