[MUD-Dev] Re: Room descriptions

[Koster]

> -----Original Message-----
> From: Adam Wiggins [mailto:adam at angel.com]
> Sent: Monday, September 28, 1998 6:35 PM
> To: 'mud-dev at kanga.nu'
> Subject: [MUD-Dev] Re: Room descriptions
> 
> 
> On Mon, 28 Sep 1998, Koster, Raph wrote:
> > > From: Adam Wiggins [mailto:adam at angel.com]
> > > On Mon, 28 Sep 1998, Koster, Raph wrote:
> > > > I would
> > > > not call the goal of having a fully interactive environment even
> > > > necessarily contradictory to what I am discussing. You can 
> > > > have a fully
> > > > interactive environment that still imposes a perspective or 
> > > > a worldview
> > > > on the player, I would think. 
> > > 
> > > On this particular point I don't think I can agree.  In 
> the end, if
> > > someone detonates a nuclear bomb inside your candy store it's 
> > > still going to have the same roomdesc.  I don't think 
> this sort of 
> > > constructed world
> > > will ever be interactive in the way that I'm thinking; it 
> will always
> > > be a sophisticated Zork spinoff, which is to say you only have the
> > > interactions which are coded by whoever created the area.
> > 
> > Yes, well, this is the case for all the interactive world 
> interactions
> > you might have, too.
> 
> I don't agree, precisely, and I'm surprised you would say 
> this, considering
> your work on Ultima Online.

Maybe I learned something about exactly how much it takes to get to
where you're going. ;) Seriously, though, what we learned with UO was
that you can provide a billion little interactions, and you're still not
gonna get to that full simulationist world--and that once you get there,
it may not be fun.

> If I write a very specific area, typical of most muds, where there are
> highly specialized room descriptions and most of the interactions are
> coded to be specific to that area, any functionality you don't cover
> will not work in your area. 

Correct. I am not saying this is a good idea. :)

>  If we build the candy store via typical
> methods - writing room descs, adding extra descs, adding little spec
> procs to handle different things the player might do - it can't deal
> with something not coded, including extreme effects like the atom
> bomb, but also effects like someone rearranging the furniture, or
> killing the candy store owner and taking his clothes and masquerading
> as him, or painting all the walls black and installing a 
> movie projector
> and turning it into a movie hall, or even something simple 
> like dropping
> a rotting carcass into the center of the floor (which would certainly
> draw attention away from all the normal decorations).

All the examples above except the painted walls and movie projector are
possible in UO; nonetheless, it's gonna feel like a static environment
anyway. The question is how many levels of sim are you gonna toss in to
make it feel worthwhile.

> If, however, the candy store is built the way Orion suggests 
> - by actually
> creating each piece of furniture, all the candy stocks, the 
> guy behind the
> counter, a full uniform for him, big window panes, and a 
> shiny tile floor,
> all the interactions the game supports are automatically 
> enabled for that
> room.  You can now drop an atom bomb in it, or kick over the 
> tables, or
> track mud all over the floor, or turn the room into an animal 
> shelter, and
> the room description will never not make sense.

You can only drop an atom bomb in it if you have the behavior coded for
atom bombs across the board. (And yes, there are many good ways to do
this so that it requires only a single databse entry and some simple
code--I am not saying that it's difficult to do). Do you support mud
tracking in your mud? How do you handle mud (on the floor) erosion over
time? How do floors get cleaned? Does the mud (on the floor) serve any
purpose?

"all the interactions the game supports" is my point... 

One of the lessons we learned with UO (which supports a LOT of
interactivity, with *billions* of discrete objects with individual
functionality, individually tracked statistics, etc) is that it's still
not enough to make it feel interactive. You might have mud tracking but
not wood polish; you might be able to burn a painting but not slash it,
or throw turpentine on it...

> One way you get the ultimate in flexiblity and freedom to 
> create a specific
> feeling and environment, at the expense of difficult to add or change
> interactivity. 

Correct.

> In the other case you loose the builder flexbility and
> probably a lot of mood, but you get the ultimate in dyanmicly 
> built worlds,
> which can truly be torn down to nothing and then built up 
> again if there
> are players willing to do it.  I'm not suggesting that one is better,
> just attempting to point out the differences...

Players will be willing to tear it down, certainly. Build it up, I don't
know. ;) You tend to get more people who have fun by destroying than
those who have fun by building. :(

> > If you leave out atom bombs from your code, after
> > all, there's no way one is going to be detonated in the 
> candy store. :)
> 
> Sure, this is the old "Doc, it hurts when I do this"/"Well 
> don't do that!"
> fixit. 

I wasn't suggesting it as a fix for the problem; I was saying that this
is what WILL happen to any simulationist enterprise. How to state this
clearly?

The Simulationist's Dilemma: "One simulates a world to provide greater
interactivity and submersion in the fictional environment. But
simulating a world can never reach the level of detail that reality can.
And the more realistic the simulated environment, the more jarring a
break in the fiction is. Therefore, the further you go into realism, the
faster and harder you will fall out of immersion."

This is not to say that the simulationist's enterprise is doomed; it's
just that I don't know that it generally achieves what you want even in
the designs being discussed. Maybe someday someone will actually do it
well enough. :) A noble goal, to be sure. A fundamental question is
whether many of those physical interactions are even desirable in the
environment. It's very easy to code destruction, and it is easy to do.
You can assume that most players will therefore engage in it once
supplied. And building will have the following problems:

- a strict realism means that builders will be bale to make highly
disjointed environments that may subtract from the sense of realism
(look ma, I can perch a chair on the chimney!)
- it's a lot harder to code the ability to create objects in the world
using a sim model than it is to code destroying things
- you've got fewer people willing to build than you've got willing to
destroy, and those people will not take kindly to people burning the
chair on their chimney

>  It certainly is a handy excuse, but it hardly 
> represents good design,
> and in the end will lead to a very inflexible and proprietary 
> functionality.
> This may or may not be a Bad Thing.

Right--the issues I wanted to raise were ones of how directed the
players are, and how much choice and freedom they have.

> > Similarly, if you don't have "burned" long descs for all the various
> > objects in your database, then once the fire is over, you 
> may not have
> > achieved the effect you want. We're always stuck with "only the
> > interactions which are coded" after all...
> 
> Yes - but assuming that you build the room out of standard components
> (wooden furniture, latex paint, plaster walls, glass windows) 
> then this should already exist.

Should, but probably won't. :)

> Again, it's just a question of whether you want to spend a lot of time
> building/coding many interactive, reusable pieces and then 
> building a world
> out of those, or whether you want to spend your time making 
> hand-crafted
> areas and objects which are highly proprietary.

The irony is: you'll still need hand-crafted things, and you'll also
never get quite enough reusable pieces to get what you want out of the
world.

Or rather, you might, but it'll take you a REALLY long time.

-Raph