[DGD]Re:Rooms with views

[Luwenth the Lewd]

	Disclaimer: I'm not hacking on any muds right now, this is all
coming from Computer Graphics and RPG experience, so if I'm wrong, just
blast me. :)

Gregory D Lewis <glewis at maths.adelaide.edu.au> writes:
=> I don't like rooms.  I also don't like coordinates, although the reasons
=> behind that are much harder to explain.  I'm thinking more along the lines
=> of a web of objects, joined by spatial relationships ("a little to the
=> south-east of", "on top of" "a long way north" - I'm inclined to make the
=> distances fuzzy rather than precise measurements, but I'm not sure why).
=
=Surely you need to keep a system of coordinates at least internally?  
=Otherwise you will have difficulty resolving things such as "can I move
=from my current location to location B" or "can I move from my current
=location x units of distance south" or "can I see object foo from my
=current location".

	Why would you need to keep a coordinate system internally?  From
any point in the mud, if you pick the data set of possible movements up
from that particular room, you're going to get a very ugly 3-dimensional
tree sort of shape.  From room X, there could be 1 exit or 24 exits, some
of which may be uni-directional (trapdoor, teleport device, shredder
to rejuvination clinic, chute) or 3-directional (walking through
the door one way puts in you room Y, walking back through puts you in Z) or
if you want to drive players truly insane, the exits rotate whenever one of
them is used, or when all are closed or something along those lines.  
	"Can I move..." is going to end up being a search, hopefuly 
breadth-first, of the possible single-direction hops.  In my mental mudlib, 
one of the things that I was considering was a 'smart helper' which you 
could say 'take me to foo' and it would know the absolute shortest route 
to get there.  Of course, it would also sometime misunderstand where you 
wanted to go, but hey.
	"Can I see..." is going to be even more difficult and is the one
that I'll gripe the most about.  Image a set of three rooms, you're
standing in X, looking towards Y, with Z on the far side of Y.  If there is
a large object in between the entrances from X and to Z in Y, then
you're not going to be able to see squat!  As well as if the door from X is
in one corner, the door to Z is in the other corner, and you're standing in
the middle of X.  Line of sight can only be done with pure geometric
mapping of the entire mud including where characters are standing, and then
doing some form of ray-trace to see what should be described, and how much.
If you can just see the leg of a table, are you going to know if it's a
table, chair, or bed?

=There have been some lengthy discussions on these sorts of things on the
=MUD-Dev list over the years (check the mailing list archives on kanga.nu).

	I've probably just restated some of the problems that they raised
with trying to do line of sight, though I've not read that discussion at
all.
	-- Ricky

--
'94 ST1100 'Turvy'				luwenth at netscum.com
      "I don't know what's wrong with people: they don't learn by 
       understanding; but by some other way -- by rote, or something.  
       Their knowledge is so fragile!"  -- Richard Feynman

List config page:  http://list.imaginary.com/mailman/listinfo/dgd