Choosing your IF setting / genre

28 views
Skip to first unread message

Jason Noble

unread,
May 17, 1994, 3:55:25 AM5/17/94
to

Greetings all,

I've been tinkering with my chosen development system long enough now that I
feel it's time to start on a fully-fledged game. One issue that I want to
carefully think through is the setting of the game-to-be. There's no end of
valuable advice in Graham Nelson's latest authorship guide, the TADS manual,
and G. Kevin Wilson's guide; all three of these authors warn of the many
issues to consider.

However, I thought I'd try to start some discussion on one particular aspect
that's plaguing me: the "historical" period of the game. In short, whether
it's set in a (modified) past, present or future.

It seems to me that games set in the "past", where the level of technology
available to the player is limited to classic adventure game items like
swords, sacks and magic wands, can be *much* easier to implement because
most objects can't do a lot to affect the game world. For example, the
sword is good for attacking living creatures and cutting any ropes, etc. and
maybe as a lever in a few defined situations. Not really a nightmare to
code.

This is in marked contrast to present or, even worse, future games: what if
the player gets hold of a flamethrower, or a supply of petrol? Surely the
player should be allowed to use big, powerful destructive implements like
this anywhere in the game. The intended (by the author) use of the
flamethrower might be to burn away some spiderwebs and the spiders within.
But what's to stop the player from trying to burn everything in the
lounge-room, or set the forest on fire, or torch the car? Short of
preventing the player from doing these things with a lame excuse like "You
don't need to do that to complete this game", the author has to try to code
large scale effects in the game world, such as a forest fire covering tens
of locations. After such a fire, the description of each location should
change markedly, and any actors or creatures living in the forest must be
seen to be burnt to a crisp or to have run away. Intelligent actors should
also be curious about the fire afterwards. This sounds like way too much
work to implement, especially if burning down the forest is not important to
the plot.

At this point, you may be thinking, OK, why doesn't he just write a
swords-and-sorcery type game? To tell the truth, I'd actually rather avoid
this, because such settings have become so hackneyed both in IF and other
literary domains (I may yet write such a game, and I would do my best, but
there have been far too many hobbit-clones loosed on the world already, I
think). I suppose the discussion I'm trying to start is about how do you
set your game in the present/future, with plenty of appropriate items lying
around, and appear to give the player complete freedom while sneakily
avoiding having to code an accurate physical simulation of the whole world?

One solution, which does not appeal to me, is to set the game in an
environment that's low on useful items and complex actors, and big on
indestructible scenery. Another, that I think has more promise, is to set
up the player as a particular character who would not be likely to go around
trashing the place (eg. Curses, where the protagonist starts out as an
aristocrat exploring his own attic - not much chance of flamethrower use
here).

I know that many of you out there are working on games right now. I'm not
trying to steal your secrets, but is there anyone who'd like to share with
the newsgroup their approach to this particular IF problem?

Regards,

---------------------------------------------------------------------------
Jason Noble | jno...@bunyip.bhs.mq.edu.au
National Centre for HIV Social Research | jno...@laurel.ocs.mq.edu.au
Macquarie University, Sydney, Australia | ph. (61 2) 850 8667
---------------------------------------------------------------------------

David Baggett

unread,
May 18, 1994, 12:08:17 PM5/18/94
to
In article <2r9t9d$s...@sunb.ocs.mq.edu.au>,
Jason Noble <jno...@laurel.ocs.mq.edu.au> wrote:

>This is in marked contrast to present or, even worse, future games: what if
>the player gets hold of a flamethrower, or a supply of petrol?

So you don't put any flamethrowers in. It's your game, you put whatever
you want in. This may seem like a no-brainer, but it's easy to get locked
into choices you've made that could easily be changed.

I'll give you an example that I heard a while ago: A writer was working on
a book that featured a river in a particular area. The writer had gone to
great pains to describe the river as narrow (for some reason that I've
forgotten). Later, another author suggested a plot change that involved
some characters floating down the river on a raft. "I can't have them
float down the river in a raft," she argued, "because the river's way too
narrow for that!" The other author reminded her, "It's your river -- make
it as wide as you want!"

The point is that you can put in or omit anything you want -- it's *your*
story to tell. I don't think leaving out flamethrowers or cans of petrol
or atom bombs will make your game seem "less like the present".

Also, keep in mind that the reader is a willing participant, much more
likely to go with the flow than you are as the author. The player, unless
given a dramatic display to the contrary, assumes your game is *correct* --
i.e., bug-free and the way you intended it to be. Given this, players
aren't going to be looking for things wrong. "Ah ha! No flamethrower.
This world is obviously flawed!" is not something a player would ever say.
(Unless you're talking about TABU members, but then they try things
like "scrape canary".)

The problem you're really alluding to here is the "too powerful object"
problem. Don't ever put in any object that's capable of mass destruction,
or purports to alter many different objects, unless you're willing to do a
lot of special-case coding. This applies to fantasy magic items as well as
to high-tech weapons. Think about a "wizard lock" wand that permanently
locks anything you zap. You're going to have to do a lot of work to make
sure that doesn't mess up any puzzles. Ditto for the complementary
"skeleton key".

>Surely the player should be allowed to use big, powerful destructive
>implements like this anywhere in the game.

This brings up alternative two: cleverly restrict where the player can go
with the object so that you don't get into troublesome situations. These
restrictions can range from utterly transparent ("Nope, I'm not going to
let you go to the bridge with the quark bomb.") to amusingly blatant ("You
head north with the Zap-O-Dragon gun, but you trip over a rock and set off
the weapon, which blasts a nearby tree, which falls on you.") to eminently
logical ("As you head north, you notice that the Zap-O-Dragon gun seems to
have stopped working. Perhaps the wizard's abode houses its power source?").

>One solution, which does not appeal to me, is to set the game in an
>environment that's low on useful items and complex actors, and big on
>indestructible scenery.

This is an extreme way of putting it. You can have plenty of useful items
and destructible scenery, just no massively destructive items or uselessly
destructible scenery. Actor complexity seems tangential to me.

>I know that many of you out there are working on games right now. I'm not
>trying to steal your secrets, but is there anyone who'd like to share with
>the newsgroup their approach to this particular IF problem?

I generally avoid super-powerful items. I prefer to make puzzles that
require you to think of less-than-obvious uses for everyday items. You can
do a lot with a paper clip and a wad of chewing gum -- just ask McGyver.

Dave Baggett
__
d...@ai.mit.edu MIT AI Lab NEW .sig Lite!
ADVENTIONS: We make Kuul text adventures! Email for a catalog of releases.

Gareth Rees

unread,
May 18, 1994, 11:47:39 AM5/18/94
to
Jason Noble writes:
> What if the player gets hold of a flamethrower, or a supply of petrol?

It may sound obvious, but the player can only get hold of a flamethrower
or a supply of petrol if you give her one! If you include in your game
something which has many possible uses (NPCs are more usual examples
than flamethrowers) then you can expect it to be difficult to code.

Think of a work of interactive fiction as being another sort of game
between the player and the writer. The player wants to pick holes in
your universe, to find inconsistencies or arbitrary and annoying
restrictions on what she can do; the writer wants to have covered all
these cases with reasonable and/or witty responses.

Obviously the writer is the loser here, especially if there are people
in the game that the player can talk to (I've never seen any NPCs that
were remotely believable once you got them cornered and started talking
to them), but you have some tricks you can play:

* Pouring petrol everywhere and setting fire to it is the kind of thing
that in any sensible game kills the player instantly rather than
causing the sort of consequences that Jason's worrying about.

* The game can fight back:

> strike match
A sudden gust of wind blows the match out. You have four matches
left.

> pour petrol on grass
A park ranger comes up to you. "Well, well," he says, "thought we'd
have a little barbecue, did we?"

* You can make powerful items difficult to transport and so restrict
them to a small area of the game in which you're more likely to be
able to code all the reasonable consequences of using them. For
example, a flamethrower might need a large tank of fuel that won't fit
through certain doors, or it might be attached to a tank which can't
drive off the island. The supply of petrol maybe available in
location A, but not in location B because in order to get to location
B you'd need to have used up the supply to refuel the helicopter. And
so on.

* NPCs can get bored with the player or suddenly find other business
that's more important than chatting to her; when they return in 20
moves time or the next day or whatever's appropriate then it's
reasonable that they'll have forgotten exactly what they said to her
last time and to start repeating themselves. Legend's game 'Gateway'
makes good use of this trick.

--
Gareth Rees

Russell Wallace

unread,
May 18, 1994, 9:31:11 PM5/18/94
to
[Comments about being able to blow things up deleted...]

My opinion about this is that it should be thought of as an interesting
challenge, not something to be avoided. For example, people don't
restrict games to verb-noun sentences because sophisticated parsers are
harder to write, they write them anyway... Now, in my opinion, the
ability to go around torching everything in sight with a flamethrower
(even if it was only provided for one puzzle and is only useful there)
or cutting things up with an axe (ditto) would do far more to make the
game believable, than the ability to type TELL THE ROBOT TO BURN ALL THE
BOOKS EXCEPT THE BLACK AND RED ONES or suchlike; if people are prepared
to spend huge amounts of effort on the latter, why not the former? It
needn't be *that* difficult... just record what material everything is
made of, and have a section of code that says objects made of paper will
burn easily (e.g. with a match), objects made of wood will burn with
difficulty (e.g. with a flamethrower), metal and stone won't burn at
all...

--
"To summarize the summary of the summary: people are a problem"
Russell Wallace, Trinity College, Dublin
rwal...@cs.tcd.ie

Gerry Kevin Wilson

unread,
May 20, 1994, 2:35:31 PM5/20/94
to
Dang. I posted a long response to this message, but it got ate somewhere
on the way. Grr. Well, I'll repost a brief summary. Here are the basic
time periods for a game:

distant past - Unrecorded, or poorly documented time in the past. This
setting doesn't require as much work as many of the others in
that there's not all that much research to be done.
alternate past - The What If? past. What if Napoleon had conquered all
of Europe? This probably is one of the hardest. Not only do you
have to do research, but you have to engage in speculation.
near past - Recorded history. Requires research to do well.
present - Natch.
near future - The next 100 years. Requires some attempt at speculation.
far future - 100+ years. No telling WHAT will happen at that point, so
just make up what you like.
mythic past, present, or future - Take any period in history, present or
future. Add magic. Bring its myths to life. There you go.

A Word on Euro-Centric Game Writing

When writing mythic games, authors often manage to forget that
there is an entire world filled with diverse cultures and folklores.
Just about every mythic game I've seen has been based in some form on the
AD&D image of England, or Tolkien's works. The rest of the world has
been passed by in favor of this very narrow genre. Even TSR, the makers
of AD&D, have recently been trying to rectify this situation in their
games by introducing such boxed sets as The Horde, based on Genghis
Khan's army. Take any period in history, anywhere in the world, and add
magic, then bring its myths to life. What about the American Indians?
They have some fascinating legends about Coyote and other fantastic
beings. Of course, I should add that the legends are different for each
tribe. What about the eskimos, the Australian aborigines, the early
polynesians, the ex-USSR's people, before their folklore turned to bible
parables, or after, if you like, and many more. All it takes is a book
on folklore, and a book on history for the period and place you want to
concentrate on.

[End of non-Vertigo related stuff, begin author request.]

Vertigo Software is looking for authors with game ideas. If you
want to write a game and make some money, but don't want to deal with all
the hassle of advertising, distributing, taking orders, shipping,
packaging, etc, then drop me a line and I'll send you more info whenever
I'm around. I may be gone for a few weeks, until I can wrangle a summer
account, but I'll get back to you.
--
<~~~~~~~S~W~~~~~~~~~~~~~~~AVALON~~~~~~~~~~DUE~IN~AUGUST~~~~~~~~~~~~~~~|~~~~~~~>
< ERT O A In the midst of the Vietnam War, one man dies, and is | ~~\ >
< V IGO F R charged with a quest from King Arthur. Live the quest! | /~\ | >
<_______T_E_____________...@uclink.berkeley.edu__|_\__/__>

Bob Newell

unread,
May 19, 1994, 9:34:50 PM5/19/94
to
>that's plaguing me: the "historical" period of the game. In short, whether
>it's set in a (modified) past, present or future.
(discussion about the problems posed by various technology levels deleted)

To me the interesting question here deals with the research needed to get
good period feel. I'm working on one game set in the recent past (20 years
back or so), and have literally bogged down in researching the setting.
Well-known settings are SO easy to get wrong.

This makes fantasy, and to some lesser extent future settings, a little bit
to a great deal easier.

Many I-F pieces have a certain period feel, without much detail (Plundered
Hearts, Abbey Montglane). But to really get into representation of a certain
era, a lot of background work seems needed. At random, here are a few
musings on eras and locations I'd find interesting:
Boston or Berkeley, 1968
French Indo-China, around 1950
India, around 1870
Of course, making these into a _game_ would be quite a proposition.

The Grim Reaper

unread,
May 20, 1994, 11:30:11 PM5/20/94
to
In article <1994May19....@cs.tcd.ie>,

Russell Wallace <rwal...@cs.tcd.ie> wrote:
>[Comments about being able to blow things up deleted...]
>
[deletia]

>(even if it was only provided for one puzzle and is only useful there)
>or cutting things up with an axe (ditto) would do far more to make the
>game believable, than the ability to type TELL THE ROBOT TO BURN ALL THE
>BOOKS EXCEPT THE BLACK AND RED ONES or suchlike; if people are prepared
>to spend huge amounts of effort on the latter, why not the former? It
>needn't be *that* difficult... just record what material everything is
>made of, and have a section of code that says objects made of paper will
>burn easily (e.g. with a match), objects made of wood will burn with
>difficulty (e.g. with a flamethrower), metal and stone won't burn at
>all...

Sheesh, you have got to be kidding. Implementing any sort of vaguely
realistic physics into an i-f game is difficult, if not impossible. Just
saying that wood objects will burn under certain conditions isn't enough.
We'll have to change the description, possibly create ashes, alter the object's
weight, change NPC's reactions to the object, possibly have certain puzzles
change (a bucket with the bottom burnt out isn't going to be much use), check
if the object's containing anything so that we can see if those things burn,
etc, etc. Physics is a much harder problem than parsers, which is why we
see more of the latter than the former.

>--
>"To summarize the summary of the summary: people are a problem"
>Russell Wallace, Trinity College, Dublin
>rwal...@cs.tcd.ie

+----------------------------------------------------------+
| One .sig to rule them all, one .sig to find them... |
| One .sig to bring them all and in the darkness bind them |
+----------------------------------------------------------+
| The Grim Reaper (Reaper of Souls, Stealer of .sigs) |
| scy...@u.washington.edu |
+----------------------------------------------------------+

David Baggett

unread,
May 21, 1994, 8:58:07 AM5/21/94
to
In article <1994May19....@cs.tcd.ie>,
Russell Wallace <rwal...@cs.tcd.ie> wrote:

>It needn't be *that* difficult... just record what material everything is
>made of, and have a section of code that says objects made of paper will
>burn easily (e.g. with a match), objects made of wood will burn with
>difficulty (e.g. with a flamethrower), metal and stone won't burn at
>all...

And then do what, change the description of the object to be "The X looks
burned?" How about if the player torches himself or another actor? What
about things that melt or vaporize? There are hundreds of objects you need
to account for here, and burning most of them will have some nontrivial
effect specific to *those* objects.

The whole point is that the player wants to be able to burn everything in
sight because he wants the game to model the world realistically. No code
that checks an object's general composition and sets a flag "burned = true"
is going to be realistic.

This gets back to the basic "realism vs. playability" argument. Though
many people think this kind of realism is good for IF, I'd argue that the
end result is a poorly designed game which is no fun at all. (You torched
that newspaper early on? Too bad, you needed to read that to find out
about the secret headquarters on Galapaos Island... But I won't tell you
that, because in a "realistic" simulation you don't get this kind of
feedback.) Ugh.

In real life, doing anything nontrivial is a complete pain. (Oh geez, how
the hell am I going to get to *Galapagos Island*? Gotta call a travel
agent. And how am I going to pay for this? And what about my chemistry
final?). Consequently, trying to make a game model the real world is
actually a *bad* idea -- what you want is a game world that amplifies the
"highs and lows" and glosses over the boring stuff. Impressionism, not
realism, is the traditional basis for good games.

Sean Barrett

unread,
May 21, 1994, 2:02:09 PM5/21/94
to
David Baggett <d...@ai.mit.edu> wrote:
>And then do what, change the description of the object to be "The X looks
>burned?" How about if the player torches himself or another actor? What
>about things that melt or vaporize? There are hundreds of objects you need
>to account for here, and burning most of them will have some nontrivial
>effect specific to *those* objects.

I agree, putting burning in correctly is non-trivial. It
is also not impossible (just a lot of work). Perhaps you
are willing to put in a lot of work writing creative long
descriptions for rooms and objects. Perhaps I am willing
to go to a lot of work to make objects which can interact
in much more complex ways than is traditional.

Perhaps _for you_ the balance is such that the former is
far faster than the latter. But perhaps for me the opposite
is true?

>No code that checks an object's general composition
>and sets a flag "burned = true" is going to be realistic.

Only if there isn't appropriate other code to make it realistic.

>This gets back to the basic "realism vs. playability" argument.

Misleading. Certainly you must work at it to make it playable.
But it's not a tradeoff; if you want more realism, you have to
work harder to make it playable; you don't give up playability.

>Though many people think this kind of realism is good for IF,
>I'd argue that the end result is a poorly designed game
>which is no fun at all.

Only if you assume the game is poorly designed.

(You torched
>that newspaper early on? Too bad, you needed to read that to find out
>about the secret headquarters on Galapaos Island... But I won't tell you
>that, because in a "realistic" simulation you don't get this kind of
>feedback.) Ugh.

And in a "well-designed" game of this sort, you either have other
ways to find out, or you don't need to get there to "win" the game.

>Consequently, trying to make a game model the real world is
>actually a *bad* idea -- what you want is a game world that amplifies the
>"highs and lows" and glosses over the boring stuff. Impressionism, not
>realism, is the traditional basis for good games.

I don't know, by "good games" do you only mean adventure games?
Again, there's no doubt you don't want infinite realism in a game,
since it would be unlikely to be fun. But that falls under the
"tuning playability" issue.

All adventure games start and are founded in some sort of realism;
otherwise they'd be unplayable. There is a set of consistent
"laws of the universe" and objects within that universe follow
those laws.

As you allow those laws to become more complex, the system becomes
more and more of a simulation.

I don't think there's anything wrong with that; simulations can
still be games (e.g. SimCity). And I think the same can be
said of the "text adventure" medium.

There are, I guess, two fundamental pulls. One is story-telling,
and the other is simulation. I don't think either is "more
valid" than the other. Perhaps some people will claim that
the "fiction" in "interactive-fiction" means only story-telling
is valid, but I think that's an amazingly unjustiifiable
stretch. Nobody knows what interactive fiction "should be";
it's something people are still experimenting with.

I admit it, I like to tell stories. But I also like
simulation. I like integrating the two, but I'm also
aware of the suspicious flaw of trying to tell a story
to someone playing a game (unless you force the player
to do the right thing to advance the plot, the player
doesn't know how to advance the plot--how to tell the
story; the "game" devolves into brief story-telling
segments and intermediate time with the player trying
to figure out how to advance it). Pure simulation allows
the player to make up the story entirely on their own.
Perhaps, ideally, the optimum is somewhere in the
middle; perhaps there is no optimum, there are simply
different sorts of games.

But this whole narrow-minded anti-simulation "it can't
be done" is exceedingly silly. "I don't want to do it
in my games" or even "I wouldn't be interested in
playing a game that did that" is fine. I realize that
I have slightly taken the poster above out of context
in this regards; but it reads to me as over-generalization.

As to a practical answer to "how to do it"; you create,
for each object class, a method executed when the object
is burned. Yes, you have to write code--not for every
object, but for every class of objects. You probably
have to write a separate description for every object
which can become burned but still remain an object.

One of the fundamental ideas here in achieving something
like this is knowing up front that you want to support
it, and building it into the game from the beginning.
Each time you write a behavior for an object, you ask
yourself, "Can they do this if it's burning/burnt/soaking
wet?" and adjust the code as appropriate.

Sean Barrett
speaking from experience again

Russell Wallace

unread,
May 21, 1994, 3:41:07 PM5/21/94
to
scy...@u.washington.edu (The Grim Reaper) writes:

>In article <1994May19....@cs.tcd.ie>,
>Russell Wallace <rwal...@cs.tcd.ie> wrote:
>>[Comments about being able to blow things up deleted...]
>>
>[deletia]
>>(even if it was only provided for one puzzle and is only useful there)
>>or cutting things up with an axe (ditto) would do far more to make the
>>game believable, than the ability to type TELL THE ROBOT TO BURN ALL THE
>>BOOKS EXCEPT THE BLACK AND RED ONES or suchlike; if people are prepared
>>to spend huge amounts of effort on the latter, why not the former? It
>>needn't be *that* difficult... just record what material everything is
>>made of, and have a section of code that says objects made of paper will
>>burn easily (e.g. with a match), objects made of wood will burn with
>>difficulty (e.g. with a flamethrower), metal and stone won't burn at
>>all...

>Sheesh, you have got to be kidding. Implementing any sort of vaguely
>realistic physics into an i-f game is difficult, if not impossible. Just
>saying that wood objects will burn under certain conditions isn't enough.
>We'll have to change the description, possibly create ashes, alter the object's
>weight, change NPC's reactions to the object, possibly have certain puzzles
>change (a bucket with the bottom burnt out isn't going to be much use), check
>if the object's containing anything so that we can see if those things burn,
>etc, etc. Physics is a much harder problem than parsers, which is why we
>see more of the latter than the former.

Only if you demand perfection or nothing. That attitude to parsers
would have nobody ever write one better than GO NORTH because it
wouldn't be able to understand TO BE OR NOT TO BE, THAT IS THE QUESTION,
so why bother? For wood and paper objects burning, it suffices for game
purposes to simply make them disappear.

David Baggett

unread,
May 21, 1994, 3:15:35 PM5/21/94
to
In article <2rlib1$o...@news.tamu.edu>, Sean Barrett <se...@stat.tamu.edu> wrote:

>>This gets back to the basic "realism vs. playability" argument.
>
>Misleading. Certainly you must work at it to make it playable.
>But it's not a tradeoff; if you want more realism, you have to
>work harder to make it playable; you don't give up playability.

First off, cite a *single* example of a successful work of interactive
*fiction* that adheres to the requirement that it be this realistic -- even
a prototypical one. If you can't, explain why there isn't one yet. (This
is a genuine request, not just a challenge.)

The more realistic your game is, the more things "should work" to solve
"the puzzles" and advance "the plot". The *very idea* of a plot is
unrealistic (modulo ideas about predestination which I suppose we could
debate as well). Dramatic pacing is an unnatural notion. Yet these things
are essential, I would argue, to a "fulfilling" experience.

You later allude to works in which the player creates his own "fiction".
The problem with this is that only a handful of people (if any) are capable
of effortlessly improvising a compelling fictional experience. Here we
have a kind of fiction that makes such demands on the audience that it is
largely inaccessible. To the rest of us, it amounts to random meandering
through story-space. *This is a flaw in the form, not in the proponents*.

>As you allow those laws to become more complex, the system becomes
>more and more of a simulation.

...and less and less of a *game*, which is what I was trying to get it.
Yes, it comes down to where you draw the line, which is a matter of
personal taste. But in the extreme (which is where I put allowing the
player to torch everything in the game with a flamethrower), I think
the desire to make games "realistic" is counterproductive.

>There are, I guess, two fundamental pulls. One is story-telling,
>and the other is simulation. I don't think either is "more
>valid" than the other.

The ultimate arbiter of what is "valid" entertainment or "art" is the
(extemporal) public. I suspect that, in the much the same way that the
public rejected serialist music because it demanded too much and resonated
too little with listeners' thoughts and feelings, so too will heavily
"simulationist" player-as-author IF fail to really "make it" as an art
form.

>Nobody knows what interactive fiction "should be"; it's something people
>are still experimenting with.

Not true at all! We have thousands of years of literature to start from.
Interactive fiction is NOT a topic in a vacuum, with no precedent. The
most basic purpose of fiction is to communicate the author's thoughts and
feelings about *something*. Simulationists seem to reject this out of
hand, or at least ignore its importance.

Again there are parallels to serialism. Schoenberg et al decided that what
the world needed was a new kind of music -- one devoid of all the trappings
of what previous generations called music. So they started rebuilding
"music" from the ground up. It didn't work -- they didn't have 1000 years
of musical genius in them; only a single lifetime's worth, so they couldn't
pull it off.

You can't ignore the storytelling methods that it's taken mankind several
thousand years to develop. It's clear from your comment that "the player
doesn't know how to advance the plot" that you're not totally ignoring
these issues, but I thought it was worth making the point once and for all.

>"I don't want to do it in my games" or even "I wouldn't be interested in
>playing a game that did that" is fine.

How about this: "I don't know anyone personally who likes this kind of
'fiction', haven't heard anyone ecstatically espousing its virtues or
joyously recounting its pleasures, and can't really imagine it going over
with the public at large."

(Now you're probably going to say "that's your opinion", "you're
generalizing", and "you're awfully presumptuous to think you know what the
public at large will like", all of which are true.)

>Sean Barrett
>speaking from experience again

Aren't we all!

Phil Goetz

unread,
May 21, 1994, 5:10:29 PM5/21/94
to
In article <2rjv83$h...@news.u.washington.edu>,

The Grim Reaper <scy...@u.washington.edu> wrote:
>In article <1994May19....@cs.tcd.ie>,
>Russell Wallace <rwal...@cs.tcd.ie> wrote:
>>[Comments about being able to blow things up deleted...]
>>
>[deletia]
>>(even if it was only provided for one puzzle and is only useful there)
>>or cutting things up with an axe (ditto) would do far more to make the
>>game believable, than the ability to type TELL THE ROBOT TO BURN ALL THE
>>BOOKS EXCEPT THE BLACK AND RED ONES or suchlike; if people are prepared
>>to spend huge amounts of effort on the latter, why not the former? It
>>needn't be *that* difficult... just record what material everything is
>>made of, and have a section of code that says objects made of paper will
>>burn easily (e.g. with a match), objects made of wood will burn with
>>difficulty (e.g. with a flamethrower), metal and stone won't burn at
>>all...
>
>Sheesh, you have got to be kidding. Implementing any sort of vaguely
>realistic physics into an i-f game is difficult, if not impossible. Just
>saying that wood objects will burn under certain conditions isn't enough.

It is good enough. In _Inmate_, if I recall correctly, being "burnt"
was equivalent to being "broken". A broken object could not be used
in certain ways. For example, a burnt match could not be lit.
If you burnt or broke an object, that object would acquire the property
"broken" and its name would be prepended by "burnt" or "broken".
I don't remember if I burnt the objects inside a burning object
recursively, but that would be easy. I didn't go to the trouble of
distinguishing between strengths of fires, which would be a nice
next step.

Physics is hard, but using that as an excuse to avoid even the most
basic aspects, such as broken objects, is being lazy.

Phil Goetz

Ms. Pauline Benzies

unread,
May 21, 1994, 9:34:01 PM5/21/94
to
% >It is good enough. In _Inmate_, if I recall correctly, being "burnt"
% >was equivalent to being "broken". A broken object could not be used
% >in certain ways.

Convince me that the author anticipated this one...

> EXAMINE CHANDLIER
It appears to be held to the roof by a rope securely attached to the wall.

> CUT ROPE WITH BROKEN BOTTLE
The chandelier falls to the ground, revealing that it was made of diamonds.
With the wealth you have now gained, you decide to give up playing computer
games and settle down in the South of France.

Ms. Pauline Benzies

unread,
May 21, 1994, 9:43:28 PM5/21/94
to
In article <2rlib1$o...@news.tamu.edu> se...@stat.tamu.edu (Sean Barrett) writes:
> (You torched
>>that newspaper early on? Too bad, you needed to read that to find out
>>about the secret headquarters on Galapaos Island... But I won't tell you
>>that, because in a "realistic" simulation you don't get this kind of
>>feedback.) Ugh.
>
>And in a "well-designed" game of this sort, you either have other
>ways to find out, or you don't need to get there to "win" the game.

Why is this different to publishing that you only need to type 'IDDQD' in Doom
and suddenly none of your stupid actions can kill you. Just charge on through
ignoring the bad guys and push the button marked "END".

The point in concealing clues is that they are *concealed*. Having a player
who says "BURN ALL WITH FLAMETHROWER" and then complains that because they
torched the Mona Lisa, they can't convince the police to let them out of the
jail to finish the game, seems to have missed the point.

>to figure out how to advance it). Pure simulation allows
>the player to make up the story entirely on their own.

At this point, the game author ceases to be an *author*, don't they?

>Perhaps, ideally, the optimum is somewhere in the
>middle; perhaps there is no optimum, there are simply
>different sorts of games.

I tend to agree with this last, that the two forms are not compatible

>Each time you write a behavior for an object, you ask
>yourself, "Can they do this if it's burning/burnt/soaking
>wet?" and adjust the code as appropriate.

Lets see, how do I code the "broken" bottle. Obviously, if they used
"BREAK BOTTLE CAREFULLY" there will be a clean edge that can be used to cut.
On the other hand, "BREAK BOTTLE WITH AXE" tends to leave long splinters.

etc.

Sean Barrett

unread,
May 22, 1994, 2:56:19 AM5/22/94
to
Let me establish some fundamentals here.

The only "works" I have played which had
pretentions of being "interactive fiction"
were Infocom games. Now, I realize, there's
a lot of much more advanced research going
on, and "interactive fiction" is intended
to mean something more.

Yet, most _practical examples_ of interactive
fiction that most of the people here have
experienced are quite similar--they are,
at heart, adventure games.

I have been asked if I can name a game
which is "more simulator" than traditional
adventure games. I cannot name one; but
I do not know what this proves. I cannot
name any game which has more complex character
interactions than the old infocom games,
either; does this mean interactive fiction
should not aspire to better? (Yes, perhaps
you can name one; but that misses both the
point that just because I can't name one
doesn't mean there isn't one, and the point
that what exists is irrelevent to what
we might want to exist.)

Let us now turn to one person's viewpoint
on a related issue.

Ms. Pauline Benzies <benziesp@ponderosa> wrote:


>se...@stat.tamu.edu (Sean Barrett) writes:
>>And in a "well-designed" game of this sort, you either have other
>>ways to find out, or you don't need to get there to "win" the game.
>
>Why is this different to publishing that you only need to type 'IDDQD' in
>Doom and suddenly none of your stupid actions can kill you. Just charge
>on through ignoring the bad guys and push the button marked "END".

Fine, yes. My point was that we make a presumption of
competence on the part of the author. Saying "you
can't do that because then you break the game" is
presuming incompetence; it was to that I was responding.

If you insist on writing a game in which you must have
one particular hidden clue, then, yes, you should make
sure it is not accidentally destroyed, or you must
accept that the player can get in a no-win situation.

>>to figure out how to advance it). Pure simulation allows
>>the player to make up the story entirely on their own.
>
>At this point, the game author ceases to be an *author*, don't they?

With pure simulation? Yes, on the surface; but no, in the
end. The author chooses the nature of the simulation.
The author determines the possible paths and the possible
outcomes.

But, overall, I am not advocating pure simulation. I
was pointing out the other extreme.

>>Each time you write a behavior for an object, you ask
>>yourself, "Can they do this if it's burning/burnt/soaking
>>wet?" and adjust the code as appropriate.
>
>Lets see, how do I code the "broken" bottle. Obviously, if they used
>"BREAK BOTTLE CAREFULLY" there will be a clean edge that can be used to cut.
>On the other hand, "BREAK BOTTLE WITH AXE" tends to leave long splinters.

Yes, what is your point?

Why is it that people seem to think that the

SIMULATION

which goes on in a traditional adventure game
"doesn't count", and is in fact the end-all
and be-all of the elemtns of simulation in
a traditional computer game?

A traditional adventure game allows:

objects which have spatio/temporal coordinates
which are quantized to discrete locations and times
objects which can act as "containers" for other objects
objects which are "openable" and "closeable"
objects which are "lockable" and "unlockable"

Essentially, I can draw an analogy between your complaints
(along with every other "anti-simulation" post I've seen)
and a hypothetical complaint against someone suggesting we
add some of the above facilities to an even-more simple
world. (This is a "flatland" kind of analogy.)

Suppsoe we have an adventure game in which there are no
containers and no openable/closeable doors (and hence
nothing lockable).

I write a post mentioning that doing something like adding
containers wouldn't be that big a deal. I receive followups
with the following sorts of suggestions: "That would be
much too complicated. Real containers have a limited
amount of space in them. And they break if you put too
much stuff by weight or by volume. Real containers tend
to stretch and squish when you put things in them. And
how will you deal with being able to fill containers with
water if you have a river in your game?"

Of course, we all know the answers to those questions. Some
of them are even thorny problems that often aren't quite
simulated as well as we might like. But we live with it;
it's only a game/story, after all. Perhaps someone points
out that real containers are a pain in the butt to deal
with packing and unpacking, and not only will simulating
the layout of objects in containers be a pain, so will
it be a pain for players to deal with them. We all know
how that problem is solved these days.

How about I suggest adding openable and closeable doors,
and perhaps lockable. "Well, you should be able to put
objects in the way, so you can't open the doors. Or
prop them open. And you should be able to pick the
locks." Does anyone implement any of that? No (except
in the case of specific puzzles). Does anyone care?
Not really.

How many games implement "burning" candles? Do they
make sure they're set up to give off light (if not,
it's a bug)? Do they make them drip wax, or provide
the ability to get the soft wax from the candle?
No. Is it a bug? No, not unless you have something
in your game that they might really want to use
candle wax on.

If you take the simulation out of adventure games,
you're left with choose-your-own-adventures. You're
writing a "menu maze"; each element of interaction
"advances the plot" in some way.

Adventure games aren't anything like that. Much of the
effort expended in adventure games is simulation--moving
from place to place, examining things, picking things
up, dropping things.

This simulation helps "immerse" the player, drawing
them into the game, making them an active participant
instead of the recipient of a story being handed
down by the gods.

Now, this is a good time to remind you--interactive
fiction doesn't have to be about adventure games. There's
definitely stuff out there considered interactive
fiction which doesn't involve simulation in the slightest.
But here, in the discussion people have been carrying out,
there seems little doubt that the thing under consideration
is adventure games.

Why is there a push to create more complex characters
in interactive fiction? Do more complex characters make
it easier to create plot, pacing, foreshadowing, or any
fundamental literary techniques?

As far as I can see it, creating more complex AIs allows
the author to have better "indirect" control. The author
can get more complex behavior in-response-to-arbitrary-inputs;
that is, the author can create an actor that deals with
situations-not-anticipated-by-the-author in ways
desired-by-the-author (hyphens for clarity).

As far as I can see it, the goal of improving the author's
indirect control is to allow the author to "relax" control
of the story, to make it _more_ of a simulation, while
still achieving the desired results.

All of these things, to me, make it very difficult for me to
understand the "anti-simulation" orientation of most of
the posters on this newsgroup. I'll repeat those
"all of these things" for clarity:

1. every existing adventure game consists of
a simulation of a universe whose laws were
set out in Adventure
2. interactive fiction without simulation is
choose-your-own-adventure or a hypertext
extravaganze
3. development of complex AIs promotes use of
AIs in simulation

To me, simulation is fundamentally a part of every
adventure game I've ever seen, and therefore to
most things which qualify as "interactive fiction".
From this viewpoint, I have extreme difficulty
understanding the idea that people could be
entirely anti-simulation. To put it more bluntly,
I find it very disturbing that to me, this is
fundamentally obvious, but to most of the people
out there, it apparently is unbelievable.

Perhaps I merely misunderstand the complaint.

Perhaps the claim is that it is a matter of
degree of simulation, not the idea in and of itself.
But I'm confused as to why the degree of simulation
chosen for some ancient Fortran program (I think)
is the "best" degree of simulation. Certainly most
people want to advance the state of AIs; what's
wrong with advancing the state of the rest of the
universe?

Examples like the broken bottle are all well and
good. Some objects do indeed behave in rather
interesting ways under certain conditions. That
is the fundamental complexity of simulation.
However, implying that this generalizes to every
object is extremely misleading. And presuming
that the author is going to create a breakable
bottle but not think about what happens when
the bottle is broken is silly. Perhaps you in
particular wouldn't want to deal with this case
(you obviously can think of it); in which case
you'd simply make the bottle unbreakable.

Some of the complaint against simulation seems to
be of the form "you will create an expectation in
the player for realism which you cannot live up
to". Suggesting this is misunderstanding the nature
of the increased realism. Players confront limited
realism every day when playing an adventure game;
but they learn to understand the laws of the
universe in which they are playing (the limitations
upon the reality that is simulated). Offering
improved simulation doesn't change that fundamental;
it will be more real, not real. The only goal
is to achieve consistency within that universe;
to achieve predictability, so that players find
their universe behaves as they would expect it to.

ObSimulation:

Here are a number of ways in which you can
take a step forward over traditional systems. I
do not believe any of these ways are excessively
complex; they are complex, no doubt, but not complex
beyond implementability.

1. 3D

Define the universe with a proper 3D model.
Still divide areas up into rooms and the like.
Have to beat on the user interace to make it still
convenient to get things done. The addition of
3D coordinates will allow better simulation of
"spatial" things like tied ropes, moveable obejcts,
etc.

2. Liquids

You can define containers to be liquid-permeable
or not. Liquids propogate in and out of containers.
Objects "interact" with liquids according to their
materials. See the explanation of "interaction" below.

3. Fire

Things can "catch on fire", burn for a length
of time, and then burn out. Other things can catch
fire from them (this overlaps somewhat with liquid
propogation).

4. Gravity

Implementing a variant of "cartoon gravity", in
which objects fall instantaneously, is relatively
straightforward. If you have some method of allowing
player freedom to fly, gravity is very important.
If you haven't implemented a full 3D environment,
then you need to build extra rooms for the player
to fly in. It is possible to build reasonably
believable models of cliffs and the like; however,
you then want to allow things like climbing up cliffs,
tying ropes at the tops of cliffs and being able to
climb down them, etc. Still, I'd rather see this
implemented in a consistent way rather than hacked
in like in Zork III.

5. Seeing into other rooms

See Deadline (and a few other Infocom games) for
extremely simplistic examples of this. Only very
interesting if your world is "active", that is there
are things going on around you. After you've played
a game which supports this everywhere, traditional
games feel unreasonably claustrophobic.

Etc.

Interaction, as described for liquid and fire, refers
to the fact that the given event may have effects relating
both to the type of object something is, and to what it
is made out of. However, for a given game, you can
enumerate the fundamental sorts of things which can happen
to objects; then code for each class of objects and each
material the behaviors in response to certain stimuli.
Use of multiple inheritence, delegation, or forwarding
are all appropriate techniques to deal with this sort
of thing.

One final observation. A common problem with the
"graphic" adventure games is the "use object X
with object Y" command. The authors have generally
not considered every possible X vs. Y combination,
for there are far to many.

People seem to think simulation suffers from the
same problem. It suffers from a related one, not
the same one. In general, simulation offers an
improvement over the "use with" model, in fact,
in terms of combinatorial explosion.

Simulation is about defining a set of "laws of
the universe" as instantiated in some sort of
"simulation engine". Then objects are given
parameters that describe their behavior for
that engine. The engine then works out the
pairwise details from the parameters and the
laws.

The canonical example of this is in containers.
There is no need to determine for each pair
of objects whether one can be put in the other;
instead, objects are charcterized in terms of
the amount of space they have inside them, and
the amount of space they take up.

Careful planning and design reveals that all
of the above examples can avoid many-to-many
combinatorial explosions [1].

Sean Barrett
[1] With the exception of liquid mixing. If
you wish to allow liquid propogation, and you
have multiple types of liquids, you have a many
to many problem determining what happens when
two liquids mix, unless you want to attempt
to implement a large quantity of chemistry, and
characterize liquids by a vast number of chemical
properties (much less simulation at the molecular
level). One thing you can do, which is what
I did, is simply invent an entirely new chemistry
for your universe, which determines what happens
when liquids mix in a striaghtforward but
unrealistic way. Most of your players aren't
chemists anyway.

David Baggett

unread,
May 22, 1994, 2:05:49 PM5/22/94
to
In article <2rmvmj$n...@news.tamu.edu>, Sean Barrett <se...@stat.tamu.edu> wrote:

>I have been asked if I can name a game which is "more simulator" than
>traditional adventure games. I cannot name one; but I do not know what
>this proves.

It doesn't *prove* anything. But it suggests, to me, that either making
such a game is harder than you claim, or that the end product is less
desirable than you think (so no one wants to do it), or both. (Sure, you
can aruge that this "doesn't prove anything," but I've already given
arguments why I think it's not technically feasible to write a playable
game that simulates the world as deeply as you want. By the same token, we
could have a dialogue like:

S: It's not that hard to make a game where the characters converse with you
in prefect English -- just carefully special case things. Maybe I'm
just an exceptional programmer, but I know I could do it.
D: That sounds hard. People have been working on that for a long time.
Can you give an example of a successful system that does anything
close to this?
S: No, but that doesn't prove anything. Maybe the right person hasn't
come along yet, so no one's done it.

Obviously, this is not something you would argue, but I feel the same way
in the current argument that I would if we were having that one: not only
does the combinatorial explosion you have to deal with in your special case
coding render infeasible the level of simulation you were talking about
(flamethrower use), but I don't think deeper simulation necessarily
enhances the player's experience *any*. More on this to come.)

>Why is it that people seem to think that the
>
> SIMULATION
>
>which goes on in a traditional adventure game
>"doesn't count"

I don't think anyone thinks that. As you pointed out yourself, this is
degenerating into an argument about the extremes. At the risk of
belaboring the point, I'll try to explain in different terms what I think
makes good interactive fiction.

Simulation is one tool the IF author needs to use. It's especially
interesting in that it's one of the only tools that conventional authors
don't have. But there are many other tools that static fiction authors
*do* use, and hence have learned (over several millenia) how to use well.

Good programs tend to have lots of features. Where two programs are
equally easy to use, the one with more features is generally regarded as
the better one. Because of this, programmers tend to have "creeping
featuritis" -- the malady that convinces a hack that he need add "just one
more feature" for his program to be the best.

Good interactive fiction is NOT about adding features. Since we must (at
this point, at least) be programmers to write interactive fiction, we have
to be extremely careful not to fall into the trap of equating "more
features" with "better interactive fiction". My argument against the undue
emphasis on simulation (at its worst in pure-simulation, player-as-author
interactive fiction) stems from this. Simulation is only one aspect of
interactive fiction. Just as "adding more characters", or even "adding
more detailed characters" won't necessarily improve static fiction, I'd
argue that deepening the simulation, in addition to being a significant
amount work and impractical as you approach the extreme of molecular-level
simulation, isn't necessarily going to add anything to the overall quality
of the work.

The notion that making simulation more detailed will improve interactive
fiction is simplistic. IMHO, only one thing will improve interactive
fiction: authors getting better at using the tools in the interactive
fiction toolbox. Making the "simulation" tool better will only help if
authors are capable of *using* the added power. My feeling (as you'd say,
"from extensive experience") is that the existing "simulation tool" is
already more powerful than we can handle, so making it better won't do
anything good. (And let me qualify that by recognizing that there is a
difference between adding somewhat better handling of liquids and
supporting things like "torch * with flamethrower".)

The likely response is, "you really think that what we have now is the best
we need?" to which I'd have to say, for the most part, yes. I'd say the
same thing for parsing technology, and I've argued as much here before.
(Remember the now-infamous "take book from troll sneakily" adverb debate?)
The *real* problem is that some things (syntactic parsing) have proven to
be easier than others (knowledge representation and semantics), so we're
stuck with using a sort of brain-damaged toolbox. *Authors* need to bridge
the gap between such things as horrible knowledge representation and
parsers that can handle "which report did you file without reading?" by
skill alone. This is what makes wrting interactive fiction unexpectedly
difficult. Again, throughout all this, it is key to remember that we're
writing *fiction*, not *programs*.

>you're left with choose-your-own-adventures. You're writing a "menu
>maze"; each element of inter

This extreme is just as bad; no doubt about that.

>This simulation helps "immerse" the player, drawing
>them into the game, making them an active participant

Yes, but good static fiction authors don't put meaningless, pointless
details in their works. A good work of interactive fiction demands only as
deep a simulation as is necessary for the reader to suspend his disbelief.
It's clear from the success of even early games like "Colossal Cave" that
the simulation does not have to be that great for this to happen. Let me
reiterate: IMHO, more powerful simulation is only useful if the author is
capable of exploting the extra capabilities to improve the work in a
literary (not featuritis) way.

>Why is there a push to create more complex characters
>in interactive fiction?

I'm not sure -- I think this is yet another example where these problems
apply. What we desperately need is better knowledge representation, so the
machine, rather than author, can bear the burden of making the characters
seem intelligent and the game world "seem more real". Until there is a
major breakthrough, interactive fiction will be stuck in its retarded,
quirky state, and authors will struggle to overcome the limitations.

Gareth Rees

unread,
May 23, 1994, 8:06:11 AM5/23/94
to
Pauline Benzies writes:
> The point in concealing clues is that they are *concealed*. Having a
> player who says "BURN ALL WITH FLAMETHROWER" and then complains that
> because they torched the Mona Lisa, they can't convince the police to
> let them out of the jail to finish the game, seems to have missed the
> point.

This is a fair point, but it seems to me that IF authors have to do more
than just make challenging puzzles; they also have to try to produce a
game that sustains the player's interest long enough for her to want to
try to solve the puzzles.

A game which is easy to get into states from which you can't solve it
and in which it isn't easy to see that you've destroyed your chance of
winning is going to be frustrating (because all your saved files are
useless and you have to go back to the beginning and re-create them),
and if the player has to spend all her time wondering if the reason she
can't solve the current puzzle is because she didn't do something right
a hundred turns ago then she's more likely to give up.

[I don't mean to suggest that the flamethrower is an example of this
problem - I think it's obvious that torching everything is a bad idea.]

--
Gareth Rees

Andrew C. Plotkin

unread,
May 23, 1994, 11:39:00 AM5/23/94
to
Excerpts from netnews.rec.arts.int-fiction: 22-May-94 Re: Simulations
(was Re: Ch.. David Bag...@min.ai.mit (6451)

> Simulation is one tool the IF author needs to use. It's especially
> interesting in that it's one of the only tools that conventional authors
> don't have. But there are many other tools that static fiction authors
> *do* use, and hence have learned (over several millenia) how to use well.

> [...]

> Simulation is only one aspect of
> interactive fiction. Just as "adding more characters", or even "adding
> more detailed characters" won't necessarily improve static fiction, I'd
> argue that deepening the simulation, in addition to being a significant
> amount work and impractical as you approach the extreme of molecular-level
> simulation, isn't necessarily going to add anything to the overall quality
> of the work.

Hm. I would say that simulation is the root of interactive fiction --
the game simulates some part of reality. The very simplest games just
contain a set of rooms and objects, and simulate the actions of walking,
getting, and dropping; more complicated games get into containers,
liquids, generalized burning procedures, and all the other examples that
have flown by.

This is a difference in definition, though, I think. In practical terms
I agree with you; increasing the level of simulation doesn't necessarily
make a better game. As with any other kind of simulation (game,
scientific, etc), there is a optimum level of detail to get the effect
you want without wasting programmer time or CPU power.

On the third hand (gripping hand?) I would like to see games with more
simulation detail than has been "traditional" up to now. IMHO, *all
other things being equal*, more simulation would make a better game. I
can say this without contradicting myself because other things never are
equal. The programming involved would be a major effort, which would be
at the expense of other parts of the game. More significantly, many of
the plot constraints we're used to are partially supported by the
simulation constraints. For example, there are dozens of game puzzles
(Infocom and others) of the sort "you have two turns to do something,
then bam." That blatantly makes use of the nature of the parser; in
God's Perfectly Realistic Game Environment, you could probably cross the
room at a dead run, sweep both objects into a bag with one hand, and
kick the lever while diving out through the opposite door. Or whatever.
More detailed simulation vastly increases the player's flexibility,
which inevitably increases the number of possible solutions to
situations. If we can figure out how to write good games on that sort of
basis -- and I don't think anyone has, certainly not me -- there would
be some awfully impressive results.

Whew. End visionary mode. I don't know if the system I'm thinking of is
practical on today's home machines. But it can't be too far off, given
the CPUs and storage media that are running around.

--Z

"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."

Michael Booth

unread,
May 23, 1994, 1:45:20 PM5/23/94
to

Sean Barrett said:
> 3. development of complex AIs promotes use of
> AIs in simulation

In fact, this is my primary motivation for focusing on the 'simulation'
aspect of IF. In order to make a NPC behave reasonably, it needs to
have information about the world it is in, and how that world works.
The same information (to some extent) needed to make the game interface
more realistic to the players. A nice symmetry there.

Mike

--
| Michael S. Booth (bo...@cs.uiowa.edu) |
| Computer Science Graduate Student, Univ. of Iowa, Iowa City, IA |
| (An Engineer pretending to be a Computer Scientist) |

Michael Booth

unread,
May 23, 1994, 1:28:32 PM5/23/94
to
In article <1994May21.1...@cs.tcd.ie>,

Russell Wallace <rwal...@cs.tcd.ie> wrote:
>scy...@u.washington.edu (The Grim Reaper) writes:
>
>>In article <1994May19....@cs.tcd.ie>,
>>Russell Wallace <rwal...@cs.tcd.ie> wrote:
>>>[Comments about being able to blow things up deleted...]
>>>
>>[deletia]
>>>(even if it was only provided for one puzzle and is only useful there)
>>>or cutting things up with an axe (ditto) would do far more to make the
>>>game believable, than the ability to type TELL THE ROBOT TO BURN ALL THE
>>>BOOKS EXCEPT THE BLACK AND RED ONES or suchlike; if people are prepared
>>>to spend huge amounts of effort on the latter, why not the former? It
>>>needn't be *that* difficult... just record what material everything is
>>>made of, and have a section of code that says objects made of paper will
>>>burn easily (e.g. with a match), objects made of wood will burn with
>>>difficulty (e.g. with a flamethrower), metal and stone won't burn at
>>>all...

>>Sheesh, you have got to be kidding. Implementing any sort of vaguely
>>realistic physics into an i-f game is difficult, if not impossible. Just
>>saying that wood objects will burn under certain conditions isn't enough.

>>We'll have to change the description, possibly create ashes,
>>alter the object's
>>weight, change NPC's reactions to the object, possibly have certain puzzles
>>change (a bucket with the bottom burnt out isn't going to be much use), check
>>if the object's containing anything so that we can see if those things burn,
>>etc, etc. Physics is a much harder problem than parsers, which is why we
>>see more of the latter than the former.

>Only if you demand perfection or nothing.

You need to think this through a bit farther. The problem is, in fact, that
it _is_ almost perfection or nothing. To make supposedly simple things work
for any _reasonable_ number of situations, you need a very complex system.

>That attitude to parsers
>would have nobody ever write one better than GO NORTH because it
>wouldn't be able to understand TO BE OR NOT TO BE, THAT IS THE QUESTION,
>so why bother? For wood and paper objects burning, it suffices for game
>purposes to simply make them disappear.

Perhaps in some games, but most definitely not most. Even so, there is
_still_ the problem of what effects such disappearing objects have on
the rest of the world/game.

The problem is not hacking one special case thing, like you mentioned
with the burning materials, but the _interaction_ of even a few of these
special case hacks. Very quickly the hacks used to make the special
case hacks work together becomes a tangled mess.

My suggestion is, if it is so easy, build one. We'd all love to see it.

Phil Goetz

unread,
May 23, 1994, 2:03:47 PM5/23/94
to
In article <2rl0gv...@life.ai.mit.edu>,

David Baggett <d...@ai.mit.edu> wrote:
>In article <1994May19....@cs.tcd.ie>,
>Russell Wallace <rwal...@cs.tcd.ie> wrote:
>
>And then do what, change the description of the object to be "The X looks
>burned?" How about if the player torches himself or another actor?

*** talking about Inmate again mode ON ***

In _Inmate_, burnt is generally equivalent to broken. If the property
"broken" is assigned to an animate object (person, dog), the adjective
"dead" is prepended instead. Animate objects can only do things when
they are not broken. So you can kill people in _Inmate_, and they will
stop moving around. More to the point, you could (I suppose, though I
don't remember if people are burnable) burn a person, and the physics
would _automatically_ infer that that person was dead, and they would
stop whatever they were doing.

*** _Inmate_ mode off ***

>The whole point is that the player wants to be able to burn everything in
>sight because he wants the game to model the world realistically. No code
>that checks an object's general composition and sets a flag "burned = true"
>is going to be realistic.

Better than saying "I don't think that's a good idea."

>This gets back to the basic "realism vs. playability" argument. Though
>many people think this kind of realism is good for IF, I'd argue that the
>end result is a poorly designed game which is no fun at all. (You torched
>that newspaper early on? Too bad, you needed to read that to find out
>about the secret headquarters on Galapaos Island... But I won't tell you
>that, because in a "realistic" simulation you don't get this kind of
>feedback.) Ugh.

If you burn a newspaper without reading it in IF, you get what you deserve.

>In real life, doing anything nontrivial is a complete pain. (Oh geez, how
>the hell am I going to get to *Galapagos Island*? Gotta call a travel
>agent. And how am I going to pay for this? And what about my chemistry
>final?). Consequently, trying to make a game model the real world is
>actually a *bad* idea -- what you want is a game world that amplifies the
>"highs and lows" and glosses over the boring stuff. Impressionism, not
>realism, is the traditional basis for good games.

I think you're confusing a realistic PLOT with a realistic PHYSICS.
I wouldn't read a book about my life, because it would be boring.
I like books and movies about highly improbable, unrealistic stories.
But I want the world they take place in to behave orderly.

Phil go...@cs.buffalo.edu

Sean Barrett

unread,
May 23, 1994, 7:26:51 PM5/23/94
to
I've never much cottoned to people who respond to
critics with, "Do you think you can do better?"

Clearly, the ability to do something and the ability
to determine whether that thing is done well are
not one and the same. Often times (such as in the
case of movie criticism), the complaint is pointless,
as the critics clearly have neither the money nor
the time.

Michael Booth <bo...@grant.cs.uiowa.edu> wrote:
>The problem is not hacking one special case thing, like you mentioned
>with the burning materials, but the _interaction_ of even a few of these
>special case hacks. Very quickly the hacks used to make the special
>case hacks work together becomes a tangled mess.
>
>My suggestion is, if it is so easy, build one. We'd all love to see it.

Again, I'll say this. If you design in from the beginning
what the laws of the universe are, you will have an easier
time.

Returning to my original point above--(1) If this is a valid
argument, then this can clearly be given to anyone who posts
to the net with ideas about how to do more sophisticated
character AIs. I think it's a flawed, "status quo", argument.
Perhaps you don't; that's just a difference of opinion.

But my second response, again in light of what I said
above, is simple: Ok, how much will you pay me?

I've already designed much more advanced text simulations,
and my team got halfway through implementing the engines
(as opposed to the actual objects) before we various personal
circumstances forced us to disband. If we had been getting
paid to do it, I'm sure we would have stuck it out.

Of course, the system we were generating would best be
seen as a "research platform" w.r.t. adventure games,
since it was a MUD. Simulation without the plot (insofar
as we had gotten; we had to prove that the technology
was feasible). Given that we completed the design and
never ran into any serious problems in the implementation,
_I_ believe from my personal experience that it's achievable.

I don't, however, think it's the sort of thing one person
will do in his spare time. If anything, our failure was
in trying to do it all, to push every simulation boundary
we could think of.

Sean Barrett

Jason Noble

unread,
May 23, 1994, 8:11:26 PM5/23/94
to
In article <YhsAsI600...@andrew.cmu.edu> "Andrew C. Plotkin" <ap...@andrew.cmu.edu> writes:

>More detailed simulation vastly increases the player's flexibility,
>which inevitably increases the number of possible solutions to
>situations. If we can figure out how to write good games on that sort of
>basis -- and I don't think anyone has, certainly not me -- there would
>be some awfully impressive results.
>
>Whew. End visionary mode. I don't know if the system I'm thinking of is
>practical on today's home machines. But it can't be too far off, given
>the CPUs and storage media that are running around.

I agree with what you're saying about simulation, Andrew, but I feel
compelled to jump in on that last paragraph. Not being a hardware expert,
people should feel free to jump in and tell me I'm wrong, but it seems to
me that hardware is not the limiting factor here.

Forgive the PC bias, but with 486's and beyond becoming increasingly common,
with 8 or 16 MB of RAM and effectively limitless secondary storage (CD-ROM
or big hard drives), I don't think it's the hardware that's holding
interactive fiction back.

As Dave Baggett said earlier in this thread, it's the ability of authors /
programmers to utilise and flesh out the medium that is the limiting factor.
(And I know you said this too, Andrew; I'm not disagreeing with you.) Dave
cautioned against bothering with increased simulation power in games when
it's so hard to get a decent plot / story happening with the level of detail
we have now. As much as I'm a fan of simulation, I fear Dave may be right.
If you think about the fact that Zork (and the rest) ran on tiny machines
like the C-64, TRS-80, etc., and then think about the fact that essentially
similar games are now running on computers 100 times as big / powerful,
well....

(Now Phil Goetz will step in and tell me just how long a SNEPS-based game
would take to run on a PC).

Greg Ewing

unread,
May 24, 1994, 1:33:47 AM5/24/94
to
In article <2rjv83$h...@news.u.washington.edu>, scy...@u.washington.edu

(The Grim Reaper) writes:
|> Physics is a much harder problem than parsers, which is why we
|> see more of the latter than the former.

The real problem, I think, is not that physics is too hard to
simulate. It's not easy, but it wouldn't be excessively hard
to simulate more physics than adventure games currently do.

The real problem is to describe the results to the player in
a literary manner.

It's not hard to arrange for all objects that should be
burnable to be burnable. But if all you see when you
burn most things is "The x burns" then burning things
gets boring very quickly.

To make the experience satisfying in the literary sense,
you need to see things like "The paper napkin flares briefly
into flame, leaving a wisp of ash." "The wet match gives off
a little smoke, but not much else." "The petrol-soaked wood
pile explodes with an enormous ball of flame, nearly setting
your hair on fire."

In other words, if boring repitition is to be avoided, the author
must hand-craft an evocative description of each possible event,
individually. Until that can be automated, the "fiction" and
"simulation" sides of IF will remain difficult to reconcile.

|> +----------------------------------------------------------+
|> | The Grim Reaper (Reaper of Souls, Stealer of .sigs) |
|> | scy...@u.washington.edu |
|> +----------------------------------------------------------+

Greg Ewing, Computer Science Dept, +--------------------------------------+
University of Canterbury, | A citizen of NewZealandCorp, a |
Christchurch, New Zealand | wholly-owned subsidiary of Japan Inc.|
gr...@cosc.canterbury.ac.nz +--------------------------------------+

Sean Barrett

unread,
May 24, 1994, 1:49:40 PM5/24/94
to
Greg Ewing <gr...@huia.canterbury.ac.nz> wrote:
>The real problem is to describe the results to the player in
>a literary manner.
>
>It's not hard to arrange for all objects that should be
>burnable to be burnable. But if all you see when you
>burn most things is "The x burns" then burning things
>gets boring very quickly.

> GET STICK
You get the stick.

> GET MATCH
You get the match.

> GET PAPER
You get the paper.

> BURN STICK
The stick burns.

> BURN MATCH
The match burns.

> BURN PAPER
The paper burns.

Sure, it's a little stupid sounding, but so
is "get" and "drop".

Of course, most games replace "You get the stick"
with "Ok."

The idea is that getting/dropping are fundamental
actions in the universe, with one very standard
"outcome" for the normal cases.

You make burning things that fundamental, it's
acceptable to replace it with "Ok." as well.

I think if you actually think about most of
the objects in a typical adventure game, few
of them need to burn in a significantly different
manner.

Yes, the results will not be very "literary".
But neither is the result of most of the
fundamental actions built into a game.

Or perhaps I'm not being fair; perhaps you code
games to have every exit have a different message
for your movement, and every object has a
different message when you pick it up and drop it,
and every pair of object-container has a different
message when you put things in and take them out.

In which case I don't think the addition of a
burn message for every burnable object is a big deal.

>In other words, if boring repitition is to be avoided, the author
>must hand-craft an evocative description of each possible event,
>individually. Until that can be automated, the "fiction" and
>"simulation" sides of IF will remain difficult to reconcile.

Certainly this is true, in some sense. But I don't think people
mind the "boringness" of the core simulation, especially if
their primary goal is to achieve some particular effect;
then they may want "just the facts, ma'am".

Sean Barrett

Michael Booth

unread,
May 25, 1994, 1:49:55 PM5/25/94
to

This is precisely where Mr. Barrett and I differ. I agree completely
with Greg Ewing, in that the most difficult part of the simulation
game is generating decent prose.

It is not neccessarily acceptable to just have the game say 'OK' when
something is dropped, taken, burnt, and so on. That is your opinion,
quite obviously, but it presents a rather bleak atmosphere. I would
rather see something like:

> BURN MATCH
The match sputters to life with a sharp hiss, the smell of sulfur
filling your nostrils.

> BURN PAPER WITH MATCH
The match ignites the paper, which curls and blackens as the flame
devours it.

Much more colorful, in my opinion.

>I think if you actually think about most of
>the objects in a typical adventure game, few
>of them need to burn in a significantly different
>manner.

A match ignites and burns very differently from a candle, which burns
differently from a log, which burns differently than the Evil Chancellor
tied to the stake. How about the wooden mug with the tin handle?
The handle doesn't burn...

>Yes, the results will not be very "literary".
>But neither is the result of most of the
>fundamental actions built into a game.

Yes and no. That entirely depends on the type of game experience you
are creating. As you said in another post, you are from the MUD arena,
and it shows. Not that MUDs have to be any different that 'regular'
single person IF. In fact, they should be much better due to the
interactions of real people. However, this has not been the general
case, in my experience. If your only plot is how to kill that nasty
goblin in the Dungeon of Dreck, who will re-appear twenty minutes after
he has been dispatched, then simplistic prose is fine.

>[...deletia...]


>
>>In other words, if boring repitition is to be avoided, the author
>>must hand-craft an evocative description of each possible event,
>>individually. Until that can be automated, the "fiction" and
>>"simulation" sides of IF will remain difficult to reconcile.
>
>Certainly this is true, in some sense. But I don't think people
>mind the "boringness" of the core simulation, especially if
>their primary goal is to achieve some particular effect;
>then they may want "just the facts, ma'am".

This is true also. No need to overstate everything with flowery
language. However, there is a happy medium.

The biggest problem then becomes not the simulation system
(which is formidable in itself), but the natural language generation
system, which is as tangled as the english language. Just spitting
out canned phrases is not what I'm referring to here, but rather having
the system generate interesting prose describing the environment, such
as a portion of a room description, or the description of an object.
Something that is NOT a series of disjoint sentances like 'You are a strong
human male. You are tired. You are carrying a burned torch, a sword,
and a gem. You have a scar on your left cheek. You are not feeling very
well.'

Bill Burdick

unread,
May 24, 1994, 5:41:30 PM5/24/94
to
>>>>> On Mon, 23 May 1994 18:03:47 GMT, go...@cs.buffalo.edu (Phil Goetz) said:
Phil> Nntp-Posting-Host: hydra.cs.buffalo.edu

Phil> In article <2rl0gv...@life.ai.mit.edu>,


Phil> David Baggett <d...@ai.mit.edu> wrote:
>In article <1994May19....@cs.tcd.ie>,
>Russell Wallace <rwal...@cs.tcd.ie> wrote:
>
>And then do what, change the description of the object to be "The X looks
>burned?" How about if the player torches himself or another actor?

In my old MUD, we had objects called forwarders that would encapsulate
their targets and intercept all of the messages to them (delegating
most of the messages). When we wanted to model broken things, burned
things, etc., we made a forwarder that changed the target's behavior
to reflect the new property. When the previous condition was
restored, we simply removed the forwarder. In a nutshell, forwarders
slap extra behavior and state on to an object (this behavior can also
override current behavior) in an organized way. This is hard to do in
most OO languages, but Sludge (courtesy of Sean Barrett) and Self
support this very well (at the time, we were using my own version of
Smalltalk {Mob} with explicit support for forwarders).
--

-- Bill Burdick
bur...@ars.rtp.nc.us

Steven Hugg

unread,
May 25, 1994, 6:01:31 PM5/25/94
to
"Andrew C. Plotkin" (ap...@andrew.cmu.edu) wrote:
:
: On the third hand (gripping hand?) I would like to see games with more

: simulation detail than has been "traditional" up to now. IMHO, *all
: other things being equal*, more simulation would make a better game. I
: can say this without contradicting myself because other things never are
: equal. The programming involved would be a major effort, which would be
: at the expense of other parts of the game. More significantly, many of
: the plot constraints we're used to are partially supported by the
: simulation constraints. For example, there are dozens of game puzzles
: (Infocom and others) of the sort "you have two turns to do something,
: then bam." That blatantly makes use of the nature of the parser; in
: God's Perfectly Realistic Game Environment, you could probably cross the
: room at a dead run, sweep both objects into a bag with one hand, and
: kick the lever while diving out through the opposite door. Or whatever.
: More detailed simulation vastly increases the player's flexibility,
: which inevitably increases the number of possible solutions to
: situations. If we can figure out how to write good games on that sort of
: basis -- and I don't think anyone has, certainly not me -- there would
: be some awfully impressive results.
:
: Whew. End visionary mode. I don't know if the system I'm thinking of is
: practical on today's home machines. But it can't be too far off, given
: the CPUs and storage media that are running around.

I've been thinking along the same lines. A text adventure could
incoprorate a knowledge engine, like Cyc -- which IMO is not powerful enough
to embody actual intelligence, but could be put to use to create an
entertaining game. We don't have to simulate every detail, but we just
have to LOOK like we're simulating every detail -- we'd just have the
notion in the knowledge base that given a particular state of the
universe, the following things COULD happen ... and so we'd sort of
randomly generate the universe as we go along. Say, you're in a hotel
lobby. What's the probability that there's a desk in this lobby? Pretty
high; let's put one there. What else? People work at a hotel, and some of
them sit behind the front desk. Make two people. And so on.

I agree the time of the text adventure has past us, and that a system
like this would help to revive it.

--
Steven E. Hugg
hu...@cs.fsu.edu

Jamieson Norrish

unread,
May 26, 1994, 8:02:40 AM5/26/94
to
In article <2rmdc0$p...@harbinger.cc.monash.edu.au> benziesp@ponderosa
(Ms. Pauline Benzies) writes:

[Burning newspaper which contains clue to finish adventure.]

>And in a "well-designed" game of this sort, you either have other
>ways to find out, or you don't need to get there to "win" the
>game.

Why is this different to publishing that you only need to type
'IDDQD' in Doom and suddenly none of your stupid actions can kill
you. Just charge on through ignoring the bad guys and push the
button marked "END".

It is different in that it allows for "mistakes", and it also avoids
railroading the player along a certain path - which for its own sakes
is a good thing. It is not removing puzzles, it is simply requiring
that the player find a different set, by removing one lot of problems.

>to figure out how to advance it). Pure simulation allows
>the player to make up the story entirely on their own.

At this point, the game author ceases to be an *author*, don't
they?

Not entirely. It is the author who decides what is included in the
game, what premises and laws it is based on, the situation at which
the simulation begins, and so forth. I think that simulations can be
games, and that the author can have a large say in what happens, by
defining and limiting the world in which the game is set.

>Perhaps, ideally, the optimum is somewhere in the
>middle; perhaps there is no optimum, there are simply
>different sorts of games.

I tend to agree with this last, that the two forms are not
compatible

I do not think that that is what was being meant. The two forms are
compatible, and that combining the two "ingredients" in different
measures results in different types of games - just as changing the
genre will change the game.

>Each time you write a behavior for an object, you ask
>yourself, "Can they do this if it's burning/burnt/soaking
>wet?" and adjust the code as appropriate.

Lets see, how do I code the "broken" bottle. Obviously, if they
used "BREAK BOTTLE CAREFULLY" there will be a clean edge that can
be used to cut. On the other hand, "BREAK BOTTLE WITH AXE" tends to
leave long splinters.

The coding for something along these lines would be hideous, of
course. But that is not to say that it is impossible, or worthless.

Jamie

Jamieson Norrish

unread,
May 26, 1994, 8:19:06 AM5/26/94
to
In article <2rlmkn...@life.ai.mit.edu> d...@min.ai.mit.edu (David
Baggett) writes:

>>This gets back to the basic "realism vs. playability" argument.
>Misleading. Certainly you must work at it to make it playable.
>But it's not a tradeoff; if you want more realism, you have to
>work harder to make it playable; you don't give up playability.

First off, cite a *single* example of a successful work of
interactive *fiction* that adheres to the requirement that it be
this realistic -- even a prototypical one. If you can't, explain
why there isn't one yet. (This is a genuine request, not just a
challenge.)

Whoa, here! From what you quoted, I can see nothing relating to a
"requirement" that a game be "this" (I assume this refers to being
able to burn everything in a game with a flamethrower) realistic.
Rather, all that is being said is that increasing realism does not
necessarily mean a decrease in playability.

The more realistic your game is, the more things "should work" to
solve "the puzzles" and advance "the plot". The *very idea* of a
plot is unrealistic (modulo ideas about predestination which I
suppose we could debate as well). Dramatic pacing is an unnatural
notion. Yet these things are essential, I would argue, to a
"fulfilling" experience.

This is nonsense. Is "plot" necessarily unrealistic, given that
realism is a very variable thing? Can not plot and dramatic pacing be
accomplished with realism? I think it can, if only by intervention of
other characters and events which the author creates.

>As you allow those laws to become more complex, the system becomes
>more and more of a simulation.

...and less and less of a *game*, which is what I was trying to get
it. Yes, it comes down to where you draw the line, which is a
matter of personal taste. But in the extreme (which is where I put
allowing the player to torch everything in the game with a
flamethrower), I think the desire to make games "realistic" is
counterproductive.

Just why does it become less of a game? And why is being able to torch
everything counterproductive? Certainly not counterproductive to the
work being a game, as I understand the term.

Jamie

Jamieson Norrish

unread,
May 26, 1994, 8:25:54 AM5/26/94
to
In article <2rmcq9$o...@harbinger.cc.monash.edu.au> benziesp@ponderosa
(Ms. Pauline Benzies) writes:

Okay, I will. The author anticipated that by composing the chandelier
of diamonds, and not including the physics which would make most of
those diamonds be smashed (and therefore massively reduced in value)
on the floor. Probably also by including in the game characters who
would get rather upset at vandals destroying their valuable property.

Regardless, this is a fatuous argument. Are you really suggesting that
the addition of something which has as one of its "cons" the
requirement that the author actually think is a bad thing? I would
hope not.

Jamie

Jamieson Norrish

unread,
May 26, 1994, 9:09:44 AM5/26/94
to
In article <2ro6tt...@life.ai.mit.edu> d...@min.ai.mit.edu (David
Baggett) writes:

>This simulation helps "immerse" the player, drawing
>them into the game, making them an active participant

Yes, but good static fiction authors don't put meaningless,
pointless details in their works. A good work of interactive
fiction demands only as deep a simulation as is necessary for the
reader to suspend his disbelief.

One of the great differences between fiction and interactive fiction
is that the player has no requirement to look at everything that is
present in the work (whereas most people would feel obliged to read
the entirety of a book, for example). No one forces a player to go
around with a flame-thrower burning everything, and most people won't,
but that does not mean that having the option is a bad thing. I for
one get a great enjoyment out of discovering what such things I as a
player can do in a game - I spent a long time in Zork II trying to
discover if exploding the brick in any other places would have an
interesting effect. Does this make it less interactive fiction, and
more a simulation? I don't think so.

Also, why should authors of interactive fiction limit themselves to
what is demanded by the perception of what makes a good game. Space
invaders was considered a great arcade game when it was first
released, but now it's a primitive relic. Saying that it provided all
that was demanded of a good game does not mean that other arcade games
should not aspire to something different. The same certainly applies
to interactive fiction.

Jamie

Russell Wallace

unread,
May 26, 1994, 5:12:17 PM5/26/94
to
d...@min.ai.mit.edu (David Baggett) writes:

>In article <JAMIE.94M...@akeake.its.vuw.ac.nz>,
>Jamieson Norrish <ja...@akeake.its.vuw.ac.nz> wrote:

>>I for one get a great enjoyment out of discovering what such things I as a
>>player can do in a game - I spent a long time in Zork II trying to discover
>>if exploding the brick in any other places would have an interesting
>>effect.

>You seem to agree that, while amusing, such things provide little in the
>way of artistic merit. Is it worth spending enormous amounts of time
>adding something silly to the work, even when this has no effect on the
>overall quality? In a perfect world, we'd be able to put an infinite
>amount of work into our games. But since we have only a finite number of
>hours to work on these games, we have to prioritize things. I'd assign
>"burn X with flamethrower" and other "deep simulation" things very low
>priorities. Sure, they're cute. But at best they give the player a
>chuckle. Shakespeare didn't spend 90% of his time on Romeo and Juliet
>writing the little jokes the soldiers tell each other in the beginning --
>they're were other "higher priority" aspects of the work.

>>Does this make it less interactive fiction, and more a simulation?
>>I don't think so.

>I still claim that pointless details *detract* from the work. Supporting
>actions that can radically alter the game world (like burn X with
>flamethrower) undermines the author's need to impose a plot on the player.
>Without a plot, you (the player) just wander around aimlessly. As I
>mentioned before, I don't think this kind of "art" is accessible to most
>people.

>I'll work this example out a little further, since it seems to be confusing
>people. If you allow "burn X with flamethrower", it's going to be hard to
>guarantee that you (the author) have anticipated all the possible ways the
>player could overcome an obstacle. Overcoming obstacles is the basic way
>we advance the plot in these games. The obstacles provide plot checkpoints
>that the author needs. (Example author's dialogue: "OK, I want this to be
>a dramatic moment, so I don't want the player to get into the dragon's
>chamber until the lamp's burned out; that way, the player will be surprised
>when the dragon's fire lights up the room.")

>In a completely realistic system, the author can plan no such thing. The
>lamp won't necessarily be out, because the player could have gone to the
>hardware store and bought kerosene with money he's earned by mowing lawns.
>Or a zillion other sensible things.

>We need to limit what things "work" so that the author has a tractable list
>of player actions to consider. Even now, we miss some good ones, as the
>fine playtesters in TABU will tell you. With Legend, which has the
>standard amount of "realism/simulation", I had a heck of a time
>guaranteeing plot points. It is by no means easy, even within a system
>where almost nothing "works," to force the player to follow a particular
>path through the game. (I don't mean move-for-move, I mean something along
>the lines of "I want the player to meet Doctor Trolovich before going in
>the inn, because I want the people in the inn to give the player
>information about Trolovich".)

This I think is going too far. Remember it's supposed to be
*interactive* fiction, i.e. the player is supposed to have meaningful
choices to make. If you, the author, have predetermined every plot
point in advance, then the player might as well just read a transcript
of someone else solving the game instead of going through the motions of
playing it (other than the intellectual challenge of solving each
puzzle; but that's a form of entertainment akin to a crossword rather
than literature). Or read a good book, which as literature will be far
better than any adventure game will.

The whole point of having interactive fiction instead of just having
non-interactive fiction is if the plot can vary depending on the
player's actions. And the best way to do that is to allow things in the
game to interact in ways that the author may not have foreseen, so that
the author is not faced with the impossibly large task of creating by
hand all the alternative plots.

thomas.j.epstein

unread,
May 19, 1994, 9:54:25 AM5/19/94
to
In article <2r9t9d$s...@sunb.ocs.mq.edu.au> jno...@laurel.ocs.mq.edu.au (Jason Noble) writes:

[Deletions]
>However, I thought I'd try to start some discussion on one particular aspect
>that's plaguing me: the "historical" period of the game. In short, whether
>it's set in a (modified) past, present or future.

[Discussion of multy-use objects and genre-related problems deleted.]

>I know that many of you out there are working on games right now. I'm not
>trying to steal your secrets, but is there anyone who'd like to share with
>the newsgroup their approach to this particular IF problem?
>
>Regards,


>
>---------------------------------------------------------------------------
>Jason Noble | jno...@bunyip.bhs.mq.edu.au
>National Centre for HIV Social Research | jno...@laurel.ocs.mq.edu.au
>Macquarie University, Sydney, Australia | ph. (61 2) 850 8667
>---------------------------------------------------------------------------

Jason,

Have you considered a fantasy setting other than Swords & Sorceryesque
type IF? Fantasy allows you to write all the rules for your universe.

For example, in a TRON-like universe *everything* can be as limited
or open-ended as you are inclined to design. You are familliar with
TRON? :-)

I'd be interested to see an adventure where the player is sleeping and
interacts with his/ her dreams; trying to solve the riddle of the
character's own sub-conscious nightmare.

--Tom

Steven Hugg

unread,
May 26, 1994, 6:35:00 PM5/26/94
to
David Baggett (d...@min.ai.mit.edu) wrote:
: In article <2s0hrr$8...@mailer.fsu.edu>, Steven Hugg <hu...@cs.fsu.edu> wrote:
: >I agree the time of the text adventure has past us, and that a system

: >like this would help to revive it.

: If it's true that text adventures have no more public appeal, I don't think
: that making them better will actually cure that. The problem really seems
: to be that game players are not readers -- the games that sell well are
: more like movies than books. I'm not sure if this is a sign of something
: significant, or whether it's just a by-product of poor marketing of games
: to book-lovers.

Don't ignore the popularity of MUDS, which seem to be the Zork substitutes
of the '90s. There are so many people adding on to the things with rooms
and objects -- some MUDS have a server dedicated to them, and handle 200
players at a time! So obviously the text adventure is not dead, it's just
"hiding" in the bowels of the Internet...

Tho MUDS aren't IF really... but if you make a text adventure too
realistic does it cease to be fiction?

Ms. Pauline Benzies

unread,
May 26, 1994, 7:27:22 PM5/26/94
to
In article <JAMIE.94M...@akeake.its.vuw.ac.nz> ja...@akeake.its.vuw.ac.nz (Jamieson Norrish) writes:
>Okay, I will. The author anticipated that by composing the chandelier
>of diamonds, and not including the physics which would make most of
>those diamonds be smashed (and therefore massively reduced in value)

Putting aside the notion that diamond is one of the toughest substances known
to man and is unlikely to be "smashed" -- at least, thats the sort of physics
that I understood...

>Regardless, this is a fatuous argument. Are you really suggesting that
>the addition of something which has as one of its "cons" the
>requirement that the author actually think is a bad thing? I would
>hope not.

The point is that the author expected the player to have to find a knife
somewhere, not use the "sharp" property that the bottle picked up because it
was also "broken".

Remember, this was initially suggested to counter the argument that "broken"
== "not-usable".

David Baggett

unread,
May 26, 1994, 12:14:46 PM5/26/94
to
In article <JAMIE.94M...@akeake.its.vuw.ac.nz>,
Jamieson Norrish <ja...@akeake.its.vuw.ac.nz> wrote:

>I for one get a great enjoyment out of discovering what such things I as a
>player can do in a game - I spent a long time in Zork II trying to discover
>if exploding the brick in any other places would have an interesting
>effect.

You seem to agree that, while amusing, such things provide little in the


way of artistic merit. Is it worth spending enormous amounts of time
adding something silly to the work, even when this has no effect on the
overall quality? In a perfect world, we'd be able to put an infinite
amount of work into our games. But since we have only a finite number of
hours to work on these games, we have to prioritize things. I'd assign
"burn X with flamethrower" and other "deep simulation" things very low
priorities. Sure, they're cute. But at best they give the player a
chuckle. Shakespeare didn't spend 90% of his time on Romeo and Juliet
writing the little jokes the soldiers tell each other in the beginning --
they're were other "higher priority" aspects of the work.

>Does this make it less interactive fiction, and more a simulation?


>I don't think so.

I still claim that pointless details *detract* from the work. Supporting

Dave Baggett

David Baggett

unread,
May 26, 1994, 11:42:32 AM5/26/94
to
In article <2s0hrr$8...@mailer.fsu.edu>, Steven Hugg <hu...@cs.fsu.edu> wrote:
>I agree the time of the text adventure has past us, and that a system
>like this would help to revive it.

If it's true that text adventures have no more public appeal, I don't think


that making them better will actually cure that. The problem really seems
to be that game players are not readers -- the games that sell well are
more like movies than books. I'm not sure if this is a sign of something
significant, or whether it's just a by-product of poor marketing of games
to book-lovers.

My question is: who buys all these SF/Fantasy books? Why don't they buy
text games? Is there something inherently annoying about reading text on a
screen that puts these people off? Do they not know that such games
exists? Is it price? Is it that text games don't have the literary
credibility that books do, because most of them are silly?

There are clearly lots of people who read books, even though there are more
good movies out there than you could possibly watch.

Perhaps this has something to do with the game industry itself. I was
looking in the latest CGW (for the Rylvania review, yay!) and noticed that
they had an awards section. Looking over the categories and the comments
about the games, it struck me that no text game could ever win an award. I
remember having this same discussion with Ross Erickson, who runs Game
Bytes. His opinion was that a text game should compete with Day of the
Tentacle in the "Adventure" category. I just don't think you'd ever see a
text game beat DOTT, even if it were the interactive equivalent of King
Lear.

What's the point? Not that it's unfair -- that's pretty much to be
expected. It's that the situation shows that there's no place for games
for people who like to read (as opposed to games for people who like to
watch movies and TV) in the gaming community overall. You could trumpet
this as the beginning of the demise of civlization, but I don't think it's
anything so dramatic.

Perhaps it comes down to this: given currently technology, graphics games
can be more like movies than text games can be like good books. (I.e., if
you "solved AI" tomorrow, the situation might change.)

Michael Booth

unread,
May 25, 1994, 1:13:06 PM5/25/94
to
Sean Barrett <se...@stat.tamu.edu> wrote:
>I've never much cottoned to people who respond to
>critics with, "Do you think you can do better?"
>
>Clearly, the ability to do something and the ability
>to determine whether that thing is done well are
>not one and the same. Often times (such as in the
>case of movie criticism), the complaint is pointless,
>as the critics clearly have neither the money nor
>the time.

Yes, you are correct. It did seem to be a logical response to your
original post, however, since you claimed implemeting burning/burned
objects to be simple. It is not, unless you consider simply flagging
an object as 'burned' sufficient, which I don't.

>Michael Booth <bo...@grant.cs.uiowa.edu> wrote:
>>The problem is not hacking one special case thing, like you mentioned
>>with the burning materials, but the _interaction_ of even a few of these
>>special case hacks. Very quickly the hacks used to make the special
>>case hacks work together becomes a tangled mess.
>>
>>My suggestion is, if it is so easy, build one. We'd all love to see it.

First of all, I must admit my last statement was not necessary. Chalk it
up to lack of sleep that day. Sorry. However, my earlier statements
are valid.

>Again, I'll say this. If you design in from the beginning
>what the laws of the universe are, you will have an easier
>time.

Yes, completely true. In fact, I expect this is a given. The difficulty
is how are those laws defined within the computer such that all
(or a decent majority) of the interactions between objects in the world
are handled in a reasonable way?

>Returning to my original point above--(1) If this is a valid
>argument, then this can clearly be given to anyone who posts
>to the net with ideas about how to do more sophisticated
>character AIs. I think it's a flawed, "status quo", argument.
>Perhaps you don't; that's just a difference of opinion.

I think you misunderstood, due to the fact I was less than clear.

>But my second response, again in light of what I said
>above, is simple: Ok, how much will you pay me?

Nothing. It seems you are concentrating on my single, facetious
statement rather than the meat of my argument.

>I've already designed much more advanced text simulations,
>and my team got halfway through implementing the engines
>(as opposed to the actual objects) before we various personal
>circumstances forced us to disband. If we had been getting
>paid to do it, I'm sure we would have stuck it out.
>
>Of course, the system we were generating would best be
>seen as a "research platform" w.r.t. adventure games,
>since it was a MUD. Simulation without the plot (insofar
>as we had gotten; we had to prove that the technology
>was feasible). Given that we completed the design and
>never ran into any serious problems in the implementation,
>_I_ believe from my personal experience that it's achievable.

I think we are operating from different views of what a 'complete'
implementation is.

However, just to continue to be obnoxious :), how did your
system handle burning a whole building down? How about merely burning
a table which was decribed in the text description of the room?
Most MUD-like systems I have seen either don't actually have objects for
things seen in the room's static description, ignore changes to these
objects, or disallow actions on those objects. None of these are
satisfactory, in my opinion.

A system which can cope with such events is quite fascinating, and
extremely difficult to create.

>I don't, however, think it's the sort of thing one person
>will do in his spare time. If anything, our failure was
>in trying to do it all, to push every simulation boundary
>we could think of.

Although I think you said nothing about the argument I posed, I would
prefer it if we just let the matter drop. My response, though valid,
was written on a bad day, and I find this whole discussion a giant
black hole of wasted time. You are right, from your definition of
the problem, and I am right from mine. Either way, we both want the
same thing, and that is a realistic world with consistent physical
rules. Not perfect, just reasonably consistent.

Ms. Pauline Benzies

unread,
May 26, 1994, 7:35:09 PM5/26/94
to
In article <JAMIE.94M...@akeake.its.vuw.ac.nz> ja...@akeake.its.vuw.ac.nz (Jamieson Norrish) writes:
>It is different in that it allows for "mistakes", and it also avoids
>railroading the player along a certain path - which for its own sakes
>is a good thing. It is not removing puzzles, it is simply requiring
>that the player find a different set, by removing one lot of problems.

Actually, the last time I sprayed the office with a flame-thrower, my boss
fired me. I told him it was a mistake and that I should still be able to
finish off the rest of my employment. Curiously, he wasn't interested.

Why should the simulation let the player make destructive mistakes? In general
life doesnt! What are we simulating anyway? Remember, in life, when the old
man staggers up to you and hands you a document proving sinister machinations
in the political party of your choice that will blow the lid off the countries
government, you don't get the option of burning it and then looking around for
another way to do the same thing!

> Lets see, how do I code the "broken" bottle. Obviously, if they
> used "BREAK BOTTLE CAREFULLY" there will be a clean edge that can
> be used to cut. On the other hand, "BREAK BOTTLE WITH AXE" tends to
> leave long splinters.
>
>The coding for something along these lines would be hideous, of
>course. But that is not to say that it is impossible, or worthless.

I agree. However, it pushes this thread back into the dreaded "<verb> <noun>
<adverb>" discussion and thats the point that most people get off. Myself
included.

Mark Hughes

unread,
May 28, 1994, 11:39:16 PM5/28/94
to
On 29 May 94 11:57:46, Jamieson Norrish (ja...@akeake.its.vuw.ac.nz) spake:
: As for artistic merit, I
: don't see that this is necessarily harmed by having a greater level of
: simulation than is presently used. If you've experienced such, think
: of good roleplaying experiences, where the players are free to make
: all their choices, and the person running the game can only give
: sensory feedback, determine the outcome of actions, and manipulate the
: rest of the world in such a way as to provide "plot". And the players
: need never take up the plot strands they are offered, but rather
: create their own. Why be constrained to what the author has
: specifically set out. Surely presenting the world, and manipulating
: that, is better than manipulating the player(s)?

I agree - my vision of the perfect IF game is an electronic GM, capable of
running as good and interesting an adventure as a human GM. That's not possible
yet, but a lot of things can be done towards that goal: open environments with
several plots possible, actual campaign-type events, and NPCs who can be
interacted with semi-rationally and do things on their own. That's one of the
things I liked best about Seastalker - it was a simple game, but the NPCs
actually DID things on their own, and the interaction with them wasn't too bad.
It actually felt like a real universe, rather than one of those "Silent Earth"
-type movies where the world is deserted...

In a way, IF is really two entirely different fields - people like Dave want
to write Choose-Your-Own-Adventure stories with some artistic merit, and I want
to write adventure modules and campaign settings that don't require a GM. To
each their own, I guess - it's just convenient that both groups need the same
tools.

-Mark Hughes

David Baggett

unread,
May 27, 1994, 3:49:27 PM5/27/94
to
In article <1994May26.2...@cs.tcd.ie>,
Russell Wallace <rwal...@cs.tcd.ie> wrote:

>Remember it's supposed to be *interactive* fiction, i.e. the player is
>supposed to have meaningful choices to make.

Yes, but it's supposed to be interactive *fiction*, i.e., the author is


supposed to have meaningful choices to make.

:)

We could go on and on -- it's all a matter of perspective. One reason I've
come to view my bent as "better" is that it seems to me that the _fiction_
part needs the most work. The player does have choices to make in current
IF games. Most players do not seem to feel deprived when they discover
that they couldn't really radically affect the outcome of the game.

What is still missing are games that are good literature as well as being
fun. I'm not talking about the basic craft of fashioning sentences from
words, either -- many games around now are competently written. I'm
talking about the greater goals of literature: to communicate in a way that
is universal -- to deliver (a) message(s) in a way that transcends genre;
to create characters that *live*, because the reader sympathizes.

There are moments in existing IF. Many people feel like they "know" Floyd.
But we have yet to see a work that maintains this kind of intensity
throughout. Floyd is a brilliant stroke on an otherwise ordinary canvas.

>If you, the author, have predetermined every plot point in advance, then
>the player might as well just read a transcript of someone else solving the
>game instead of going through the motions of playing it (other than the
>intellectual challenge of solving each puzzle; but that's a form of
>entertainment akin to a crossword rather than literature).

You're reducing this case to absurdity. Although the overall plot of
Colossal Cave is predetermined, the player never feels as though he's
"trapped in a glorified novel". He feels like he's exploring a game world.
When things happen, they certainly don't "feel" predetermined (at least not
the first time through the game). I agree that such plotted games have
less replay value, but only to the extent that books do too. (You may say,
"Big deal, I know he's going to kill the king -- why bother?", but Hamlet
is still compellling the 100th time around.)

>Or read a good book, which as literature will be far better than any
>adventure game will.

It is truly maddening to me that most people here seem to be willing to
accept this without a fight. Interactive fiction is clearly a superset of
fiction -- you can just have the "player" read a novel on the screen.
Given this, IF works have the potential to be every bit as good as any work
of static fiction. (Yes, I realize it's an oversimplification, but you
can't disagree with it in the abstract!)

There are many ways you could try to bring "good literature" to IF. In
Legend, I've tried one particular approach: I've put in quite lengthy
sections of static prose in order to capitalize on the existing "literary
tradition", so to speak. That's where I develop the characters (since
"actors" are no more than cardboard cut-outs now), and, to a certain
extent, where I develop the overall theme of the work. It's clumsier than
I'd like, but as a first attempt at the method I think it works pretty
well. Perhaps once a few r.a.i-f readers have played through it, they'll
be able to explain where I'm coming from better than I (who seem to be
failing miserably).

>And the best way to do that is to allow things in the game to interact in

>ways that the author may not have foreseen..

This sounds good but is unworkable in practice. If the author hasn't
anticipated it, the prose resulting from the action will be lame. Good
text generation is years off. If there's one thing IF readers *have*
voiced concern about, it's prose quality.

Also, there still *is* a big difference between plotted (i.e., traditional
adventure game) IF and static fiction. When reading a book, you can learn
about people and witness events that make an indelible impression on you.
Plotted interactive fiction has the same potential, but you get to *meet*
the people and *live* the events.

This may sound overblown, but there really is truth to it. We recently got
email from a player who was quite upset about having to solve a particular
puzzle in the way it had to be solved. Without going into details -- I
respect this player's views and privacy -- it made absolutely clear to me
that even fairly straight-ahead IF games really can draw the player in, to
the extent that he feels he is *responsible* for his fictional character's
actions. This is a powerful medium indeed.

David Baggett

unread,
May 28, 1994, 5:07:26 PM5/28/94
to
In article <1994May28....@cs.tcd.ie>,
Russell Wallace <rwal...@cs.tcd.ie> wrote:

>Perhaps I expressed myself badly; I did not mean to imply that interactive
>fiction can never be as good as a novel, just that all current examples in
>fact are not as good.

Well, we shouldn't expect the best IF works to be as good as the best
novels (for a long while, at least), but I still think that most IF works
now are better than a lot of novels out there. (It's just that that's not
saying much.)

>This Legend game, is it currently available?

It's finished but I'm still in the beta-testing and bug-fixing phase. I'll
put a demo on ftp.gmd.de as soon as I'm sure it's production-quality.

>I would still rather be able to do something, even if the feedback I get
>is in unimaginative prose, than be unable to do it at all by arbitrary
>fiat.

I appreciate this point of view. I'm not sure that this kind of game would
go over well with people who are naive about the technology, though. I can
just see people saying things like, "It was a fun game, but a lot of the
time it sounded really computer-y."

Jamieson Norrish

unread,
May 29, 1994, 7:57:46 AM5/29/94
to
In article <2s2htm...@life.ai.mit.edu> d...@min.ai.mit.edu (David
Baggett) writes:

In article <JAMIE.94M...@akeake.its.vuw.ac.nz>,
Jamieson Norrish <ja...@akeake.its.vuw.ac.nz> wrote:

>I for one get a great enjoyment out of discovering what such
>things I as a player can do in a game - I spent a long time in
>Zork II trying to discover if exploding the brick in any other
>places would have an interesting effect.

You seem to agree that, while amusing, such things provide little
in the way of artistic merit.

I think that what made exploding the brick only an interesting and
amusing exercise, rather than something which really added to the
game, was that the effects of my actions were minimal. Blocking the
princesses path by exploding a room either made no difference, or made
the game insolvable (I can't remember which). The only unstable area
in the entire underground was in the location where the brick was
meant to be used. The demon did not make any attempt to stop the brick
exploding in the summoning room, or doing anything after it had gone
off. Had such things been accounted for, using the brick in different
places to achieve different things would have become far more a part
of the game rather than a cute diversion. As for artistic merit, I


don't see that this is necessarily harmed by having a greater level of
simulation than is presently used. If you've experienced such, think
of good roleplaying experiences, where the players are free to make
all their choices, and the person running the game can only give
sensory feedback, determine the outcome of actions, and manipulate the
rest of the world in such a way as to provide "plot". And the players
need never take up the plot strands they are offered, but rather
create their own. Why be constrained to what the author has
specifically set out. Surely presenting the world, and manipulating
that, is better than manipulating the player(s)?

I doubt, too, that players would wander around aimlessly - if they
could come up with their own things to do, there would always be the
author's intended plots to follow, which might later present
opportunities for the players to do their own thing.

Jamie


Bob Newell

unread,
May 28, 1994, 8:17:48 PM5/28/94
to
>>I agree the time of the text adventure has past us, and that a system

I don't agree with the original author's premise here. Let us rather say
that the time of mass/commercial appeal of the text adventure is past.



>If it's true that text adventures have no more public appeal, I don't think
>that making them better will actually cure that. The problem really seems
>to be that game players are not readers -- the games that sell well are
>more like movies than books. I'm not sure if this is a sign of something
>significant, or whether it's just a by-product of poor marketing of games
>to book-lovers.

I once posted something from a woman with whom I was attempting to
collaborate on a text game; this collaboration has now revived and I have
another long posting to make from her if I can locate the text of it. But
she and I seem to have this discussion all the time. She has now played the
Rylvania demo, and some of Klaus and Bureaucracy. Her (unfavorable)
impressions haven't changed much; she compares text games to fiction she has
read and it comes up short in her mind. She is truly a book lover to whom
text games have not been marketed well.

>My question is: who buys all these SF/Fantasy books? Why don't they buy
>text games? Is there something inherently annoying about reading text on a
>screen that puts these people off? Do they not know that such games
>exists? Is it price? Is it that text games don't have the literary
>credibility that books do, because most of them are silly?

I wonder if she would say "yes" to all of the above. The typical text game
player simply _won't_ read a lot of text on screen if all it does is provide
color and atmosphere. He (she) is too busy looking for clues and
descriptions, and objects to examine, take, and use. Small wonder there is
no literary credibility- it's the paradigm that's at fault, I believe.

Text adventures have developed in certain directions:
-they rely on the examine-take-use model
-they are predicated on puzzle solving
-they rarely have a serious theme, and even if so (Trinity, for instance)
they are presented in a less-than-serious, unrealistic manner (I don't mean
here realism=simulation; but my main objections to Trinity are foolishnesses
such as paper birds turning into spacecraft).
-the quality of the prose is generally not what it could or should be.

(Not that I could do any better myself, mind you....)
These are the things we come to expect in a text adventure; these are the
things that when we write, we tend to emulate. Breaking the paradigm is
required to move the text game into new fields of accomplishment. This
raises all of the issues of playability, interest, etc., that we've
discussed here so often. No, electronic page turning is not very
interesting. Yes, reading on screen is inherently less satisfying than
reading on paper...

>Perhaps it comes down to this: given currently technology, graphics games
>can be more like movies than text games can be like good books. (I.e., if

This is a very good, accurate, and useful insight. Perhaps the answer is
to then explore what it would take to make text games more like good books.
Solving the AI-complete problem is not it; graphics games fall even shorter
than text games here, I think. Again, I believe it comes back to our basic
paradigm for the text game.

I'd like to mention as an example a non-text game which I just began
exploring, "Myst". Graphics and music are used to create a startling sense
of place and atmosphere. But what is so fascinating is how Myst breaks many
of the old paradigms of a game of its type:

Russell Wallace

unread,
May 28, 1994, 10:11:39 AM5/28/94
to
d...@case.ai.mit.edu (David Baggett) writes:

>In article <1994May26.2...@cs.tcd.ie>,
>Russell Wallace <rwal...@cs.tcd.ie> wrote:

>>Remember it's supposed to be *interactive* fiction, i.e. the player is
>>supposed to have meaningful choices to make.

>Yes, but it's supposed to be interactive *fiction*, i.e., the author is
>supposed to have meaningful choices to make.

>:)

Yep... It's a bit like creating and running face to face roleplaying
games, the same sort of issue comes up: Some GMs have the plot pretty
much predetermined, but I dislike this style. What I do when running a
game is create the background, the NPCs and so on, which I find leaves
plenty of room for creativity, but I *don't* predetermine the plot, I
let that happen as it will, from the interactions of the players and
NPCs. Of course, to at least some extent this is a matter of taste.


>We could go on and on -- it's all a matter of perspective. One reason I've
>come to view my bent as "better" is that it seems to me that the _fiction_
>part needs the most work. The player does have choices to make in current
>IF games. Most players do not seem to feel deprived when they discover
>that they couldn't really radically affect the outcome of the game.

>What is still missing are games that are good literature as well as being
>fun. I'm not talking about the basic craft of fashioning sentences from
>words, either -- many games around now are competently written. I'm
>talking about the greater goals of literature: to communicate in a way that
>is universal -- to deliver (a) message(s) in a way that transcends genre;
>to create characters that *live*, because the reader sympathizes.

>There are moments in existing IF. Many people feel like they "know" Floyd.
>But we have yet to see a work that maintains this kind of intensity
>throughout. Floyd is a brilliant stroke on an otherwise ordinary canvas.

I agree; indeed, I will agree with you that the lack of meaningful
literary content is more important than the lack of options for the
player (though I do think the latter is also an important lack).
Trinity is the only adventure game I can think of which I found
genuinely meaningful for pretty well most of the game.

>>If you, the author, have predetermined every plot point in advance, then
>>the player might as well just read a transcript of someone else solving the
>>game instead of going through the motions of playing it (other than the
>>intellectual challenge of solving each puzzle; but that's a form of
>>entertainment akin to a crossword rather than literature).

>You're reducing this case to absurdity. Although the overall plot of
>Colossal Cave is predetermined, the player never feels as though he's
>"trapped in a glorified novel". He feels like he's exploring a game world.
>When things happen, they certainly don't "feel" predetermined (at least not
>the first time through the game). I agree that such plotted games have
>less replay value, but only to the extent that books do too. (You may say,
>"Big deal, I know he's going to kill the king -- why bother?", but Hamlet
>is still compellling the 100th time around.)

A matter of opinion; I personally find that the linear nature of most
adventure games (i.e. being a series of puzzles each with a fixed
predetermined solution) is a serious lack.

>>Or read a good book, which as literature will be far better than any
>>adventure game will.

>It is truly maddening to me that most people here seem to be willing to
>accept this without a fight. Interactive fiction is clearly a superset of
>fiction -- you can just have the "player" read a novel on the screen.
>Given this, IF works have the potential to be every bit as good as any work
>of static fiction. (Yes, I realize it's an oversimplification, but you
>can't disagree with it in the abstract!)

Perhaps I expressed myself badly; I did not mean to imply that


interactive fiction can never be as good as a novel, just that all
current examples in fact are not as good.

>There are many ways you could try to bring "good literature" to IF. In


>Legend, I've tried one particular approach: I've put in quite lengthy
>sections of static prose in order to capitalize on the existing "literary
>tradition", so to speak. That's where I develop the characters (since
>"actors" are no more than cardboard cut-outs now), and, to a certain
>extent, where I develop the overall theme of the work. It's clumsier than
>I'd like, but as a first attempt at the method I think it works pretty
>well. Perhaps once a few r.a.i-f readers have played through it, they'll
>be able to explain where I'm coming from better than I (who seem to be
>failing miserably).

This sounds very good. This Legend game, is it currently available? If
so, how can I obtain a copy?

>>And the best way to do that is to allow things in the game to interact in
>>ways that the author may not have foreseen..

>This sounds good but is unworkable in practice. If the author hasn't
>anticipated it, the prose resulting from the action will be lame. Good
>text generation is years off. If there's one thing IF readers *have*
>voiced concern about, it's prose quality.

I would still rather be able to do something, even if the feedback I get


is in unimaginative prose, than be unable to do it at all by arbitrary
fiat.

>Also, there still *is* a big difference between plotted (i.e., traditional


>adventure game) IF and static fiction. When reading a book, you can learn
>about people and witness events that make an indelible impression on you.
>Plotted interactive fiction has the same potential, but you get to *meet*
>the people and *live* the events.

>This may sound overblown, but there really is truth to it. We recently got
>email from a player who was quite upset about having to solve a particular
>puzzle in the way it had to be solved. Without going into details -- I
>respect this player's views and privacy -- it made absolutely clear to me
>that even fairly straight-ahead IF games really can draw the player in, to
>the extent that he feels he is *responsible* for his fictional character's
>actions. This is a powerful medium indeed.

In this I agree, though I would say that its power is currently more
potential than actual.

Jamieson Norrish

unread,
May 29, 1994, 8:04:54 AM5/29/94
to
In article <2s3b8q$i...@harbinger.cc.monash.edu.au> benziesp@ponderosa
(Ms. Pauline Benzies) writes:

Putting aside the notion that diamond is one of the toughest
substances known to man and is unlikely to be "smashed" -- at
least, thats the sort of physics that I understood...

Diamonds are very hard, but I think (and may well be wrong) that they
can be cracked under such circumstances as were used in the example.

The point is that the author expected the player to have to find a
knife somewhere, not use the "sharp" property that the bottle
picked up because it was also "broken".

Why is strict adherence to what the author expected the player to do
necessary, or even beneficial? I would much rather have the experience
of "Now I can do what I want to do", rather than "Here I am, about to
embark on an intellectual game of jumping through pre-determined
hoops." I don't see the problem in this specific instance of allowing
a broken bottle to be used to cut something, rather than a knife. It
doesn't require very much more programming, and makes the game far
more interesting.

Remember, this was initially suggested to counter the argument that
"broken" == "not-usable".

And the coding to achieve this in this particular case may be tedious,
but is certainly not especially challenging.

Jamie

Andrew C. Plotkin

unread,
May 29, 1994, 12:06:27 PM5/29/94
to
bne...@delphi.com (Bob Newell) writes:

> Here I am replying to my own message because my mail program is so erratic.
> What got cut off was: Myst doesn't even have inventory and manipulation
> commands, it doesn't have menu dialog, etc. It creates and unfolds an
> environment in a very effective way, without relying on these "cliches".
> Objects tend to get manipulated in-place, not carried around and used in
> n(n-1)/2 attempts to make it work.
>
> If the paradigms can be so effectively broken in a graphics game, why not in
> a text game?

Because, IMHO, Myst managed this simply by throwing away many of the
things I most enjoy about interactive games. The story was involving,
but the interactivity was nearly nil; at most points in the game,
there was a thing in front of you and you did it. There was never any
question of which thing to apply to which other thing, which to me is
the basic expression of the player's free will. (Yes, the n(n-1)/2 stuff.)

Hmm, I bet that will require some explaining. Have I confused you?

Put another way: I claim (and I think you would agree) that if Myst
were rewritten as a text game, say in TADS, it would be an extremely
simple text game. "get", "drop", "put", "examine", "read", "turn", and
"pull" would cover the verb list, and the game-logic would be minute.
Does this mean that Myst has broken the paradigm of interactive
fiction? I don't think so, because nothing has been *added* except for
500 Meg of graphics and sound. My immediate reaction to the game was
"wow, this would be *great* if they had managed to combine this level
of visual artistry with the game technology that Infocom worked out
ten years ago. Get on the *stick*, people!" (:-)

What I want to try is exactly that. I think a CD-ROM has enough
storage to do a classic text game graphically -- that is, render
enough versions of each scene that it doesn't feel "stitched-together"
or restricted. Do what Myst did with rooms; three or four versions of
each, in different directions, plus closeups of interesting furniture
and variants for open / closed doors and so forth. Then ten or twenty
versions of each object, so that you can *realistically* scatter any
pile of objects around a room by superimposing images. (Twenty would
probably do it. Various distances, various directions of illumination,
a few variants for objects on or in particular other objects.)
This would take
a lot more space than Myst used, so you'd have to give up most of the
animation -- oh well. Put in an inventory window. Keep the simple
point-click-drag interface; that covers most of the verbs that I like
to throw into a text game, and I believe you can do a good piece of IF
without the strange one-situation verbs like "hold breath". You'd need
mouse conventions for things like "put A in B" (the basic Mac
drag-drop works) and "eat A" (maybe a mouth icon, or just have a
"double-click to use" rule.) Talking with other game characters is
harder. You might have to skip that. The game theme would probably
pull in a few more bits of interface; a list-of-spells window, or
whatever.

Conclusion: such a game form is more limited than the classic text
game; but I don't think the limitations need affect the quality of the
fiction. The graphics would be more limited than Myst, since there
would have to be many more images and thus fewer and shorter
animations, and the animations did play a large part in creating the
environment of Myst. But for such a game, I think I would be willing
to deal.

So would such a thing combine the strengths of text games and current
CDROM games? Or combine the weaknesses? Give me six Quadras and two
years to work, and I'll try to find out. Sigh.

--Z

"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."

Bob Newell

unread,
May 28, 1994, 8:42:09 PM5/28/94
to
>I'd like to mention as an example a non-text game which I just began
>exploring, "Myst". Graphics and music are used to create a startling sense
>of place and atmosphere. But what is so fascinating is how Myst breaks
many
>of the old paradigms of a game of its type:

Here I am replying to my own message because my mail program is so erratic.

Jamieson Norrish

unread,
May 30, 1994, 8:22:37 AM5/30/94
to
In article <2s3bnd$i...@harbinger.cc.monash.edu.au> benziesp@ponderosa
(Ms. Pauline Benzies) writes:

>It is different in that it allows for "mistakes", and it also avoids
>railroading the player along a certain path - which for its own sakes
>is a good thing. It is not removing puzzles, it is simply requiring
>that the player find a different set, by removing one lot of problems.

Actually, the last time I sprayed the office with a flame-thrower,
my boss fired me. I told him it was a mistake and that I should
still be able to finish off the rest of my employment. Curiously,
he wasn't interested.

Umm, excuse me, but we've moved away from that specific example. If
you throw away a newspaper and it gets shredded by the garbage
disposal people, that is a "mistake" which might be allowed for by
having the same information present elsewhere, or by the inclusion of
other possible storylines in the game.

Why should the simulation let the player make destructive mistakes?
In general life doesnt! What are we simulating anyway? Remember,
in life, when the old man staggers up to you and hands you a
document proving sinister machinations in the political party of
your choice that will blow the lid off the countries government,
you don't get the option of burning it and then looking around for
another way to do the same thing!

Well, maybe you live in a different world from me, but "general life"
certainly does let me and everyone else make destructive mistakes. I
can burn a vital document just as easily as I can burn garbage - life
doesn't change the laws of physics just because the former is
important. The *effects* of my action might not be easily overcome,
however.

As for your example - sure, the information may be gone forever, and
that story is thus closed to you. Sounds like a fairly weak way of
getting the character involved, with only one slim hook, but that
doesn't stop there being other story-lines to follow.

Besides, you seem to yourself be advocating not allowing the player to
do things which will make the game impossible to solve. Why so?

> Lets see, how do I code the "broken" bottle. Obviously, if they
> used "BREAK BOTTLE CAREFULLY" there will be a clean edge that can
> be used to cut. On the other hand, "BREAK BOTTLE WITH AXE" tends to
> leave long splinters.
>The coding for something along these lines would be hideous, of
>course. But that is not to say that it is impossible, or worthless.

I agree. However, it pushes this thread back into the dreaded
"<verb> <noun> <adverb>" discussion and thats the point that most
people get off. Myself included.

Your example did bring up the adverb question, as I'm sure it was
intended to. However, please don't pretend that all elements of
simulation rely on adverbs. That would be patently absurd.

Jamie

Bob Newell

unread,
May 29, 1994, 9:52:17 PM5/29/94
to
>Put another way: I claim (and I think you would agree) that if Myst
>were rewritten as a text game, say in TADS, it would be an extremely
>simple text game. "get", "drop", "put", "examine", "read", "turn", and
>"pull" would cover the verb list, and the game-logic would be minute.
>Does this mean that Myst has broken the paradigm of interactive
>fiction? I don't think so, because nothing has been *added* except for
>500 Meg of graphics and sound. My immediate reaction to the game was

If Myst (or similar) were a TADS game, it would indeed be very simple at the
parser/response level. But to complete the analogy, suppose it had 500 meg
of rich, vibrant prose? (OK, maybe not quite 500 meg, I'd settle for 2).

But perhaps I am getting ahead of your intruiging idea:



>What I want to try is exactly that. I think a CD-ROM has enough
>storage to do a classic text game graphically -- that is, render
>enough versions of each scene that it doesn't feel "stitched-together"
>or restricted. Do what Myst did with rooms; three or four versions of
>each, in different directions, plus closeups of interesting furniture
>and variants for open / closed doors and so forth. Then ten or twenty

(detail deleted)



>Conclusion: such a game form is more limited than the classic text
>game; but I don't think the limitations need affect the quality of the
>fiction. The graphics would be more limited than Myst, since there

>So would such a thing combine the strengths of text games and current
>CDROM games? Or combine the weaknesses? Give me six Quadras and two

The productions to date haven't been so great (I'm thinking of most of the
text and static graphics games out there, including some on CDROM). But
your approach could be very interesting; it amounts to creating a rich
static environment for a hybrid type of game.

Still, I'd like