Thoughts on new interfaces for I-F

25 views
Skip to first unread message

George E Caswell

unread,
May 13, 1996, 3:00:00 AM5/13/96
to

Hi there.

During an idle brainstorming session (the only way I get anything
conceptually done with my Inform game. I get no real work done. :( ) I
was thinking about how certain interface and world elements of I-F haven't
changed, really, at all since the days of Colossal Cave, including
graphical adventures, Infocoms, TADSes, Informs, etc... All based around
Objects, rooms, and a third person command situation... (well, there are
a FEW exceptions to every rule, but, mostly...) These, well, you can't
really call them limitations, but these self-imposed restrictions don't
seem to fit everything... Examples like the white house in Zork or the
large lab in Seastalker don't fit the room-and-exit idea too well... so
was kinda wondering if anyone had tried to 'push the envelope' of
technique, so to speak?
One idea I've been tinkering with is to keep the room idea and possibly
the exit idea, but have actual player movement take place in scaled
real-space. The room would have quantification of its size and shape,
every object in the room would have a location... I understand an
adventure game, or any simulation for that matter, tends to only become
more muddled in details and require more and more implementation when
attempting a higher level of realism (example: New Zork Times mentioned
the boat in Zork, which required modifying the rules for water, and the
axe, which required consideration fo the existance of trees..) I also
think such a world might be hard to navigate in plaintext... but...

You're in a large rectangular room. It appears to be about 10 meters by
20, give or take. You're about 3 meters from the middle of the north
wall, facing southwest... There's a marble column one meter in front of
you, standing 8 meters tall with a diameter of 2 meters.

>Turn half-right. Walk 5 meters.
[Ahead.]

You're about 6 meters from the middle of the north wall.
You see a sign on the south face of the column. You can't read it from
here.

>Turn around. Turn half right. Walk 5 meters.
[Ahead.]

You're approximately in the center of the large room. You can now clearly
read the sign on the column.

>Read it.

It reads, "This space intentionally left blank."

-

Of course, the example is a bit muddled and arbitrary- in the past it's
always been assumed your character wouldn't be so brain-dead as to be
unable to navigate a room himself... it's an idea in progress, and I'm
not sure just how ill-conceived it may be, especially to attempt it in a
text-only situation. I think ideally such a thing would have an overhead
2-d map, or at least well-detailed paper maps...

I guess it's a question of style-- in the opinions of the authors and
players 'present', is the room-and-exit approach something that should be
abandoned, or at least reconsidered? How does quantification rank against
natural, if limited, prose? Especially, how do you quantify things and
create details and less limited interaction without risking the entire
game becoming dull drudgery (like Detective, or parts of Seastalker, most
Infocom mazes, the swamp in Return to Zork, etc...)

Also, any thoughts anyone has on the difficulty of implementation of
the system I created the fictional transcript for--? As far as I can
tell, if I don't want to get too fancy keeping track of what the player
does and does not know the existance of, can and can not see, etc, it just
takes a grid-description of each room, a re-working of the exit system to
put the player at a specific point in the destination room, and new verbs
for moving, and new rules to handle movement (IE movement moves the player
within the room's grid, until he reaches a boundary, then he crosses over,
a-la DOOM-esque sectoring...) What can I say? I'm a hacker at heart, I
program. <g>

This intricate, convoluted question brought to you by-

T I M B U K T U
________________ _/>_ _______
<___ ___________// __/<___ /
// <> _____ <_ > / ____/
// /> / / __/ / / <___________
// </ </</</ <_ _/ <_____________/
</ </ ---The same folks who brought you
this nice sig...


PAZ SALGADO

unread,
May 13, 1996, 3:00:00 AM5/13/96
to

George E Caswell (timb...@wpi.edu) wrote:
: You're in a large rectangular room. It appears to be about 10 meters by

: 20, give or take. You're about 3 meters from the middle of the north
: wall, facing southwest... There's a marble column one meter in front of
: you, standing 8 meters tall with a diameter of 2 meters.

: >Turn half-right. Walk 5 meters.
: [Ahead.]

: You're about 6 meters from the middle of the north wall.
: You see a sign on the south face of the column. You can't read it from
: here.

: >Turn around. Turn half right. Walk 5 meters.
: [Ahead.]

Too much complex. This kind of details are very good for 3-D adventures, like
DOOM or its older versions. I think it's a very good idea try to mix the
posibilities of this kind of adventures with I-Fs; making better things than
the normal role-playing by computer (like EOB or ISTHAR). But the computer need
to calculate a lot of graphics for a lot of situations, and the programmer
need to draw the graphics for every puzzle. Just a bit impossible nowadays.

I couldn't understand anything if I must to walk through a *real* laberinth
step by step just only asking for every step (asking in english!).

Well, I think it is no a good idea.If you want this kind of details you may
do graphical adventures or DOOM's ones.


Meliton Rodriguez, from 115 room

Gareth Rees

unread,
May 13, 1996, 3:00:00 AM5/13/96
to

George E Caswell <timb...@wpi.edu> wrote:
> [...] keep the room idea and possibly the exit idea, but have actual

> player movement take place in scaled real-space. The room would have
> quantification of its size and shape, every object in the room would
> have a location. [...]

>
> You're in a large rectangular room. It appears to be about 10 meters by
> 20, give or take. You're about 3 meters from the middle of the north
> wall, facing southwest... There's a marble column one meter in front of
> you, standing 8 meters tall with a diameter of 2 meters.

Obviously you wouldn't print text like this, because only very rarely
would the dimension be important. Instead, the implementation would do
any necessary interpretation of the player's commands. For example
(internal machinations in square brackets):

A room. There is a window in the north wall and a table in the middle
of the room.

> open window
[ player approaches window ]
[ player tries to touch window, but fails because the player is
1.6m tall and the window is 2.3m off the ground ]
You can't reach the window, it's too high up.

> push the table over to the north wall
[ player approaches table ]
[ player tries to push the table; success ]
[ check for obstacles between the table and the north wall; none ]
OK.

> stand on table
OK.

> open window
[ player approaches window ]
[ player tries to touch window: success ]
[ player tries to open window, but window is locked ]
The window is locked shut.

So it looks to the player like an ordinary adventure game, but
internally objects have sizes and shapes and locations within rooms.

I'd be interested in seeing such a game engine (I've thought about this
issue in the past), but I imagine that the gains wouldn't be worth the
implmentation difficulties (although making the implementation would be
a very interesting research problem, involving AI planning techniques).
It's very rare that I write a puzzle where shapes, sizes and positions
are important, and where they are necessary it is usually easy to code
them up in some hackish fashion. For example, the above puzzle with the
table and window is the work of half a dozen lines of code in Inform or
TADS.

--
Gareth Rees

John Elliott

unread,
May 13, 1996, 3:00:00 AM5/13/96
to

In article <Pine.ULT.3.93.960513...@wpi.WPI.EDU>, George E Caswell (timb...@wpi.edu) wrote:

: You're in a large rectangular room. It appears to be about 10 meters by


: 20, give or take. You're about 3 meters from the middle of the north
: wall, facing southwest... There's a marble column one meter in front of
: you, standing 8 meters tall with a diameter of 2 meters.

: >Turn half-right. Walk 5 meters.
: [Ahead.]

: You're about 6 meters from the middle of the north wall.
: You see a sign on the south face of the column. You can't read it from
: here.

: >Turn around. Turn half right. Walk 5 meters.
: [Ahead.]

: You're approximately in the center of the large room. You can now clearly
: read the sign on the column.
:
: >Read it.

: It reads, "This space intentionally left blank."

I suspect that trying to move my player about like that without some
sort of graphical view would pretty quickly drive me round the bend.

: it's an idea in progress, and I'm


: not sure just how ill-conceived it may be, especially to attempt it in a
: text-only situation. I think ideally such a thing would have an overhead

: 2-d map...

Sounds better. But this is all reminding me a bit of my schooldays:

______
/ \
/ \
\ /
\_|\___/
|/

? REPEAT 6 [FD 50 RT 30]

: Also, any thoughts anyone has on the difficulty of implementation of


: the system I created the fictional transcript for--?

I'd say the answer's pretty obvious by now... use LOGO :-)

-------------------- http://users.ox.ac.uk/~sjoh0132/ ---------------------
John Elliott |BLOODNOK: "But why have you got such a long face?"
|SEAGOON: "Heavy dentures, Sir!" - The Goon Show
:-------------------------------------------------------------------------)

David Baggett

unread,
May 13, 1996, 3:00:00 AM5/13/96
to

In article <yxsohns...@ouse.cl.cam.ac.uk>,
Gareth Rees <gd...@cl.cam.ac.uk> wrote:

>It's very rare that I write a puzzle where shapes, sizes and positions
>are important, and where they are necessary it is usually easy to code
>them up in some hackish fashion.

More importantly, we often write puzzles that fail to conform to real-world
physics. You'd have to go to extra trouble to coerce a library enforcing
physical constraints to this degree to allow "put anchor in magical bottle
cap" (or what have you).

Dave Baggett
__
d...@ai.mit.edu
"Mr. Price: Please don't try to make things nice! The wrong notes are *right*."
--- Charles Ives (note to copyist on the autograph score of The Fourth of July)

Den of Iniquity

unread,
May 14, 1996, 3:00:00 AM5/14/96
to

On Mon, 13 May 1996, George E Caswell wrote:
>snip<

> You're in a large rectangular room. It appears to be about 10 meters by
> 20, give or take. You're about 3 meters from the middle of the north
> wall, facing southwest... There's a marble column one meter in front of
> you, standing 8 meters tall with a diameter of 2 meters.
>
> >Turn half-right. Walk 5 meters.
> [Ahead.]
>
> You're about 6 meters from the middle of the north wall.
> You see a sign on the south face of the column. You can't read it from
> here.
>
> >Turn around. Turn half right. Walk 5 meters.
> [Ahead.]
>
> You're approximately in the center of the large room. You can now clearly
> read the sign on the column.
>
> >Read it.
>
> It reads, "This space intentionally left blank."

Ooh, no, I don't like this...

You're in the kitchen, facing south west, 3 metres north-north-east of the
door.

-> turn quarter left. walk 1 metre

You go forward 50cm then bump into the table. Ouch.

-> turn half right. walk 50cm

OK (...descriptive message...)

-> turn half left. Walk 2 metres

You go forward 32.5cm then bump into the table. Ouch. The vase rocks
slightly.

-> move

Bump!

-> move. move. move. move

Bump! (crash, tinkle!)

You need a great deal of textual information to navigate accurately, and
this would stop being playable after about three or four frustrating
mistakes in the first dozen or so moves. This sort of thing has to be
done visually really.

But as for 2D overhead maps - why keep them in the mind of the Z-machine
when you can stick them on the screen? Does anybody with plenty of
Z-machine skill out there fancy writing a special maze library which would
allow an author to create a 'maze of twisty passages all alike' which is
viewed a la Moria? Maybe the player should only have a limited view of the
maze - say a 7x7 block or something... I remember a couple of old Atari
VCS games did that sort of thing - the one I remember best was called
(interestingly enough) 'Adventure' and had a few mazes in which you only
had a certain radius of vision to go by. There was another called
something like 'Haunted House' which had a similar effect. If you enter a
room (or find an object) within the maze (perhaps by deliberately moving
the cursor on to a designated square [a non-space, non-wall character]) ,
the game could return to a normal description. Ideally the player should
have the option to play without this little 'subgame'.

I'm not saying that this idea would be the best thing since magical
burins but it would be different and if implemented well could make a
pleasant change.

Other new interfaces, people?

Den

Matt Ackeret

unread,
May 17, 1996, 3:00:00 AM5/17/96
to

In article <4n7afs$s...@tid.tid.es>, PAZ SALGADO <ja...@tid.tid.es> wrote:
>Too much complex. This kind of details are very good for 3-D adventures, like
>DOOM or its older versions. I think it's a very good idea try to mix the
>posibilities of this kind of adventures with I-Fs; making better things than
>the normal role-playing by computer (like EOB or ISTHAR). But the computer need
>to calculate a lot of graphics for a lot of situations, and the programmer
>need to draw the graphics for every puzzle. Just a bit impossible nowadays.

If you think of Wizardry I-type wireframe graphics, something like this could
be done in text, and if it were well done could be engaging.
--
unk...@apple.com Apple II Forever
These opinions are mine, not Apple's.

Steven Howard

unread,
May 19, 1996, 3:00:00 AM5/19/96
to

In <Pine.ULT.3.93.960513...@wpi.WPI.EDU>, George E Caswell <timb...@wpi.edu> writes:
>
>Of course, the example is a bit muddled and arbitrary- in the past it's
>always been assumed your character wouldn't be so brain-dead as to be
>unable to navigate a room himself... it's an idea in progress, and I'm
>not sure just how ill-conceived it may be, especially to attempt it in a
>text-only situation. I think ideally such a thing would have an overhead
>2-d map, or at least well-detailed paper maps...
>
> I guess it's a question of style-- in the opinions of the authors and
>players 'present', is the room-and-exit approach something that should be
>abandoned, or at least reconsidered? How does quantification rank against
>natural, if limited, prose? Especially, how do you quantify things and
>create details and less limited interaction without risking the entire
>game becoming dull drudgery (like Detective, or parts of Seastalker, most
>Infocom mazes, the swamp in Return to Zork, etc...)
>

Now that you mention it, I'd like *less* of this kind of detail. I'd much rather type
"drive to my office" than "stand up. take keys. open front door. n. close front door.
lock door with keys. e. e. d. d. e. unlock car with keys. get in car. turn car on.
push button. n. push button. e. e. n. n. w. up. up. w. turn car off. exit. lock car
with keys. n. up. n. unlock office door with keys. open it. w. close it. sit on chair."


========
Steven Howard
bl...@ibm.net

What's a nice word for "euphemism"?

George E Caswell

unread,
May 20, 1996, 3:00:00 AM5/20/96
to

On 19 May 1996, Steven Howard wrote:

> >game becoming dull drudgery (like Detective, or parts of Seastalker, most
> >Infocom mazes, the swamp in Return to Zork, etc...)
>

> Now that you mention it, I'd like *less* of this kind of detail. I'd much rather type
> "drive to my office" than "stand up. take keys. open front door. n. close front door.
> lock door with keys. e. e. d. d. e. unlock car with keys. get in car. turn car on.
> push button. n. push button. e. e. n. n. w. up. up. w. turn car off. exit. lock car
> with keys. n. up. n. unlock office door with keys. open it. w. close it. sit on chair."
>

Kind of my thought, halfway through the post, too-- but the idea behind
it is that some sort of quantitative representation of the world, instead
of vague room/object.... So I guess what's needed is either 1: a game
interface smart enough to cut through the crap when you want it to, or 2:
a good example of a game where quantification would be very helpful...

Matthew Daly

unread,
May 20, 1996, 3:00:00 AM5/20/96
to

In article <4nnl7o$1m...@news-s01.ny.us.ibm.net> bl...@ibm.net writes:
>
>Now that you mention it, I'd like *less* of this kind of detail.
>I'd much rather type
>"drive to my office" than "stand up. take keys. open
>front door. n. close front door.
>lock door with keys. e. e. d. d. e. unlock
>car with keys. get in car. turn car on.
>push button. n. push button. e. e. n. n. w.
>up. up. w. turn car off. exit. lock car
>with keys. n. up. n. unlock office door
>with keys. open it. w. close it. sit on chair."

A macro command should be standard equipment for an IF parser, so you
could define new verbs as a combination of old words.

I also think that, if your character knows where a location is you
should be able to say "go to office parking lot entrance" and have
it issue a collection of direction statements, like it does in
Suspect, for instance.

BTW, you forgot to put your car into a parking spot before you
turned it off, so the garage is getting backed up. You might
want to clear it up before your boss shows up.... ;-)

-Matthew Daly

Cardinal Teulbachs

unread,
May 23, 1996, 3:00:00 AM5/23/96
to

George E Caswell <timb...@wpi.edu> wrote:
>You're in a large rectangular room. It appears to be about 10 meters by
>20, give or take. You're about 3 meters from the middle of the north
>wall, facing southwest... There's a marble column one meter in front of
>you, standing 8 meters tall with a diameter of 2 meters.

>>Turn half-right. Walk 5 meters.
>[Ahead.]

>You're about 6 meters from the middle of the north wall.
>You see a sign on the south face of the column. You can't read it from
>here.

>>Turn around. Turn half right. Walk 5 meters.
>[Ahead.]

>You're approximately in the center of the large room. You can now clearly
>read the sign on the column.
>
>>Read it.

>It reads, "This space intentionally left blank."

I guess I'm a bit late getting in on this one, but here goes anyway:

If one of the intents of i-f is to be in any way literary, then this
approach would not work at all. Imagine reading a book in which every
character's actions were plotted out in the minutest detail:

"Joe walked 3.2 meters east and, after turning 26 degrees, continued
on in a roughly northeasterly direction for another 7.6 meters."

Who wants to read stuff like that? What we're mostly interested in in
literature are those things which we conceive as having some
commensurate and direct bearing upon the outcome, just as in a history
the historian lists only those things he takes to be causal and
ignores the rest. For instance, walking, flying and swimming are all
modes of movement, which conceptions have in themselves no reference
to questions like "how far?" or "how high?" If the key thing, then, is
that Joe got to the treasure location--unless the precise distance
travelled has some bearing upon the story, as in the case of, say, a
treasure map which has to be followed--we're perfectly satisfied to
know simply that he walked to it. It's clear enough that he walked
*some* distance, and that's enough for us to identify the general
cause of his getting there.

The same thing should apply to interactive fiction, if the word
"fiction" is to retain any connection with traditional notions of
storytelling. While it's perfectly possible to do the sort of thing
you suggest, the question is *whether* one should do it, and in this
case the answer is no. Measurement and geometry are concerned with
precise determinations about quantity, whereas we don't live our lives
as calculating machines (and even if we did, the data necessary to
carry out such calculations is rarely available). We tend to make
rough judgments about things based upon their appearance to our
senses--which is why we have words like "seems" and "appears." This is
the language that literature is concerned with, the language of life
as it is directly experienced, and not that of the formal propositions
of mathematical science or physics.

--Cardinal T

I mean, what the hell kind of villain thwarts the hero's
progress with soup cans in the kitchen pantry?
--Russ Bryan

Cardinal, I follow up your post in the hopes that some
day I too will be quoted in your sig.
--Matthew Amster-Burton

Hey! This isn't what I said! What'd you do with my
quote?
--Bonni Mierzejewska


John Wood

unread,
May 24, 1996, 3:00:00 AM5/24/96
to

card...@earthlink.net (Cardinal Teulbachs) writes:
>
> If one of the intents of i-f is to be in any way literary, then this
> approach would not work at all. Imagine reading a book in which every
> character's actions were plotted out in the minutest detail:
>
> "Joe walked 3.2 meters east and, after turning 26 degrees, continued
> on in a roughly northeasterly direction for another 7.6 meters."
>
> Who wants to read stuff like that? What we're mostly interested in in
> literature are those things which we conceive as having some
> commensurate and direct bearing upon the outcome
[snip]

> treasure map which has to be followed--we're perfectly satisfied to
> know simply that he walked to it. It's clear enough that he walked
> *some* distance, and that's enough for us to identify the general
> cause of his getting there.
>
> The same thing should apply to interactive fiction, if the word
> "fiction" is to retain any connection with traditional notions of
> storytelling. While it's perfectly possible to do the sort of thing
> you suggest, the question is *whether* one should do it, and in this
> case the answer is no.
[snip]

Perfectly true, and I happen to agree with you in this case...but the
same arguments could be used against *most* of the actions players
type in an adventure game. Do we really want to read about
characters who keep checking to see what they are carrying, looking
around, in, under and behind things? "Mr. Darcy unlocked the front
door with the brass key, opened it, went in, and looked around..."
doesn't sound *too* bad - once. But when 50% of the whole work
consists of similar passages the book will be remaindered faster
than you can say XYZZY.

To get back to movement: I think that (contrary to popular programming
books) a greater use of "go to" would be beneficial. As someone else
pointed out, it's not particularly literary *or* convenient to say
"n. e. e. open door. s. close door. d." when "go to garage" would
suffice.

[I'm off for 10 days holiday tomorrow, so don't expect an immediate
response if you comment on this]

John


George E Caswell

unread,
May 26, 1996, 3:00:00 AM5/26/96
to

On Thu, 23 May 1996, Cardinal Teulbachs wrote:

> George E Caswell <timb...@wpi.edu> wrote:

> >You're in a large rectangular room. It appears to be about 10 meters by
> >20, give or take. You're about 3 meters from the middle of the north
> >wall, facing southwest... There's a marble column one meter in front of
> >you, standing 8 meters tall with a diameter of 2 meters.
>
> >>Turn half-right. Walk 5 meters.
> >[Ahead.]
>
> >You're about 6 meters from the middle of the north wall.
> >You see a sign on the south face of the column. You can't read it from
> >here.
>
> >>Turn around. Turn half right. Walk 5 meters.
> >[Ahead.]
>
> >You're approximately in the center of the large room. You can now clearly
> >read the sign on the column.
> >
> >>Read it.
>
> >It reads, "This space intentionally left blank."
>

> I guess I'm a bit late getting in on this one, but here goes anyway:
>

> If one of the intents of i-f is to be in any way literary, then this
> approach would not work at all. Imagine reading a book in which every
> character's actions were plotted out in the minutest detail:
>

It's possible my example was bad-- my intention isn't to make players and
the game work in piddling little detains, it's to include the ability to
work in details when necessary or useful...

> "Joe walked 3.2 meters east and, after turning 26 degrees, continued
> on in a roughly northeasterly direction for another 7.6 meters."
>

Or, alternately, "Joe walked east, then north, then east again. After
travelling a good distance to the northeast, he took his inventory,
dropped a piece of paper to mark his way, and went off to the south."

> Who wants to read stuff like that? What we're mostly interested in in
> literature are those things which we conceive as having some

> commensurate and direct bearing upon the outcome, just as in a history
> the historian lists only those things he takes to be causal and
> ignores the rest. For instance, walking, flying and swimming are all
> modes of movement, which conceptions have in themselves no reference
> to questions like "how far?" or "how high?" If the key thing, then, is
> that Joe got to the treasure location--unless the precise distance
> travelled has some bearing upon the story, as in the case of, say, a

> treasure map which has to be followed--we're perfectly satisfied to
> know simply that he walked to it. It's clear enough that he walked
> *some* distance, and that's enough for us to identify the general
> cause of his getting there.
>
> The same thing should apply to interactive fiction, if the word
> "fiction" is to retain any connection with traditional notions of
> storytelling. While it's perfectly possible to do the sort of thing
> you suggest, the question is *whether* one should do it, and in this

> case the answer is no. Measurement and geometry are concerned with
> precise determinations about quantity, whereas we don't live our lives
> as calculating machines (and even if we did, the data necessary to
> carry out such calculations is rarely available). We tend to make

...Again, the example is slightly bad. The commands included in the
example are intended fo very limited use, and would really require a
better interface.

> rough judgments about things based upon their appearance to our
> senses--which is why we have words like "seems" and "appears." This is
> the language that literature is concerned with, the language of life
> as it is directly experienced, and not that of the formal propositions
> of mathematical science or physics.

T I M B U K T U

Reply all
Reply to author
Forward
0 new messages