Games in development

20 views
Skip to first unread message

Crispin Boylan

unread,
Dec 8, 1998, 3:00:00 AM12/8/98
to
Hi

Is anyone out there planning, or about to release a game written with ALAN,
HUGO, JACL (what has happened to JACL?) or any other authoring system other
than Inform or TADS.

It would be great if you could, its for an article i'm doing for XYZZYnews.

Cheers.
Cris.

Eric O'Dell

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
On Tue, 08 Dec 1998 22:33:05 GMT, viewt...@easynet.co.uk (Crispin
Boylan) wrote:

>Is anyone out there planning, or about to release a game written with ALAN,
>HUGO, JACL (what has happened to JACL?) or any other authoring system other
>than Inform or TADS.

I'm currently planning a game called Nocoroco which will be based on
the CAVE engine, which is itself in a pre-development (i.e., design
and specification) stage. Nocoroco will combine elements of IF and
RPGs and provide a playing environment that will (hopefully) be a
dynamic world in which the story unfolds rather than a mere setting
for a story. The overall genre is intended to resemble the fantastic
realism of some Latin American literature, a la Borges and Marquez.

I'm planning to make a more detailed announcement about the features
that will be included in CAVE as soon as the specification is complete
and actual coding begins. Here is a preview of some features whose
implementation details have been mostly worked out:

* CAVE will be based on C or C++ (probably C) rather than a
special-purpose language. CAVE adventures will be executable code.

* While designed to permit the use of substantially flashier
interfaces, the default distribution will probably use Glk to make
porting traditional text-only games easier.

* The parser will understand adverbs (finally) and many kinds of
interrogatives in addition to the usual imperative commands.

* The parser (and the underlying object model) will be able to handle
collections of objects, such as "the bridge club" instead of "Betty,
Sue, Maude, and Larry", which will be able to act individually as well
as collectively. Collections will also be useful for handling money
realistically, so that "some cash" or "$5.40" can also be dealt with
as "a five dollar bill, a quarter, a nickel, and a dime". It will also
understand and be able to handle indefinitely subdivisible objects
such as water and sand, providing the necessary support for handling
phrases like "half of the sand", "six liters of water", etc.

* CAVE encodes proximity relationships with objects so that
interactions across the boundaries of "rooms" can be implemented
easily and realistically. Fr'example, it will be possible to implement
a single instance of a tower that is visible from many locations
throughout the game world, and whose appearance varies according to
distance and perspective.

* A sophisticated message-passing scheme allows events with
far-reaching effects to be implemented easily. An explosion in the
basement could automatically damage nearby objects, with the extent of
the damage diminishing with distance. All that is necessary is for
objects to have methods to handle explosions; the system figures out
the rest from proximity relationships. A similar scheme is used for
knowledge acquisition by NPCs.

* Greater support for converations with NPCs. Support for directed
dialogue, too, if you like that sort of thing.

* Excluding for the moment these (and some other less-well defined)
new features, CAVE will support much the same basic featureset as
Inform and TADS. Many features of WorldClass will be cloned as well.

CAVE will be released under a Berkeley-style license and should appear
in beta form in 12-24 months. Maybe sooner if I win the lottery and
can devote all my time to it. Aside from providing the features I need
to implement Nocoroco, a major design goal is to provide some
genuinely new capabilities and inspire other developers to push beyond
the Infocom-esque rut that, IMHO, too much current IF is caught in.


--Eric


+-------------------------------------------------------------------+
| "I have come a very long way from myself only to realize that |
| identity is a skill and self-betrayal is a habit. Once lost, the |
| former is very hard to regain; once gained, the latter is very |
| hard to lose." ---I. Corvus, _The Europe of Our Dreams_ |
+-------------------------------------------------------------------+

Palmer Davis

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to Crispin Boylan
Crispin Boylan wrote:
>
> Is anyone out there planning, or about to release a game written
> with ALAN, HUGO, JACL (what has happened to JACL?) or any other
> authoring system other than Inform or TADS.

Yes; I'm working on a five-part science fiction series in a
Scheme dialect for the Z-machine and JVM. It's sort of like
_King_Lear_, the _Ring_ cycle, and the history of the early
Hellenistic age thrown into a blender. It looks like Act I
might be playable in two hours, in which case it might turn
up as a competition entry one of these years.

Hopefully, this will turn out to be an _Avalon_ in the new sense
of "very ambitious work of IF that takes an unreasonably long
time to complete" rather than in the old sense of "highly touted
IF vaporware that never appears." Of course, the fact that I
just mentioned it in public has probably added six months or so
to its final release date....

--
Palmer Davis <pal...@secant.com>
Secant Technologies * 4853 Galaxy Parkway * Cleveland OH 44128

Martin Dransfield

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
Palmer Davis <pal...@secant.com> writes:

> Yes; I'm working on a five-part science fiction series in a
> Scheme dialect for the Z-machine and JVM.

Is this a standard JVM implementation of Scheme like Kawa or SILK? Or is
it specially adapted for IF? If the latter is it published anywhere?

I'm interested because I've been thinking of writing a translator from
a lispish source format (yet to be decided) to Inform for
compilation. Even when trying very hard I cannot bring myself to like
Inform syntax. A personal failing I'm sure :-); whereas I do like
Lisp's (lack of) syntax. A usable authoring language with a
Lisp-like syntax might be the spur I need to attempt a writing a
game. I do not, however, want to write yet another game delivery system.

> It's sort of like _King_Lear_, the _Ring_ cycle, and the history of
> the early Hellenistic age thrown into a blender. It looks like Act
> I might be playable in two hours, in which case it might turn up as
> a competition entry one of these years.

This sounds fascinating. Good luck

-- Martin

dou...@hotmail.com

unread,
Dec 10, 1998, 3:00:00 AM12/10/98
to
On Fri, 11 Dec 1998 00:12:27 +0100, elindh....@SPAMebox.tninet.se
(Peter Nilsson) wrote:

>Crispin Boylan <viewt...@easynet.co.uk> wrote:
>
>> Hi


>>
>> Is anyone out there planning, or about to release a game written with ALAN,
>> HUGO, JACL (what has happened to JACL?) or any other authoring system other
>> than Inform or TADS.
>>

>> It would be great if you could, its for an article i'm doing for XYZZYnews.
>>
>> Cheers.
>> Cris.
>
>

Hey, just wanted to let you know that I'm working on a game called
'Ravenshire Keep' using ALAN. I'm not really sure when I'll be able
to release it since I'm still trying to figure out the more advanced
aspects of its programming but I'll let you know when I finish.

Doug

Stephen Griffiths

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
Martin Dransfield wrote:
> [....]

> I'm interested because I've been thinking of writing a translator from
> a lispish source format (yet to be decided) to Inform for
> compilation. Even when trying very hard I cannot bring myself to like
> Inform syntax. A personal failing I'm sure :-); whereas I do like
> Lisp's (lack of) syntax. A usable authoring language with a
> Lisp-like syntax might be the spur I need to attempt a writing a
> game. I do not, however, want to write yet another game delivery system.

Have you had a look at the GINAS system which you can download from the
IF Archive?

The IF-Archive index description is:
GINAS: The Generic Interactive Narrative Authoring System, by Jeff
Standish. GINAS is an object-oriented LISP system for
Interactive Fiction. Current version: 0.4 (August 1995), beta

ftp://ftp.gmd.de/if-archive/programming/ginas/

It may not be what you're thinking about but there's probably some
relevant ideas there.

Regards,
SteveG

Peter Nilsson

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
Crispin Boylan <viewt...@easynet.co.uk> wrote:

> Hi
>
> Is anyone out there planning, or about to release a game written with ALAN,
> HUGO, JACL (what has happened to JACL?) or any other authoring system other
> than Inform or TADS.
>
> It would be great if you could, its for an article i'm doing for XYZZYnews.
>
> Cheers.
> Cris.


--
*** remove 'SPAM' from email address ***

*** ta bort 'SPAM' från emailadressen ***

Peter Nilsson

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
Crispin Boylan <viewt...@easynet.co.uk> wrote:

> Hi
>
> Is anyone out there planning, or about to release a game written with ALAN,
> HUGO, JACL (what has happened to JACL?) or any other authoring system other
> than Inform or TADS.
>
> It would be great if you could, its for an article i'm doing for XYZZYnews.
>
> Cheers.
> Cris.

(sorry about the other post. all these stupid machines and stuff :-))

I'm not quite sure what you mean, but I'm always planning to release a
games written with an 'alternative' system, namely ALAN.
Maybe someday...

What do you want to know?


Peter Nilsson

Martin Dransfield

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
Stephen Griffiths <stephen....@xtra.co.nz> writes:

> Have you had a look at the GINAS system which you can download from the
> IF Archive?
>

> [...]


>
> It may not be what you're thinking about but there's probably some
> relevant ideas there.

Thanks, yes I have looked at GINAS. It has (and will) inform any
design I finally come up with.

-- Martin

Magnus Olsson

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
In article <367038...@xtra.co.nz>,

Stephen Griffiths <stephen....@xtra.co.nz> wrote:
>Martin Dransfield wrote:
>> [....]
>> I'm interested because I've been thinking of writing a translator from
>> a lispish source format (yet to be decided) to Inform for
>> compilation. Even when trying very hard I cannot bring myself to like
>> Inform syntax. A personal failing I'm sure :-); whereas I do like
>> Lisp's (lack of) syntax. A usable authoring language with a
>> Lisp-like syntax might be the spur I need to attempt a writing a
>> game. I do not, however, want to write yet another game delivery system.
>
>Have you had a look at the GINAS system which you can download from the
>IF Archive?

(...)

>It may not be what you're thinking about but there's probably some
>relevant ideas there.

You may also get some ideas from AdvSys (AFAIK the first
full-featured, object oriented adventure language), also available
from the IF-archive. AdvSys is very Lisp-like. It was written up in
an article in Byte around 1986 or so.

--
Magnus Olsson (m...@df.lth.se, zeb...@pobox.com)
------ http://www.pobox.com/~zebulon ------

Martin Dransfield

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
m...@bartlet.df.lth.se (Magnus Olsson) writes:

> You may also get some ideas from AdvSys (AFAIK the first
> full-featured, object oriented adventure language), also available
> from the IF-archive. AdvSys is very Lisp-like. It was written up in
> an article in Byte around 1986 or so.

Again, thanks. I didn't know AdvSys was lisp-like. Will investigate.

-- Martin

Palmer Davis

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
Martin Dransfield wrote:
>
> Is this a standard JVM implementation of Scheme like Kawa or SILK?
> Or is it specially adapted for IF?

I'm using Kawa for all of my prototyping; all of the "IF-ness"
is done at the library level. Some of the library stuff (the
object system, the dictionary, et cetera) then turns into
basic language constructs when the Z-machine compiler is used.
I call it "Zcheme;" a legal Zcheme program should run on any
R5RS-complient Scheme interpreter, on the JVM, and if you follow
certain guidelines, it should compile into Z-code. The idea is
to provide a migration path beyond the Z-machine.

Magnus Olsson wrote:
>
> You may also get some ideas from AdvSys (AFAIK the first
> full-featured, object oriented adventure language), also available
> from the IF-archive. AdvSys is very Lisp-like. It was written up in
> an article in Byte around 1986 or so.

In fact, AdvSys was my point of departure for all of this. Way
back in the mists of time, before there was an Inform, I did all
of my IF development in AdvSys. TADS was still in version 1.x,
and wasn't free, and wouldn't run on my NeXT in any event. A
lot of what's in Zcheme now is the result of some brainstorming
on what would have to be added to AdvSys to bring it up to the
level of functionality that TADS had in those days; I found that
I could do most of it by hacking up the library without touching
the AdvSys code itself. Later, I discovered Scheme, and wound
up porting all of my old AdvSys code over.

Joe Mason

unread,
Dec 11, 1998, 3:00:00 AM12/11/98
to
Palmer Davis <pal...@secant.com> wrote:
>I call it "Zcheme;" a legal Zcheme program should run on any
>R5RS-complient Scheme interpreter, on the JVM, and if you follow
>certain guidelines, it should compile into Z-code. The idea is
>to provide a migration path beyond the Z-machine.

I like this idea a lot. I've been a little worried about the Z-machine's
restrictions: it sounds like a Zcheme program that overflows the Z-machine
will still run on JVM or Scheme interpreters, so it wouldn't require
completely reimplementing the thing.

(I'm thinking something like the way Jigsaw was split into two V5 files for
non-V8 interpreters. I could split my Zcheme program into chapters, and
distribute the complete version for JVM/Scheme, and smaller versions for
Z-Machine on small architectures.)

Joe
--
Congratulations, Canada, on preserving your national igloo.
-- Mike Huckabee, Governor of Arkansas

Dan Shiovitz

unread,
Dec 12, 1998, 3:00:00 AM12/12/98
to
In article <366f2373.1016718@news>, Eric O'Dell <eod...@pobox.com> wrote:
[..]

>* The parser will understand adverbs (finally) and many kinds of
>interrogatives in addition to the usual imperative commands.
[..]

If you think the reason current IF systems do not support adverbs is
because no one is able to program them, you are mistaken. (Hint:
adding adverbs is equivalent to, say, making "get" and "take" separate
verbs instead of synonyms. The current thrust is towards increasing
playability by creating more ways to say the same thing; why would you
want more possibilities to choose from, most of which don't do
anything?)

Interrogatives would be nice, though (and, of course, are done already
in a few games. But as a general thing, they're nice.)

>--Eric
--
Dan Shiovitz || d...@cs.wisc.edu || http://www.cs.wisc.edu/~dbs
"...Incensed by some crack he had made about modern enlightened
thought, modern enlightened thought being practically a personal buddy
of hers, Florence gave him the swift heave-ho and--much against my
will, but she seemed to wish it--became betrothed to me." - PGW, J.a.t.F.S.

David Glasser

unread,
Dec 12, 1998, 3:00:00 AM12/12/98
to
Dan Shiovitz <d...@cs.wisc.edu> wrote:

> Interrogatives would be nice, though (and, of course, are done already
> in a few games. But as a general thing, they're nice.)

Ok, so they don't teach enough grammar in school these days. What do
you mean by this? WHO IS THAT EVIL GUY? WHAT ARE MONKEYS?

--
David Glasser gla...@NOSPAMuscom.com http://onramp.uscom.com/~glasser
DGlasser @ ifMUD : fovea.retina.net 4000 (webpage fovea.retina.net:4001)
Sadie Hawkins, official band of David Glasser: http://www.port4000.com/
rec.arts.int-fiction FAQ: come.to/raiffaq

Eric O'Dell

unread,
Dec 14, 1998, 3:00:00 AM12/14/98
to
On 12 Dec 1998 06:12:34 GMT, d...@cs.wisc.edu (Dan Shiovitz) wrote:

>If you think the reason current IF systems do not support adverbs is
>because no one is able to program them, you are mistaken.

I'm well aware that adverbs are blitheringly simple to implement.
Shockingly so, actually, as the grammatical constraints on the proper
usage of adverbs are much looser than with other parts of speech.

>(Hint:
>adding adverbs is equivalent to, say, making "get" and "take" separate
>verbs instead of synonyms. The current thrust is towards increasing
>playability by creating more ways to say the same thing; why would you
>want more possibilities to choose from, most of which don't do
>anything?)

_Used with care_, adverbs could diminish the amount of verb guessing
that inevitably goes on even in the better games. Presented with a
vial of nitroglycerin, it is easier to type

gently pick up the vial

than it is to cast around for a single word that means "get gently".
The absence of adverbs and a large number of imperfect verb synonyms
can lead to (admittedly extreme) cases like

>grab the nitroglycerin
Taken.

which really should produce results like

>grab the nitroglycerin
You fly to bits.

Personally, I suspect the main reason that adverbs haven't been widely
implemented is because they impose a substantially greater planning
burden on the implementor and an equally great implementation burden,
not because the parser mechanics are particularly difficult. My
intention here isn't to advocate the use of adverbs, or necessarily
even to use them myself; I simply want to make them available so that
someone cleverer and more dedicated than myself can surprise us with
them further down the road.

As far as "the current thrust" is concerned, well, I'm philosophically
opposed to following it. The great risk of failure in diverging from
the main track is a fair price to pay for the small chance of success
in new territories.

>Interrogatives would be nice, though (and, of course, are done already
>in a few games. But as a general thing, they're nice.)

Fair warning -- CAVE will not provide the underlying knowledge
representation code (at least initially) to answer interrogatives. The
current extent of my ambitions in this area is to make sure the parser
can take a certain class of simple questions and build an AST from
them which can be conveniently handled by implementor-supplied
routines. I'm not sure how far this will go in the first version, but
at a minimum, the parser should handle most simple state-of-being
questions beginning with an interrogative pronoun, e.g.,

>Bob, what color are your eyes?
>Jill, who is your husband?
>Dave, where is the secret door?

This is being implemented as a separate interrogative parser, so that
changes can be made without affecting the imperative parser. All that
will be required of someone who wishes to alter or extend the parsers
later will be to make adjustments to the routines that distinguish
between types of commands.

Martin Dransfield

unread,
Dec 14, 1998, 3:00:00 AM12/14/98
to
Palmer Davis <pal...@secant.com> writes:

> I'm using Kawa for all of my prototyping; all of the "IF-ness"
> is done at the library level. Some of the library stuff (the
> object system, the dictionary, et cetera) then turns into
> basic language constructs when the Z-machine compiler is used.

> I call it "Zcheme;" a legal Zcheme program should run on any
> R5RS-complient Scheme interpreter, on the JVM, and if you follow
> certain guidelines, it should compile into Z-code. The idea is
> to provide a migration path beyond the Z-machine.

This sounds rather cool. I had intended using Common Lisp rather than
Scheme as a base (personal preference) but I can see the advantages of
using something that has been implemented in Java already. I will
carry on tinkering for my personal gratification but I look forward to
seeing Zcheme.

I'll be happy to assist with testing, etc. when it's ready.

Regards,

Martin

David Glasser

unread,
Dec 14, 1998, 3:00:00 AM12/14/98
to
Eric O'Dell <eod...@pobox.com> wrote:

> _Used with care_, adverbs could diminish the amount of verb guessing
> that inevitably goes on even in the better games. Presented with a
> vial of nitroglycerin, it is easier to type
>
> gently pick up the vial
>
> than it is to cast around for a single word that means "get gently".
> The absence of adverbs and a large number of imperfect verb synonyms
> can lead to (admittedly extreme) cases like

I think that Dan's point is that an even better solution would be
>PICK UP THE VIAL
In other words, *no* vocab guessing at all, just action guessing.

--
David Glasser gla...@NOSPAMuscom.com http://onramp.uscom.com/~glasser

DGlasser @ ifMUD : fovea.retina.net:4001 | r.a.i-f FAQ: come.to/raiffaq

Paul O'Brian

unread,
Dec 14, 1998, 3:00:00 AM12/14/98
to
On Mon, 14 Dec 1998, Eric O'Dell wrote:

> _Used with care_, adverbs could diminish the amount of verb guessing
> that inevitably goes on even in the better games. Presented with a
> vial of nitroglycerin, it is easier to type
>
> gently pick up the vial
>
> than it is to cast around for a single word that means "get gently".

May I suggest, then, that CAVE implementors provide in the game's
documentation a list of the adverbs it knows? Otherwise, you run the risk
of situations like this:

>CAREFULLY PICK UP THE VIAL
I don't know that adverb.

>GINGERLY PICK UP THE VIAL
I don't know that adverb.

>CAUTIOUSLY PICK UP THE VIAL
I don't know that adverb.

>PICK UP THE VIAL
You fly to bits.

Also, I think you're probably already planning this, but you do want to
accomodate both "prefix" and "postfix" (I can't think of better terms this
early in the morning) usage of adverbs: GENTLY PICK UP THE VIAL or PICK UP
THE VIAL GENTLY. Please forgive me if you've already announced you plan to
do this.

Paul O'Brian
obr...@colorado.edu
http://ucsu.colorado.edu/~obrian


Michael Straight

unread,
Dec 15, 1998, 3:00:00 AM12/15/98
to
On Mon, 14 Dec 1998, Eric O'Dell wrote:

> _Used with care_, adverbs could diminish the amount of verb guessing
> that inevitably goes on even in the better games. Presented with a
> vial of nitroglycerin, it is easier to type
>
> gently pick up the vial
>
> than it is to cast around for a single word that means "get gently".

> The absence of adverbs and a large number of imperfect verb synonyms
> can lead to (admittedly extreme) cases like
>

> >grab the nitroglycerin
> Taken.
>
> which really should produce results like
>
> >grab the nitroglycerin
> You fly to bits.

Ugh. Guess the verb is bad enough. Guess the adverb would be worse. If
I mean to break the vial, I'll type "break vial." If I want to pick it up
without damaging it, I'll type "take vial."

If what you're after is to penalize the player for carelessly grabbing
a vial before she knows what it is, then the puzzle should be figuring out
what is in the vial, after which the protagonist should automatically pick
up the nitro gently when the player types "take vial." The puzzle should
not be figuring out how to tell the parser to do what the protagonist
would naturally do.

SMTIRCAHIAGEHLT

Eric O'Dell

unread,
Dec 16, 1998, 3:00:00 AM12/16/98
to
On Mon, 14 Dec 1998 15:20:15 -0700, Paul O'Brian
<obr...@ucsu.Colorado.EDU> wrote:

>May I suggest, then, that CAVE implementors provide in the game's
>documentation a list of the adverbs it knows?

A worthy point. It would be no big deal to provide support directly in
the CAVE engine for retrieving vocabulary lists (subject, of course,
to their being deliberately hidden by the implementor).

>Also, I think you're probably already planning this, but you do want to
>accomodate both "prefix" and "postfix" (I can't think of better terms this
>early in the morning) usage of adverbs: GENTLY PICK UP THE VIAL or PICK UP
>THE VIAL GENTLY. Please forgive me if you've already announced you plan to
>do this.

I haven't announced it, but it's already planned. :) What I discovered
when I was creating syntax flow charts was that English adverbs can
appear almost anywhere in a sentence without being strictly
ungrammatical, especially when verbs are paired with prepositions as
with "PICK UP". My solution, rather than worry about the few instances
where an adverb is ungrammatical (and the special cases where certain
verb-adverb-preposition idioms _are_ legal, which never occur to you
at the time, but which will inevitably be pointed out by some beta
tester or raif poster weeks after the parser logic is so thoroughly
crufty that the thought of reworking it brings tears), is simply to
allow an adverb anywhere in the sentence. Thus

pick up the vial gently


gently pick up the vial

pick the vial up gently
gently pick the vial up

are all legal and equivalent, as are crawling syntactical horrors like

pick gently up the vial
pick up gently the vial
pick the vial gently up

In fact, as a general rule, I am trying to make all manner of
comprehensible but ungrammatical sentences possible just so long as
the sense of the sentence can still be extracted by the parser. This
is partly intended to reduce the frustration of newbies who do not yet
know if-ese, partly to make it easier to add support for other
languages later, and partly to accommodate the grunting shorthand a
lot of veteran players like to use.

(Surprisingly -- to me, anyway -- much of the difficulting in
constructing the parser lies not in parsing normal English, but in
interpreting un-English phrases like "get green" as meaning "get the
green orb".)

Andrew Plotkin

unread,
Dec 16, 1998, 3:00:00 AM12/16/98
to
Eric O'Dell (eod...@pobox.com) wrote:

> On Mon, 14 Dec 1998 22:41:31 GMT, gla...@DELETEuscom.com (David
> Glasser) wrote:

> >I think that Dan's point is that an even better solution would be
> > >PICK UP THE VIAL
> >In other words, *no* vocab guessing at all, just action guessing.

> Taken to the extreme, that road leads to using USE as an all-purpose
> verb.

But I don't advocate taking it to the extreme. I advocate leaving it where
it is, which is a well-understood mix of verbs and ranges of action.

> I'm not suggesting that adverbs are necessarily a good idea, just that
> they really haven't been formally made available to IF authors

I don't think that's true. Authors have implemented much stranger things
in both TADS and Inform. I think that anyone who wanted adverbs would have
used them by now.

--Z

--

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

Eric O'Dell

unread,
Dec 16, 1998, 3:00:00 AM12/16/98
to
On Mon, 14 Dec 1998 22:41:31 GMT, gla...@DELETEuscom.com (David
Glasser) wrote:

>I think that Dan's point is that an even better solution would be
> >PICK UP THE VIAL
>In other words, *no* vocab guessing at all, just action guessing.

Taken to the extreme, that road leads to using USE as an all-purpose

verb. Sometimes, figuring out what needs to be done is more than just
a vocabulary game, as in that venerable class of puzzles in which an
object must be used in an unusual way.

I'm not suggesting that adverbs are necessarily a good idea, just that

they really haven't been formally made available to IF authors, and it
could be that someone will think of a way to use them that will make
people say, "So _that's_ what you can do with adverbs!" (I'm guessing
that you could really go far in comedy IF with adverbs.) So I'm
providing them.

IMHO, the really interesting features of CAVE are the event message
passing scheme and the collective/subdivisible objects. (I'm also
giving some thought to implementing objects that can genuinely be
present in multiple locations, so that you could finally implement a
realistic rope without the usual ugly hacks.)

Joe Mason

unread,
Dec 16, 1998, 3:00:00 AM12/16/98
to
Eric O'Dell <eod...@pobox.com> wrote:
>allow an adverb anywhere in the sentence. Thus
>
> pick up the vial gently
> gently pick up the vial
> pick the vial up gently
> gently pick the vial up
>
>are all legal and equivalent, as are crawling syntactical horrors like
>
> pick gently up the vial
> pick up gently the vial
> pick the vial gently up

Do not pick gently up the vial,
Acid should burn and rave at close of day;
Rage, rage against the lighting of the dial.

Eric O'Dell

unread,
Dec 16, 1998, 3:00:00 AM12/16/98
to
On Wed, 16 Dec 1998 06:07:55 GMT, erky...@netcom.com (Andrew Plotkin)
wrote:

>> Taken to the extreme, that road leads to using USE as an all-purpose
>> verb.
>

>But I don't advocate taking it to the extreme. I advocate leaving it where
>it is, which is a well-understood mix of verbs and ranges of action.

Well-understood to whom? To the existing IF crowd, many of whom have
been playing IF games for fifteen years or more and who have _very_
conservative ideas about what IF is and is not? Or to a newcomer who,
upon encountering a command language that resembles English, attempts
to use the language of everyday life? At the very least, the parser
should be able to recognize adverbs for what they are, if only to
inform the player that they are not necessary to the game.

As far as "leaving it where it is" goes in general, I think that's a
large part of the problem with "modern" IF. While there have been
noteworthy advances and experiments on the artistic side of the
question, to which you yourself have contributed, the underlying
technology has largely stagnated for no reason I can see, other than
perhaps nostalgia. Other than the multimedia extensions to the
standard IF interface provided by TADS and Hugo, the basic technology
has not changed significantly since Asteroids was the state of the art
in graphical games.

This is apparently viewed as a virtue by many. Personally, I am not in
agreement with that view. It is my hope that CAVE and the other new
systems which have been announced will provide a concrete
demonstration that the Z-Machine is not the ultimate pinnacle of the
art, or even close to it. I may fail, but I may not. And I don't think
that --- with respect to what I think is the root of much of the
conservatism in these parts --- that a new or different kind of IF
will hurt "traditional" IF.

>> I'm not suggesting that adverbs are necessarily a good idea, just that

>> they really haven't been formally made available to IF authors
>
>I don't think that's true. Authors have implemented much stranger things
>in both TADS and Inform. I think that anyone who wanted adverbs would have
>used them by now.

I think that it's too early in the evolution of the artform to say
what is or is not wanted, needed, or useful. The novel languished for
two millennia between the Roman novelists and the present era of its
reign as the dominant form of prose literature. Interactive fiction
scarcely has two decades under its belt.

Andrew Plotkin

unread,
Dec 16, 1998, 3:00:00 AM12/16/98
to
Eric O'Dell (eod...@pobox.com) wrote:

> As far as "leaving it where it is" goes in general, I think that's a
> large part of the problem with "modern" IF. While there have been
> noteworthy advances and experiments on the artistic side of the
> question, to which you yourself have contributed, the underlying
> technology has largely stagnated for no reason I can see, other than
> perhaps nostalgia. Other than the multimedia extensions to the
> standard IF interface provided by TADS and Hugo, the basic technology
> has not changed significantly since Asteroids was the state of the art
> in graphical games.
>
> This is apparently viewed as a virtue by many. Personally, I am not in
> agreement with that view. It is my hope that CAVE and the other new
> systems which have been announced will provide a concrete
> demonstration that the Z-Machine is not the ultimate pinnacle of the
> art, or even close to it.

At this point, you're unfairly conflating VM technology, language
technology, library technology, parser design technology, and -- for lack
of a better term -- game-design "technology".

I both agree and disagree with you, depending on which of those you're
talking about. :-) But I don't have time for a full response right now.

You're implementing your ideas, which is all I could ask for anyway.

> >> I'm not suggesting that adverbs are necessarily a good idea, just that
> >> they really haven't been formally made available to IF authors
> >
> >I don't think that's true. Authors have implemented much stranger things
> >in both TADS and Inform. I think that anyone who wanted adverbs would have
> >used them by now.

> I think that it's too early in the evolution of the artform to say
> what is or is not wanted, needed, or useful.

It's too early to say what *will be* wanted or useful; but it's easy to
check what *has been* wanted. Dejanews covers most of it. The idea of
adverbs has been broached several times on this newsgroup, but I don't
remember anyone ever saying, hey, I *want* to use adverbs in my game, how
do I implement it?

(Or if they have, I'm betting they got a workable answer.)

George Caswell

unread,
Dec 16, 1998, 3:00:00 AM12/16/98
to
On 12 Dec 1998, Dan Shiovitz wrote:

> In article <366f2373.1016718@news>, Eric O'Dell <eod...@pobox.com> wrote:
> [..]
> >* The parser will understand adverbs (finally) and many kinds of
> >interrogatives in addition to the usual imperative commands.
> [..]
>

> If you think the reason current IF systems do not support adverbs is

> because no one is able to program them, you are mistaken. (Hint:


> adding adverbs is equivalent to, say, making "get" and "take" separate
> verbs instead of synonyms. The current thrust is towards increasing
> playability by creating more ways to say the same thing; why would you
> want more possibilities to choose from, most of which don't do
> anything?)
>

While it's true that effectively using new possibilities is often
difficult, giving the game writer more flexibility is always a good thing..

> Interrogatives would be nice, though (and, of course, are done already
> in a few games. But as a general thing, they're nice.)
>

I must agree wholeheartedly with this part-- Interrogatives are nice.
Typically the player gets all their information by doing "Look at .." - which
works, for the most part. Getting direct answers could be quite an
improvement, however.

It's my belief that the technology involved could definately stand a little
improvement-- In the early years these games had to be stuffed into most of
their target platforms with a crowbar. Now everyone and their sister has a
reasonably modern PC, and the number of systems that require the crowbar
treatment is ever decreasing... even the palmtops have some elbowroom. I
think it's a worthwhile challenge to use some of this elbow room to apply some
AI techniques-- they won't quite pass the Turing test, but it does lead to
the solutions of a lot of these challenges. Finding the best representation
for the data is certainly one of the toughest, but once that's solved a lot of
other problems become much simpler.

Offhand I can think of a bunch of fun things that could be done in IF with
a little extra data/processing space and some AI techniques.. NPC's with some
simple problem solving abilities, and the ability to explain their actions,
for example. Or if all objects had a little descriptive information in their
data fields, interrogatives could be handled, NPC's could identify and
describe objects, etc.

Now a lot of games that get made aren't going to have a lot of use for any
new features, extra processing space, AI searching and guessing and matching
techniques, etc- which actually is good, since these games will continue to
be usable on bare-bones platforms. IMHO, however, authors will be much more
keen to take advantage of such features once a few proper demonstrations have
been made of what can be done with them.

As for choice/design of new VM for the games, I'd suggest Scheme. It's a
very powerful way to express things, with a clean design, etc.

---GEC

Den of Iniquity

unread,
Dec 17, 1998, 3:00:00 AM12/17/98
to
On Wed, 16 Dec 1998, Eric O'Dell wrote:

> On Wed, 16 Dec 1998 06:07:55 GMT, Andrew Plotkin wrote:
>> I advocate leaving it where it is, which is a well-understood mix of
>> verbs and ranges of action.

> Well-understood to whom? To the existing IF crowd, many of whom have
> been playing IF games for fifteen years or more and who have _very_
> conservative ideas about what IF is and is not? Or to a newcomer who,
> upon encountering a command language that resembles English, attempts
> to use the language of everyday life?

To put it briefly, however you work it, the biggest problem with adverbs
is the increase in complexity they bring. In order to make them work
properly, you have to facilitate their use throughout; if you only
incorporate adverbs partially - that is, if most of the time, they don't
make any difference to the way your i-f progresses - then the player will
learn not to use them - maybe having a detrimental effect when they _are_
needed. So you need to implement an average of three or four adverbs (plus
synonyms) for every verb, along with all the special cases for special
objects of those verbs. You can see where this goes. Might as well
incorporate lots of prepositions: 'look to the left of' and 'look to the
right of' along with 'look under' and 'look behind'. Unless your world
already contains a pretty advanced physics model, it means a lot of extra
writing for the author.

> At the very least, the parser should be able to recognize adverbs for
> what they are, if only to inform the player that they are not necessary
> to the game.

Certainly, though surely it's enough to make a note in the average 'how to
play this game' menu selection?

> As far as "leaving it where it is" goes in general, I think that's a
> large part of the problem with "modern" IF.

There meay well be a place for it. You're doing the right thing: giving it
a go. We're just trying to draw your attention to the problems you're
facing. It's _not_ just a hankering for the traditional way. IMO, for
adverbs to work properly, you need a better fleshed-out model of the i-f
world than Inform/TADS/Hugo libraries define as standard. For example, if
you're going to do something 'quickly' or 'slowly' and allow speedy or
tardy actions throughout, you need a more precise model of the timing of
the world than the normal, weirdly flexible quanta of 'turns', in one of
which you might say hello to a sailor or press a button and watch a small
alien fish perform a variety of involuntary acrobatics before
inadvertently flying into your ear.

The z-machine isn't the be all and end all of interactive fiction. Not by
a long way. Indeed the limitations of the system to some extent define the
state of the art. But we're a long way from exhausting the possibilities
of the system, so it's not time to write it off yet.

>It is my hope that CAVE and the other new systems...

Out of curiosity, is CAVE so called after the colossal variety or the
Latin imperative, 'beware'? :)

--
Den


Eric O'Dell

unread,
Dec 17, 1998, 3:00:00 AM12/17/98
to
On Thu, 17 Dec 1998 13:30:36 +0000, Den of Iniquity
<dms...@york.ac.uk> wrote:

>To put it briefly, however you work it, the biggest problem with adverbs
>is the increase in complexity they bring. In order to make them work
>properly, you have to facilitate their use throughout; if you only
>incorporate adverbs partially - that is, if most of the time, they don't
>make any difference to the way your i-f progresses - then the player will
>learn not to use them - maybe having a detrimental effect when they _are_
>needed. So you need to implement an average of three or four adverbs (plus
>synonyms) for every verb, along with all the special cases for special
>objects of those verbs. You can see where this goes. Might as well
>incorporate lots of prepositions: 'look to the left of' and 'look to the
>right of' along with 'look under' and 'look behind'. Unless your world
>already contains a pretty advanced physics model, it means a lot of extra
>writing for the author.

[...snip...]

> For example, if
>you're going to do something 'quickly' or 'slowly' and allow speedy or
>tardy actions throughout, you need a more precise model of the timing of
>the world than the normal, weirdly flexible quanta of 'turns', in one of
>which you might say hello to a sailor or press a button and watch a small
>alien fish perform a variety of involuntary acrobatics before
>inadvertently flying into your ear.

I take it for granted that it means a lot more work for the author,
not only in terms of coding but especially in terms of planning. Every
adverb, as you note, could conceivably be combined with every verb by
the user. While CAVE will provide some tools to help cut down on the
amount of implementor labor involved[1], every verb/adverb combination
will require consideration of what to do with it, even if it is only
to ignore the adverb or to issue a witty error message.

However, as George Caswell notes a little further down the thread, one
of the reasons that traditional IF tends to be lightweight and
somewhat sparse is because it started on machines that are outclassed
by today's palmtops. We have the processor speed and the memory to do
much more, and perhaps it is time to look at what the extra effort
could do for IF in the way of providing greater realism and depth (and
hopefully increasingly playability at the same time).

>The z-machine isn't the be all and end all of interactive fiction. Not by
>a long way. Indeed the limitations of the system to some extent define the
>state of the art. But we're a long way from exhausting the possibilities
>of the system, so it's not time to write it off yet.

Perish the thought. The challenge of coding in a restricted
environment encourages creative and efficient solutions that can be
carried elsewhere. God forbid that greater resources lead to
needlessly bloated IF the way it led to 20-megabyte word processors.

>>It is my hope that CAVE and the other new systems...
>
>Out of curiosity, is CAVE so called after the colossal variety or the
>Latin imperative, 'beware'? :)

It's actually an acronym for C AdVenture Engine, and is meant as a pun
on Colossal Cave. But considering what a pain it will be to use in its
early versions -- until I have time to code some additional support
tools -- you may as well take it in the Latin sense. ;-)

-E.


[1] CAVE verb actions are not necessarily bound directly to noun
objects as they are in TADS. When the implementor defines a verb, he
may cause it to call a method in the actor, direct object, indirect
object, or the verb itself, or any combination of the above. For
example, the GET action might invoke a member of function of the verb
object which checks the weight and bulk of the direct object and
compares it to the restrictions of the actor, and then calls methods
in the actor and the direct object (and any number of other objects)
to see if there are any special cases which prohibit acting in
accordance with the default physics model. (Adverbs might also be
handled centrally in this fashion.) The author might also implement
the same arrangement in a more traditionally O-O fashion, but CAVE
leaves him free to break out of the O-O model where desirable.

Anywhere along the line, event messages can be automatically passed up
and down the chains of containment and proximity to query other
objects to see if they would like to react to the change in state --
the GOLDSKUL.T example from TADS springs to mind as a situation where
this might apply. This should make modelling of physics easier,
allowing players to use objects in realistic but unanticipated ways,
freeing precious time for the implementor to work on witty responses
to player commands like SENSUOUSLY PUT ON SOCKS. <g>

Jonadab the Unsightly One

unread,
Dec 18, 1998, 3:00:00 AM12/18/98
to
jcm...@uwaterloo.ca (Joe Mason) wrote:

> Do not pick gently up the vial,
> Acid should burn and rave at close of day;
> Rage, rage against the lighting of the dial.

[GROAN.]


- jonadab

Reply all
Reply to author
Forward
0 new messages