Spaces or rooms? (Re: [MUD-Dev] Information sharing (was: Re: Where are we now?))

[Koster]

> -----Original Message-----
> From: Ola Fosheim Gr=F8stad

> "Koster, Raph" wrote:
 
>> presentation was stuff that is publicly available raw data:
>> room-based maps versus continuous maps, for example, is something
>> that just about anyone on this list could identify as an important
>> design difference. But when I went looking for analysis of mud
>> design, I found that nobody had even classified those two very
>> basic structures.

> Uhm... I don't believe that is right.  I believe several have
> pointed this out.  Anyway, for a rather (overly) thorough look at
> the room-based one, look for Andreas Dieberger's Phd thesis.

Let me rephrase; "for game design purposes." There's been quite a lot
of academic analysis of the subject.

> If I was to write on this today, I probably would make the
> distinction between graph-structured, (segmented and nonsegmented)
> euclidean space and "non-spatial/generated" dynamic environments.

I don't really understand the logical grouping of "non-spatial" and
"generated"--they don't seem to have anything to do with one
another. And I don't know of a generation scheme that isn't
graph-structured in some manner. Unless you are defining graph more
narrowly than I am...

For that matter, I defined online worlds as using a spatial metaphor,
so non-spatial in my terminology doesn't even exist. :)

> Although I think it is much more complicated than that, because even
> in eculidean space you have rooms, and you have the field of vision
> which forms a room, you have the auras in DIVE which are rooms, you
> have the sliding window over the landscape in Alphaworld (and other
> 3D worlds) which is some sort of (annoying) room. Not to mention the
> parallell chat/action "rooms" that comes with team play and such (I
> prefer to abstract those as "spheres" in my own thinking on the
> subject)...

Well, I defined a room as a node, a database entry describing a
location, with logical links to other database entries; essentially
the same as a given block of a hypertext (in a text mud, they
literally are). By convention, the logical links are labelled or
represented with spatial relationships though these may be absolute or
relative. Since these are logical links, you can build non-Euclidean
spaces quite easily.

Methods of culling what is displayed seemed to be distinct from the
map structure; in the case of a room-based system the cull mechanism
and the map structure happen to be coincident, but in a continuous map
structure is is not the case. And when you get into continuous maps
embedded in rooms, or vice versa, attaching significance to the method
of culling starts getting silly. Consider that in a 3d environment,
there's actually two kinds of culling going on--network traffic, which
is NOT culled to the "room" created by the view frustrum; and visual
culling, which is.

Anyway, I don't see thinking of that as a room as being very fruitful.

On the spheres... I am not entirely sure what you are defining.

-Raph

_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev