From adamc@acpub.duke.edu Thu Aug 28 23:08:41 MET DST 1997
Article: 27999 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!howland.erols.net!agate!newsgate.duke.edu!usenet
From: Adam Cadre <adamc@acpub.duke.edu>
Newsgroups: rec.arts.int-fiction
Subject: Re: Nonlinearity
Date: Thu, 28 Aug 1997 16:21:08 -0400
Organization: Group 17
Lines: 10
Message-ID: <3405DDB4.1D52@acpub.duke.edu>
References: <01bcb083$e874bbe0$539f72ce@hammurabi.blarg.net> <erkyrathEFG5Kv.6Dv@netcom.com> <01bcb13c$a1ec3fe0$959f72ce@hammurabi.blarg.net> <erkyrathEFHBGy.57u@netcom.com> <01bcb379$27d95320$869f72ce@hammurabi.blarg.net> <erkyrathEFMsDG.Lvo@netcom.com>
Reply-To: adamc@acpub.duke.edu
NNTP-Posting-Host: res-9-43.dorm.duke.edu
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01Gold (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:27999

Andrew Plotkin wrote:
> The program is programmatic, but the game, or story -- the ultimate
> effect on the player -- is in the player's head, and that's the real
> goal here.

Ding!  Zarf wins our daily drawing.

  -----
 Adam Cadre, Durham, NC
 http://www.duke.edu/~adamc


From carl@earthweb.com Fri Aug 29 11:08:00 MET DST 1997
Article: 28005 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!infeed1.internetmci.com!newsfeed.internetmci.com!199.201.128.9!news.silcom.com!news.genuity.net!not-for-mail
From: Carl Muckenhoupt <carl@earthweb.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: Nonlinearity
Date: Thu, 28 Aug 1997 19:21:32 -0400
Organization: EarthWeb
Lines: 130
Message-ID: <340607FC.6327@earthweb.com>
References: <01bcb083$e874bbe0$539f72ce@hammurabi.blarg.net> <erkyrathEFG5Kv.6Dv@netcom.com> <01bcb13c$a1ec3fe0$959f72ce@hammurabi.blarg.net> <34016297.5A25@acpub.duke.edu> <01bcb1de$e14b5d40$989f72ce@hammurabi.blarg.net> <3402DD9E.295F@acpub.duke.edu> <01bcb37d$a92e50c0$869f72ce@hammurabi.blarg.net>
Reply-To: carl@earthweb.com
NNTP-Posting-Host: ben.earthweb.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.0 (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:28005

Brandon Van Every wrote:
> 
> > I also had problems with the use of letters to stand for... what?
> > Events?  Locations?  Game states?
> 
> All of the above.  If you think like a programmer :-) you tend to see all 3
> of those notions as equivalent.

I don't see them as equivalent at all, at least in terms of linearity.
Let's consider these on a high level.  Only consider significant parts
of game state (ie, "Does the player know the killer's identity?" rather
than "What are the locations of all the objects?"), only consider those
events that alter game state in that sense, and clump together groups of
mutually-accessible locations.  In other words, only consider the things
that affect what the player perceives as the story.

Now, gaining access to a new location is certainly a kind of
state-altering event.  But it's only one kind.  And certainly, the set
of available location is part of the state, but it's only part. It's
perfectly possible to alter game state without entering a new location,
and most irreversible (or not-directly-reversible; see "Tube Trouble"
for examples) changes in game state do not cut off access to old
locations.  Thus, except in certain very linear games (much of the
Sierra canon springs to mind), locations are not equivalent to events or
states.  And at any rate, any reasonably nonlinear game isn't a
frogmarch from location to location until you've been to each exactly
once, but involves wandering around a lot and revisiting familiar
locations for new reasons - thus, your A-Z analysis is scarcely
applicable to them.  A typical game would be more like
ABABABCABACBADAEFBACEBAEFABG.

Events and states are more closely related, and in a highly linear game
there would be a one-to-one relationship: each event would put the game
into a particular state.  However, in a more typical game, game state is
composed of multiple elements.

A simple example: Once you've obtained the Holy Grail and the
Philosopher's Stone, you can perform the ritual that grants you eternal
life.  (I realize that this is exactly the kind of thing you say you're
tired of; I choose the example for exactly that reason.)  The Grail and
the Stone both have other uses (healing and transmutation,
respectively), and thus possession of either is significant; perhaps
possession of one even affects how the other can be obtained.

There are three Events: gain grail, gain stone, perform ritual.  They
can be performed in two orders: GSR or SGR.  At any rate, all three
events will occur over the course of the game.

There are five States: have nothing, have grail, have stone, have both
grail and stone, and immortal.  We still have only two possible
orderings (it's still the same game, after all!), but each uses only
four of the states: NGBI or NSBI.

Thus, it is very important to your analysis whether the letters
represent Events or States.  If they represent States, then your
assertion that most games simply mix the ordering around is false.  Most
games have many states that the player never sees.

> > I don't think of stories, interactive
> > or otherwise, as a matter of "traversing nodes."
> 
> Even though this is the underlying programmatic structure of nearly all
> text adventures?  Certainly all the Infocom games were (in mathematical
> terms) a collection of rooms ("nodes") linked by passageways ("edges")
> creating the entire game universe (the "graph").
> Most adventure games are merely "traverse all the nodes of the graph until
> you've acquired all the available resources."  That's why most adventure
> games "feel" the same, it's structural.

Up until this point, I thought the graph traversal you were referring to
was a traversal of a plot DAG.  So all you want is rooms that you're not
required to enter? :)

> I would also say that the same holds true of the average Hollywood film.
> There is a very canned set of nodes that will be traversed, such as:
> shoot-out at the beginning, one person you like must be killed but it can't
> be the protagonist, obligatory sex scene between male and female lead,
> shoot-out at the end, etc.  I could go on about other genres, I'll spare us
> all.  :-)
>
> I became tired of IF when I realized that my algorithm to win any game was
> the same.
> "Traverse the nodes acquiring all available resources, use the resources to
> gain access to unavailable nodes, until the "final" node is reached."
> 

All I can say about this is that your Hollywood formula refers only to
content, and your IF formula refers only to form.  Indeed, it's more of
a format than a formula.

Not to discourage experimentation, but this complaint strikes me as a
bit like "All books are the same.  You start at the beginning, you read
words in order until you get to the end, and then you stop.  When I
realized this, I stopped reading books."  There's a lot that can be done
within a particular format, and we're far from exhausting ours. 
Furthermore, there are distinct reasons why this format has been as
popular as it is - chiefly that it provides the player with motivation.

That said, any new forms we can play with will only improve the world. 
But please, you don't need to tell us that our games are boring in order
to convince of this.

> > What does your argument help us do or understand
> > that we couldn't do or didn't understand before?
> 
> It would help engineer a simulated universe, rather than insisting on
> marching a person from start-to-finish.  It might also make that universe
> "something more than boring."  When I say "engineer," I mean that there are
> both theoretical requirements of plot device, and labor requirements for
> actually getting all the work done.  One thing I developed is how much
> extra work you've got to do, if you want to make a game based on relational
> histories instead of independent elements thrown into a grab-bag.

Please explain.  What is a "relational history"?  A history of
relationships?  Between what?

> As the
> length of the (recent) history increases, the additional work is
> unfortunately n(n-1), n(n-1)(n-2), n(n-1)(n-2)(n-3), ... etc.  This becomes
> a lot of work rather fast!

What does the value n represent?  The only thing I can think of that
would produce this effect is a grab-bag of n basically unrelated
elements, but with each combination and ordering of these elements made
into a special case for which the author writes code and/or prose.  Is
this what you're proposing?

-- 
Carl Muckenhoupt		carl@earthweb.com
EarthWeb			http://www.earthweb.com/


From mol@bartlet.df.lth.se Fri Aug 29 18:37:51 MET DST 1997
Article: 28026 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: Is it still prossible to make money off text adventures?
Date: 29 Aug 1997 18:18:33 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 87
Message-ID: <5u6sop$a6e$1@bartlet.df.lth.se>
References: <O40Ih21s8GA.281@nih2naac.prod2.compuserve.com> <01bcb46c$16ef9940$859f72ce@hammurabi.blarg.net> <5u6g3s$uvo$1@bartlet.df.lth.se> <3407ee30.2496342@news.u.washington.edu>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:28026

In article <3407ee30.2496342@news.u.washington.edu>,
Matthew Amster-Burton <mamster@u.washington.edu> wrote:
>mol@bartlet.df.lth.se (Magnus Olsson) wrote:
>
>>This is a variation of the commonly repeated myth that a free market
>>promotes products with low quality. Companies, it is thought,
>>deliberately make products of low quality so that they'll break often,
>>thus leading to higher sales.
>
>That's not an argument I've ever heard before, although it sounds like
>a variation on the (true) stories about IBM making sure each new
>system rev required more memory.

IBM is a bit of a special case; while they didn't have a true monopoly on
mainframes, the competition situation was very skewed in their favour, to
say the least. 

A similar situation holds for Microsoft today.

>>However, while there is some truth to it, it is obvious that high
>>quality is perceived as positive by consumers, hence companies compete
>>with high quality. A high-quality product can be sold for a higher
>>price, to start with. And once companies start using quality as a
>>competition weapon, we'll have a self-sustaining process. 
>
>But to capture the public eye, a company must not only deliver a
>product of high quality, but also spend a tremendous amount of money
>and person-power on advertising, i.e., convincing you that this
>product is better than competing products.  None of that work in any
>way enhances the quality of the product.

Indeed. The sad fact is that high quality is not enough to be a
commercial success - on the other hand, quality is not the only
desirable feature of a product, either.

>  Competition doesn't
>inherently promote high-quality products or low-quality ones:  if the
>consumer demand is for high quality, the market will, to some extent,
>deliver. 

That's true; I didn't mean to imply that competion automatically leads
to higher quality - I was debunking the idea that competition
automatically leads to *lower* quality, as well as the idea that
competition automatically leads to computer games with zero
replayability.

> That's asking a lot of the consumer, who can't be expected
>to have a reasonable understanding of the kind of product that could
>be delivered if cooperation, rather than competition, were the rule.

Frankly: if cooperation between the big corporations were the rule,
then we'd see very little, if any, progress at all. All proposals from
engineers to improve a product would be turned down by management,
since basically the *only* reason for management to spend money
improving products is to gain a competitive edge over the competition.

There are very good reasons for having anti-trust laws explicitly
forbidding cooperation among companies.

Of course, in a socialist economy, the government can *order* the
companies to improve their products. In theory, that is - we all know
how the practical attempts at implementation have turned out.

>Sorry to get socialist on you.

No problem. I didn't really intend to start a debate about the pros
and cons of the market economy; I just wanted to address one
particular (alleged) "con".

P.S. Market economy is far from ideal, of course. I just tend to
subscribe to the view that there's no better alternative around.

The most irritating thing about the market is that, like evolution, it
works toward *local* maxima, which means that it can work itself into
strange dead ends or vicious spirals. One such example seems to be the
market for adventure games: since everybody is doing flashy, graphical
things with lots of special effects and little depth, it makes more
financial sense to try to beat the competitors at their own
game. Which leads to even flashier and shallower new products, and so
on ad nauseam.



-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From erkyrath@netcom.com Fri Aug 29 20:36:32 MET DST 1997
Article: 28032 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!newscore.univie.ac.at!news.apfel.de!news.maxwell.syr.edu!europa.clark.net!205.252.116.205!howland.erols.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] Superclasses
Message-ID: <erkyrathEFosqn.1tI@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <T9Ng6RAEOdB0Ewxz@highmount.demon.co.uk>
Date: Fri, 29 Aug 1997 18:01:35 GMT
Lines: 82
Sender: erkyrath@netcom14.netcom.com
Xref: news.lth.se rec.arts.int-fiction:28032

Neil Brown (neil@highmount.demon.co.uk) wrote:
> Has anyone cracked superclasses? And are they available with Inform
> 6.05?

I think I've gotten them right. And the third-edition Inform manual, 
which contains the class/superclass system, says it's for Inform 6.04 and 
later.

> I have a situation where a class definition defines the before property,
> in order to capture a 'in' command in certain conditions. But several of
> the rooms that have this class also need to define the before property,
> and quite often also need to capture a direction command. Can the
> superclass feature of Inform handle this?

No. The inheritance of "before" routines is old; it goes back to Inform 5 
and probably earlier, long before the :: operator existed. It works 
differently.

The way it works is, the "before" property is "additive". This means that 
if you have 

Class roomclass
  with before [...]; ! function1

Object roomobj
  class roomclass,
  with before [...]; ! function2

then roomobj's "before" property will actually be two words long, and 
contain both function addresses. Just as if you'd done

[ function1; ... ];
[ function2; ... ];

Object roomobj
  with 
    before function2 function1;

Now, the before-clause checker knows that if a "before" property contains 
two functions like this, instead of one, it should run the first one, and 
then run the second one only if the first one returns 0.

Ignoring the technical details I've just explained:

First the object's "before" function is run. If that returns a nonzero 
value, that's all that happens. If it *does* return zero (rfalse), then 
the parent class's "before" function is run. If that also returns zero, 
the whole "before" process returns zero and the library continues to the 
standard verb routine.

So you could do stuff like this:

Class roomclass
  with before [;
    Go:
      switch (noun) {
        in_obj: "Going inside is boring.";
        default: rfalse;
      }
  ];

Object roomobj
  class roomclass,
  with before [;
      switch (noun) {
        e_obj: "The god of the east kicks you.";
        default: rfalse;
      }
  ];

and I think that's what you want.

Actually the "default: rfalse" is redundant; if there's no default, it 
just goes on to the end to the function, and hits the implicit return (0) 
at the end.

--Z

-- 

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


From mwdaly@kodak.com Sat Aug 30 15:00:16 MET DST 1997
Article: 28039 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!howland.erols.net!news-peer.sprintlink.net!news-sea-19.sprintlink.net!news-in-west.sprintlink.net!news.sprintlink.net!Sprint!159.56.2.2!thetimes.pixel.kodak.com!news.kodak.com!newsserver.rdcs.Kodak.COM!not-for-mail
From: mwdaly@kodak.com (Matthew Daly)
Newsgroups: rec.arts.int-fiction
Subject: Re: Nonlinearity
Date: Fri, 29 Aug 1997 19:59:30 GMT
Organization: Eastman Kodak Company
Lines: 95
Message-ID: <34142132.173174328@newsserver.rdcs.kodak.com>
References: <01bcb083$e874bbe0$539f72ce@hammurabi.blarg.net>  <340300A2.489@romulus.sun.csd.unb.ca>  <01bcb24c$bd7ee580$2e9f72ce@hammurabi.blarg.net> <Rc4iJGAY5zA0Iwi7@isla-mia.demon.co.uk> <87pvr0jsek.fsf@razumovsky.thecia.net>
Reply-To: mwdaly@kodak.com (Matthew Daly)
NNTP-Posting-Host: ep-psg-arrakis.ycc.kodak.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Newsreader: Forte Agent 1.01/32.397
Xref: news.lth.se rec.arts.int-fiction:28039

Dan Schmidt <dfan@thecia.net>, if that is your REAL name, said: 

>I wrote a good deal of Ultima Underworld, which is not one of the above
>two games but is fairly similar in many ways, being an Ultima.  For
>example, here are eight subquests corresponding to the eight virtues,
>etc.  Sometimes the quests depend on each other a bit (e.g., you can't
>get to part 2 of quest B until you've gotten to part 3 of quest D).  But
>there are usually a fair number of things you can do at once.  Then, at
>the end, it all gets pretty linear again - in fact, I don't think the
>last two levels succeeded that well.  RPG's do have the nice (sometimes)
>property of also having the "gain experience and items" quest, which can
>run in parallel with all the others but subtly interacts with them.

It worked for me.  Well, the endgame scene was hard to swallow, I suppose.
But, being an Ultima, the object is to prepare yourself for a large
challenge, and then to meet it.  There are a huge number of ways to prepare
yourself (do you focus on brawn or magic?  Is it worth fighting a dozen big
monsters to get a magic axe, or would you rather keep yourself focused on
the main quest?) but the ending is always going to be a dungeon march to
the climax in an Ultima.  (Well, other than U6 -- you could almost solve
that game without any violence at all if you understood the crucial item
that you started the game with.)

>One of the interesting things about Underworld is that it was a
>dungeon game, so there was (physical) level 1, level 2, etc.  We
>specifically tried to avoid making the player clear out a level at a
>time, but people were so used to playing that way (from games like
>Wizardry) that they would refuse to go down a level until they had
>cleared the current one out.  There was a monster on level 2 that
>was almost impossible to kill when you first found it, which delayed
>some people from going to level 3 for weeks...

I never felt that way for some reason.  The evil knight was the worst, but
I was happy to leave him behind and keep moving down.  As long as I felt
sufficiently comfortable that I could handle the routine monsters on the
next level (and it seems to me that usually the first experiences you had
were with NPCs and not monsters) I would dive.  It's not that different
>from the normal Ultima series where something interesting was usually
happening at Bucaneer's Den but you couldn't get there until the middle of
the game.

>We were actually very influenced by the Infocom games, and had to dumb
>down a lot of puzzles because they were puzzles, and Warren Spector
>(our producer) pointed out that in an Ultima, the player is always
>told explicitly what to do; they're not supposed to have to infer
>anything themselves.  Grrr.

Interesting.  There was one room that I thought would have been an
interesting puzzle.  I think it was a room where you had to arrange the
switches in a pleasing way so that you could jump to a room that had some
goodies.  Of course, by the time I reached there I had acquired magic boots
that would let me accomplish godlike jumping, so I didn't need to think
about it.

>I played around two thirds of Ultima 7, and remember feeling that it
>was always pretty obvious what I had to do next, but I imagine the
>game was actually designed less linearly than that.

The first phases of 7 were unfortunate, IMHO.  For those of you who are
unfamiliar with the game, you spend the opening half of the game going from
town to town trying to track down the leader of a religious order who is
always one day ahead of you.  But it doesn't work in practice -- you can go
to town A, be told that Batlin is in town B, spend a year (i.e. 365 days
and nights) in the woods honing your fighting skills, go to town B, find
out that he was here until yesterday when he went to town C, return to town
A to find that a murder scene remained essentially untouched even though it
was in the center of the town's only stable, etc.  It would have been more
interesting to have it be real-time, although that's a struggle for the
designer as well.  (Like my arrival in town B could have prompted "Batlin?
Gee, he hasn't been here in AGES, but there sure were some wierd things
that happened last time he came through....")

>Serpent Isle was a monstrosity; even Warren (the producer of that game
>too) agrees.

I thought that the story was a little better, although I was never able to
finish it.

>For what it's worth, my favorite "Ultima" was Savage Empire.  I would
>have been happy to see Origin churn out a few more games for the
>"U6-machine."

Never played that (or Martian Dreams).  I was very pleased when I got
around to the two Underworld games, which I didn't think I would originally
enjoy (given the original descriptions which made it sound more like Rogue
than Ultima).

-Matthew
--
Matthew Daly       I feel that if a person has problems communicating
mwdaly@kodak.com   the very least he can do is to shut up - Tom Lehrer

My opinions are not necessarily those of my employer, of course.

--- Support the anti-Spam amendment!  Join at http://www.cauce.org ---


From taf@isla-mia.demon.co.uk Sun Aug 31 22:36:34 MET DST 1997
Article: 28094 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!dispatch.news.demon.net!demon!isla-mia.demon.co.uk!taf
From: Terence Fergusson <taf@isla-mia.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: Nonlinearity
Date: Sun, 31 Aug 1997 17:32:32 +0100
Organization: University of Murphy
Distribution: world
Message-ID: <Z20ADJAgyZC0IwxA@isla-mia.demon.co.uk>
References: <01bcb083$e874bbe0$539f72ce@hammurabi.blarg.net>
 <340607FC.6327@earthweb.com> <01bcb53c$ff5113a0$939f72ce@hammurabi.blarg.net>
 <5uak9v$isj@nntp5.u.washington.edu> <5ubtlj$kpb$1@bartlet.df.lth.se>
NNTP-Posting-Host: isla-mia.demon.co.uk
X-NNTP-Posting-Host: isla-mia.demon.co.uk [194.222.221.92]
MIME-Version: 1.0
X-Newsreader: Turnpike Version 3.03a <zra8MeDkGOXkPB67pZZTymsN3F>
Lines: 43
Xref: news.lth.se rec.arts.int-fiction:28094

In article <5ubtlj$kpb$1@bartlet.df.lth.se>, Magnus Olsson
<mol@bartlet.df.lth.se> scribed:
>>The reason why I don't want to write these kind of games is that if
>>the player can do anything and all possibilities are played out, then
>>the author can no longer be writing any sort of story; all they're
>>doing is providing a virtual park for the player to wander around in.
>
>I think you're unnecessarily pessimistic. There could still be a story
>going on; it's just that you couldn't force the player to participate.

Hmmm....  This reminds me a lot like Daggerfall.  Nice game.  Huge game
as well.  It wasn't really as sophisticated as the hype said it was, but
if you didn't like the main plot, you could go off and do whatever you
wanted (within the game restrictions).  But the main quest never
continued until you went back to it.

Frontier: First Encounters is another like this.  This time, however,
all of the "story" quests (the ones that were unique, and were not
repeated) went on regardless.  If you didn't take them up, someone else
would, and then the papers would report it.  One brilliant example was
an assassination job you could do.  If you did it, a new quest would
appear on the bulletin boards offering a reward for *your* destruction.
If you didn't take the assassination job, someone else would take it and
complete it, and then you'd be able to hunt them down for the reward.
Or you could ignore it completely and get on with trading, bounty
hunting, pirating, etc, to build up your money reserves.  Maybe go
exploring, or buy a better ship, focus on combat, trading, and so on.

The hard part with text IF is the huge amount of detail that is usually
expected from it.  The current parser and the diversity of the English
language means that an awful lot of programming usually has to be
directed to giving sensible answers to reasonable commands.  And then
interaction of said objects is another factor.  The interface we're
working with is not helpful for creating a huge limitless universe.  But
since that's the interface we're using (and I doubt any other is more
suitable for text IF, and by interface, I'm not talking about any
particular language, but the standard parser system we've been using
since Adventure) is what we're stuck with, that's what we'll have to
make do with.  Either way, it's gonna take a lot of work.

Ciao,
                             Terence Fergusson
                          -- Student of Advanced Murphodynamics


From erkyrath@netcom.com Tue Sep  2 10:27:47 MET DST 1997
Article: 28133 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!howland.erols.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Nonlinearity
Message-ID: <erkyrathEFusr1.80I@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <01bcb083$e874bbe0$539f72ce@hammurabi.blarg.net><5uak9v$isj@nntp5.u.washington.edu> <5ubtlj$kpb$1@bartlet.df.lth.se><Z20ADJAgyZC0IwxA@isla-mia.demon.co.uk> <5ud9mc$694@nntp5.u.washington.edu> <PfTV5AA8mwC0IwgY@isla-mia.demon.co.uk> <01bcb71c$0805dae0$9a9f72ce@hammurabi.blarg.net>
Date: Mon, 1 Sep 1997 23:47:24 GMT
Lines: 20
Sender: erkyrath@netcom18.netcom.com
Xref: news.lth.se rec.arts.int-fiction:28133


Brandon Van Every (vanevery@blarg.net) wrote:

> I'm tired of puzzles that are a "key in the lock" at ANY level of
> abstraction.  Unfortunately I suspect that this is fundamental to the
> defnition of a "puzzle."  A puzzle is, generally speaking, a barricade that
> blocks you from acquiring a new resource.  Or at least that's how I see the
> notion of "puzzle" being used over and over again.  Is there some other
> possible notion of "puzzle?"

A puzzle is something which requires the player to pay attention and then 
act.

(The final cause, rather than the functional definition.)

--Z

-- 

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


From mol@bartlet.df.lth.se Wed Sep  3 18:32:30 MET DST 1997
Article: 28182 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: Disabling "all"
Date: 3 Sep 1997 18:31:42 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 125
Message-ID: <5uk3de$2ol$1@bartlet.df.lth.se>
References: <19970820030201.XAA13411@ladder02.news.aol.com> <fake-mail-2208971956300001@van-52-0328.direct.ca> <5u1j1g$emp$1@bartlet.df.lth.se> <ant2719500b0c4bn@arnod.demon.co.uk>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:28182

In article <ant2719500b0c4bn@arnod.demon.co.uk>,
Julian Arnold  <jools@arnod.demon.co.uk> wrote:
>In article <5u1j1g$emp$1@bartlet.df.lth.se>, Magnus Olsson
><URL:mailto:mol@bartlet.df.lth.se> wrote:

>> In some games, there are NPC's that have vital information which you
>> can only get by asking them about a lot of different things. This is
>> one of the few cases where I really miss "all": it would be nice to be
>> able to say "ask oracle about all" and get a total braindump in
>> return. Of course, it's not very practical to implement.
>
>I wouldn't like this at all.  I wouldn't do this as a player for the
>same reason I don't like "x all".  I wouldn't want to implement this as
>it defeats the purpose of having a questionable NPC (you know what I
>mean) in the first place.  The purpose, or at least a purpose, of such
>an NPC is for gradual and, dare I say it, realistic dissemination of
>information.  If you allow "ask npc about all" you may as well replace
>him with a pamphlet and make the player "read pamphlet."  This is also
>the reason why I wouldn't stick a Nelsonesque book in a game and allow
>the player to "consult book about all" or "look up all in book."

Even though it sounds as if I'm contradicting myself, I actually agree
with you here.

IActually, I was expressing myself a bit badly. I meant that _as a
player_ I sometimes miss the command "as Jill about all", but I didn't
mean that I actually thought it would be a good idea to implement such
a command.

The occasions I've missed "ask about all" is when the NPC knows something
about many of the objects in the game (which is good), and where the author
has sprinkled vital information in these pieces of conversation in a way
that makes it not at all obvious what I should ask about. 

In other words, it happens that you encounter NPC's where you have to ask
them about every object in them game because you feel that their answers
may contain some useful information. IMO this makes the conversation into
a brute-force puzzle, which is a sign of bad game design, not just a case
of a missing "all".

Don't get me wrong: it is very nice when you can ask an NPC about
every object in the game and get adequate responses. However, you
shouldn't be _required_ to do so because there is information hidden
in those responses. There should always be clues about which
conversation subjects are imporant. Such clues could, of course, be
given during the course of the conversation, or they could simply be
implicit in the importance of an object. If you're standing in front
of It's not reasonable to have to ask her about the cat you saw 25
rooms away, unless there's some reason to suspect that she knows
anything important about it.

The same thing applies to encyclopedias, vellum scrolls, and so on:
it's not reasonable to have to consult the encyclopedia about every
single object you encounter (though it's more reasonable than in the
case of the NPC if you can carry the encyclopedia along, which
you normally can'tdo with NPC's).

>Another problem comes to mind. Say the player knows Jack fell down the
>hill and wants to ask Jill about this, "ask jill about jack."  Jill
>reveals that Jack broke his crown during the tumble.  Now the player may
>go on to "ask jill about jack's crown."  But he only has this
>information once he has asked about Jack.  So, if "ask jill about all"
>was allowed the author could handle this in one of two ways:
>  1.
>  Topics are dynamically added to the list of topics encapsulated by
>  "all."  Thus:
>    >ask jill about all
>    [Jill tells about Jack] {Jack's crown is now known about and so
>                             is added to the "all" list}
>    [Jill tells about Jack's crown]
>  
>  2.
>  The list is static; it remains in the state it was in when it was
>  first compiled:
>    >ask jill about all
>    [Jill tells about Jack]
>
>Neither option is desirable, IMO.  #1 is, I guess, better but detracts
>greatly from the point of having an NPC, as I explained above.  #2
>suffers as it misleads the player (as you, Magnus, said in a bit I've
>snipped, I think, the player will think "well I asked about 'all' so
>there can't be anything else to ask about.").  I guess you could
>implement #2 and then tell the player he may have to "ask npc about all"
>multiple times, but this rather defeats the purpose of "ask npc about
>all," doesn't it?

Come to think of it, exactly the same thing applies to "examine all"
in the case where the author wants you gradually to discover things
about a scene. For example:

| You are standing on the porch of a white house. There is a door immediately
| to the north.
| 
| > x door
| 
| The door seems to be nailed shut. There is a sign on it.
| 
| > x sign
| 
| The sign says: "Customers are kindly referred to the side entrance."
| It is fastened with a thumb-tack.
| 
| > x tack
| 
| There's something strange about it - a magical aura, perhaps?

In this case, it's obvious (I hope) that the author wants the player to go
through this sequence of commands - perhaps to ease the info load, perhaps
to achieve an artistic effect. 

Note also that the thumb-tack isn't hidden - it would be unnatural for
it not to be in scope all the time. If it weren't, the player would be
forced to go through all the commands above every time he/she had
restarted the game, which would feel both tedious and unnatural.

In this case, "x all" runs into the same kinds of problems as "ask
Jane about all".




-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From pissedoff@having.too.much.spam Thu Sep  4 09:51:55 MET DST 1997
Article: 28194 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!dispatch.news.demon.net!demon!arnod.demon.co.uk!pissedoff
From: Julian Arnold <pissedoff@having.too.much.spam>
Newsgroups: rec.arts.int-fiction
Subject: Re: Disabling "all"
Date: Wed, 03 Sep 1997 20:44:46 +0100 (BST)
Organization: Stephen Promulgation Society
Message-ID: <ant031946b49c4bn@arnod.demon.co.uk>
References: <19970820030201.XAA13411@ladder02.news.aol.com> <fake-mail-2208971956300001@van-52-0328.direct.ca> <5u1j1g$emp$1@bartlet.df.lth.se> <ant2719500b0c4bn@arnod.demon.co.uk> <5uk3de$2ol$1@bartlet.df.lth.se> 
NNTP-Posting-Host: arnod.demon.co.uk
X-NNTP-Posting-Host: arnod.demon.co.uk [158.152.151.175]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.29]
Lines: 69
Xref: news.lth.se rec.arts.int-fiction:28194

(Spoilers for _The Light: Shelby's Addendum_ below)

In article <5uk3de$2ol$1@bartlet.df.lth.se>, Magnus Olsson
<URL:mailto:mol@bartlet.df.lth.se> wrote:
> [...snip load of stuff that has been agreed...]
>
> Come to think of it, exactly the same thing applies to "examine all"
> in the case where the author wants you gradually to discover things
> about a scene. For example:
> 
> | You are standing on the porch of a white house. There is a door immediately
> | to the north.
> | 
> | > x door
> | 
> | The door seems to be nailed shut. There is a sign on it.
> | 
> | > x sign
> | 
> | The sign says: "Customers are kindly referred to the side entrance."
> | It is fastened with a thumb-tack.
> | 
> | > x tack
> | 
> | There's something strange about it - a magical aura, perhaps?
> 
> In this case, it's obvious (I hope) that the author wants the player to go
> through this sequence of commands - perhaps to ease the info load, perhaps
> to achieve an artistic effect. 
> 
> Note also that the thumb-tack isn't hidden - it would be unnatural for
> it not to be in scope all the time. If it weren't, the player would be
> forced to go through all the commands above every time he/she had
> restarted the game, which would feel both tedious and unnatural.

Yes, this kind of situation is not uncommon.  Progressive examining like
this makes up the first "puzzle" of _The Light: Shelby's Addendum_.  In
this case you examine a fireplace (which, it is revealed, is decorated
with a mosaic of Jonah and the Whale; the whale's eye looks odd) and
then the whale's eye.  This game also supports "x all," and goes for a
static all list.  Thus, if you "x all" this includes the fireplace (so
you know about the eye), but doesn't go on to examine the eye.  Indeed
subsequent "x all"s give the same results as this first one.  ISTR that
people complained about this in an early version of _Shelby_ (not the "x
all" but rather the number of examine commands required).  In that
version the goal of your examining (the eye itself) was "deeper," so you
had to examine, say, four things to reach it.  This implies there is a
(rather low) limit to the depth of progressive examining which players
will tolerate--probably 2 or 3 levels.  I know I get ratty at more.

I think there is an identifiable difference between "x all" and "ask npc
about all."  In the former case the members of the "all" list are in
scope--it is literally a short-cut for lazy players who don't read room
descriptions. :) In the latter case the members are quite likely not in
scope--if the askable topics are limited to game objects then those
members could be any game object, whether it is, or even has been, in
scope, and if the askable topics are not limited to actual objects (ie,
as with Inform) the members could be anything at all.  Thus "ask npc
about all" is more than a short-cut--IMO it's a cheat.  (Note: there are
certainly circumstances where "x all" becomes as cheaty as "ask npc
about all.")

Jools
-- 
"For small erections may be finished by their first architects; grand
ones, true ones, ever leave the copestone to posterity.  God keep me
>from ever completing anything." -- Herman Melville, "Moby Dick"
             [ Please reply to jools@arnod.demon.co.uk ]



From gkw@pobox.com Thu Sep  4 09:54:00 MET DST 1997
Article: 28206 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!cpk-news-hub1.bbnplanet.com!cam-news-hub1.bbnplanet.com!news.bbnplanet.com!prodigy.com!nntp.earthlink.net!usenet
From: gkw@pobox.com (Gerry Kevin Wilson)
Newsgroups: rec.arts.int-fiction
Subject: [TADS] ask X about Y.
Date: 4 Sep 1997 04:58:50 GMT
Organization: None
Lines: 44
Distribution: world
Message-ID: <5ulf6a$8j9@chile.earthlink.net>
NNTP-Posting-Host: 207.217.141.100
X-Newsreader: News for Windows NT X1.0-93
Xref: news.lth.se rec.arts.int-fiction:28206

Well, it's taken 4 years to notice it, but I see that the 2.0 manual and the
current ADV.T file handle >ASK X about Y in two different ways, and the
ADV.T method is infinitely superior.

By the manual, we have:

verDoAskAbout(a, i) = {}
doAskAbout(a, i) =
{
  switch(obj)
  {
     case FredObject:
       "\"Fred's a great guy.\"\n";
      break;
     [etc]
  }
}

Looking in the ADV.T file, I note that the method has switched from
using objects to examining words (a far superior method in that it
avoids some nasty disambiguation problems with questions.)  It works
like:

askWord(word, lst) =
{
   switch(word)
   {
     case 'fred':
     case 'flintstone':
       "\"Fred's a great guy.\"\n";
      break;
     [etc]
   }
}

Granted, it's a little more work, but again, if you have three
different Freds in your  game, this can really help you out.

This is of course, probably obvious to most TADS users already,
but I figure if I can overlook it for years, then others can too.


G. Kevin Wilson: Freelance Writer and Game Designer.  Resumes on demand.



From graham@gnelson.demon.co.uk Thu Sep  4 14:10:33 MET DST 1997
Article: 28212 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!dispatch.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Inform library 6/6: DarkToDark bug
Date: Thu, 04 Sep 1997 09:53:09 +0100 (BST)
Organization: none
Message-ID: <ant0408090b0M+4%@gnelson.demon.co.uk>
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 45
Xref: news.lth.se rec.arts.int-fiction:28212


With the competition deadline approaching, I felt I should post
a fix for a newly-found bug in Library 6/6 right away.  The bug
is simply that I inadvertently botched a previous repair in such
a way that the routine DarkToDark, if you provide one (for when
the player moves from one dark place to another), is never called.

This means that any rules about moving around in the dark will not
be enforced, a problem you might not even notice when testing your
game, but which could do it some harm.

The bug is easy to fix, as follows: amend "GoSub" in "verblibm.h",
firstly to replace

[ GoSub i j k df movewith thedir;

with

[ GoSub i j k df movewith thedir old_loc;
  old_loc = location;

and secondly (on line 1163) replace

  if (df~=0) { location=j; lightflag=1; }
  else
  {   if (location == thedark)
      {   DarkToDark();
          if (deadflag~=0) rtrue;

with

  if (df~=0) { location=j; lightflag=1; }
  else
  {   if (old_loc == thedark)                    << change here
      {   DarkToDark();
          if (deadflag~=0) rtrue;

I apologise for the difficulty and for this unorthodox posting, but
I thought the bug was serious and stupid enough to be worth mention
now.  Library 6/7 will follow in due course, with this and other
(but only much more minor) bug fixes in.

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From pissedoff@having.too.much.spam Thu Sep  4 21:01:56 MET DST 1997
Article: 28220 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news.algonet.se!news.maxwell.syr.edu!news-peer.sprintlink.net!news.sprintlink.net!Sprint!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!dispatch.news.demon.net!demon!arnod.demon.co.uk!pissedoff
From: Julian Arnold <pissedoff@having.too.much.spam>
Newsgroups: rec.arts.int-fiction
Subject: Re: Summary of Nonlinearity
Date: Thu, 04 Sep 1997 17:53:59 +0100 (BST)
Organization: Stephen Promulgation Society
Message-ID: <ant041659b49c4bn@arnod.demon.co.uk>
References: <01bcb548$f6c96dc0$939f72ce@hammurabi.blarg.net> <19970830215100.RAA11608@ladder01.news.aol.com> <01bcb71a$2b5a4460$9a9f72ce@hammurabi.blarg.net> 
NNTP-Posting-Host: arnod.demon.co.uk
X-NNTP-Posting-Host: arnod.demon.co.uk [158.152.151.175]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.29]
Lines: 87
Xref: news.lth.se rec.arts.int-fiction:28220

In article <01bcb71a$2b5a4460$9a9f72ce@hammurabi.blarg.net>, Brandon Van Every
<URL:mailto:vanevery@blarg.net> wrote:
> [...]  A "nonlinear" game, according to my own definitions:
> 
> (1) has no single endpoint.  Nor does it have multiple endpoints.  In fact,
> it has no endpoints whatsoever.  Expressing this mathematically: the game
> is a general graph with loops in it.  From any node in the graph, you can
> get back to any other node in the graph somehow.
> 
> (2) has no singular goal.  The "goal" of the game is to arrive at a game
> state "that you like."  If you like it, you stay there for awhile.  If you
> don't like it, or you get bored and restless with it, you move on to some
> other game state.  In this respect the game has something in common with
> the "software toys" of Maxis, i.e. SimCity, SimAnt, etc.  The difference is
> that you might not care about building the "best" city or some such.  You
> might want to build the worst city, a city that's better in a different
> way, a mediocre city, or not build cities at all.  It depends on the
> player's current goal, their own psychological satisfactions or
> dissatisfactions with the current game.
> 
> (3) could have a single beginning.  However, it could also have multiple
> beginnings.  For example you could start in a random location, or pick from
> a list of available characters/personas.  By design, where you start could
> influence your perception of the game, but the game doesn't have to be
> designed this way to be nonlinear.
> 
> (4) allows freedom of action and freedom of intention.  The game should
> refrain from herding players from room to room.  Nor should it be necessary
> to have a specific objective or "mindset" in order to move around in the
> game.  Let's say the game, broadly speaking, was about a quest for the Holy
> Grail.  If I wanted to play very seriously and actually find the Grail,
> that should be possible.  But if I wanted to blow off the Grail and ham it
> up ala Monty Python, that should be possible as well.  Or, maybe I want to
> experience the quest as Salvador Dali might have, with the Grail as a
> psychological symbol rather than the tangible product of a series of puzzle
> solutions.  Put another way, the "authorial intent" of the game should be
> nonrestrictive.  In a nonlinear game, it is not the author's job to shout
> the narrative at the player.  It is the author's job to construct an
> environment in which the player will complete his/her own self-narration.

I haven't been following this thread, but this summary seems a good
opportunity to jump in.  Apologies if I think what someone else has
thunk afore me.  Here are some disjointed thoughts:

Firstly, it sounds to me as if you're advocating "simulationist" IF over
"puzzle-/story-based" IF, and not really talking about (non)linearity. 
I did read the first few posts in this thread, where people questioned
your definition of (non)linearity actually, and as my own understanding
of this concept seems to differ slightly from other people's (see the
Linearity v Interactivity thread from a while back) I won't go on about
this.

Secondly, I think you want such an extreme "simulationist" IF that it is
entirely open-ended and, yes, nonlinear.  Is your hypothetical Holy
Grail game from (4) an example of one of your nonlinear games?  If so,
how is (4) reconciled with (1)?  Surely finding the Holy Grail (or using
it in some way, or whatever) would constitute an endpoint?

Thirdly, ISTM that part of the enjoyment of playing IF is that you are
working within certain limitations--those naturally imposed by having a
plot, or a story, at all.  Certainly it's undesirable that these
limitations should become restrictions, and this is something that
authors have to look out for.  In (1), (2), and (3) you seem to be
saying that to you nonlinear IF has no limitations whatsoever, that the
game is (or at least can be) completely unstructured.  In (4) however,
your outline of a Holy Grail game has a certain implicit
structure--namely that the goal of the game is to find the Grail.

If your Grail game was implemented, so that everything in (4) was
possible, wouldn't you just complain that it was too restrictive still,
as you couldn't build cities a la SimCity?

Finally, to summarize your summary, you seem to be calling for a game
where you can do anything at any point, a game that is in fact less
limited than Real Life (Star Trek's holodeck, right?).  Is this right? 
I don't think such a game will be possible (even text-based) for a very
long time, if ever.

Jools

PS--this isn't an attempt to flame you.
-- 
"For small erections may be finished by their first architects; grand
ones, true ones, ever leave the copestone to posterity.  God keep me
>from ever completing anything." -- Herman Melville, "Moby Dick"
             [ Please reply to jools@arnod.demon.co.uk ]



From mol@bartlet.df.lth.se Fri Sep  5 18:11:07 MET DST 1997
Article: 28237 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: What is a language (was: I need info...)
Date: 5 Sep 1997 18:10:28 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 66
Message-ID: <5upatk$55m$1@bartlet.df.lth.se>
References: <340D4640.59D0@injersey.com> <5uk5as$77d$1@bartlet.df.lth.se> <5uo46f$d2j$1@news01.micron.net>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:28237

In article <5uo46f$d2j$1@news01.micron.net>,
Jay Goemmer  <radioguy@not.my.address> wrote:
>
>mol@bartlet.df.lth.se (Magnus Olsson) wrote:
>
>> In article <340D4640.59D0@injersey.com>, Ray <sumby@injersey.com> wrote:
>>> I was just wondering if someone could refer me to an simple and easy 
>>> program to use to write a text adventure game (and/or an internet 
>>> site where i could get info and download it) 
>>
>> There are several. Inform, TADS and Hugo are very powerful but perhaps 
>> not very "simple and easy". AGT is said to be easier, as long as you 
>> don't try to do more advanced things. 
>
>   I differentiate AGT and Alan from "The Big Three," preferring to refer  
>to AGT and Alan as IF "implementation systems," rather than actual IF 
>"languages."  

I can see the need to divide the IF systems into different groups, but
I really don't think your choice of terminology is a very good one,
fora bunch of reasons:

1) Surely the Big Three (Inform, TADS and Hugo) are "IF Implementation
Systems" as well? Come to think of it, I can see a difference between
the TADS System (which includes the run-time, the debugger, the compiler
and the library) as opposed to the TADS Language which is the language in
which TADS programs are written.

2) By any reasonable definition, both AGT and Alan are languages. Both
systems are based on textual source code; a formal description of
the game that has a well-defined syntax. This description is translated
by a computer program into a form that the run-time system can understand.

You may argue that AGT and Alan aren't Turing-complete (though I'm not
sure if they are or not), but that's not a requirement for anything
being called a language. 

3) The division is arbitrary and obfuscates the fact that AGT and Alan
really has much more in common with the big three than there are
differences.


>AGT and its ilk are easier for the beginner, who might be 
>rather intimidated by several screenfuls of Inform error messages, for 
>example.  (I'm determined to beat them, though!)  

So you mean you can't get screenfuls of error messages from AGT?
(I don't know, I've never tried it.) That's good!

>> Alan is much more modern than AGT and probably the easiest of the 
>> bunch.
>
>   I couldn't say.  I never did finish printing out the manual's .PDF  
>file, but the sample game "Jungle" enclosed with Alan seemed very stiff  
>and unplayable, most likely because it was merely a very rough  
>programming example, rather than a full-fledged game segment.  

I'd rather leave that for ThoNi to answer...




-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From jaysmith@removethis.kih.net Fri Sep  5 22:06:44 MET DST 1997
Article: 28244 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!newsfeed1-hme1!newsfeed.internetmci.com!198.60.253.107!skin02.micron.net.!news-sea-20.sprintlink.net!news-sea-19.sprintlink.net!news-in-west.sprintlink.net!news.sprintlink.net!Sprint!204.214.208.3!news.dbtech.net!sco21.kih.net!jaysmith
From: jaysmith@removethis.kih.net (Jayson Smith)
Newsgroups: rec.arts.int-fiction
Subject: The largest inventory listing ever?
Date: Fri, 5 Sep 1997 13:45:34
Organization: Nobody
Lines: 42
Message-ID: <jaysmith.2.000DC2E1@removethis.kih.net>
NNTP-Posting-Host: 207.140.101.86
Xref: news.lth.se rec.arts.int-fiction:28244

Hey you guys, check out this inventory listing which comes from
"The Museum of Inform", when I had gathered up all of the mobile
objects I could find and put them all in my samples bag.  Could
this possibly be the longest inventory listing ever created?  Or
maybe the highest number of objects ever placed inside one
container?  Check it out and see!

Start of a transcript of
MUSEUM OF INFORM
An Interactive Companion to the Designer's Manual
Release 1 / Serial number 951220 / Inform v6.13 Library 6/6 D
Standard interpreter 0.2
Interpreter 6 Version E / Library serial number 970818

>i
You are carrying your samples bag (which is open), inside which are
a moth, a feather, a lead ingot, a five-zob weight, a ten-zob
weight, a twenty-zob weight, a fifty-zob weight, a hundred-zob
weight, a Louis XV chair, a black Tyndale Bible, a gold watch, a
sprig of lavender, a green ball, a red cone, a blue pyramid, a
plain shopping bag (which can only hold 2 things), a glass box with
a lid (which is open but empty), a steel box with a lid (which is
open but empty), a bolted key, a toothed bag, a portable television
set, a tricorder, an alarm clock, a genie's lamp, a white stone, a
black stone, a fly in amber, a magnifying glass, a green pepper, a
red fried tomato, that lethal old book of Geoffrey's, that hateful
box (which is open), inside which is a matchbook (5
matches left), a platinum pyramid, a plastic fork, knife and spoon,
three hats ( a fez, a Panama and a sombrero), the letters X, Y, Z,
P, Q and R from a Scrabble set, a defrosting Black Forest gateau,
Punch magazine, a recent issue of the Spectator, a die, eight stars
(four silver, one bronze and three gold), a Chinese scroll, the
gold coins goat, deer and chicken and the silver coins robin, snake
and bison.

>script off

End of transcript.

Hehehe!  Hope you liked that!
To reply to me directly, simply remove the "removethis." from my  
e-mail address.  Thanks.


From mwdaly@kodak.com Sun Sep  7 09:50:40 MET DST 1997
Article: 28265 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!news-fra1.dfn.de!news-was.dfn.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-peer.sprintlink.net!news-pull.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!159.56.2.2!thetimes.pixel.kodak.com!news.kodak.com!newsserver.rdcs.Kodak.COM!not-for-mail
From: mwdaly@kodak.com (Matthew Daly)
Newsgroups: rec.arts.int-fiction
Subject: Re: Has anyone truly been *offended* by IF?
Date: Sun, 07 Sep 1997 00:22:04 GMT
Organization: Eastman Kodak Company
Lines: 43
Message-ID: <3412edff.266452156@newsserver.rdcs.kodak.com>
References: <owls-0609971115370001@owls.vip.best.com>
Reply-To: mwdaly@kodak.com (Matthew Daly)
NNTP-Posting-Host: ep-psg-arrakis.ycc.kodak.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Newsreader: Forte Agent 1.01/32.397
Xref: news.lth.se rec.arts.int-fiction:28265

owls@best.com (JID), if that is your REAL name, said: 

>For myself, it hasn't happened yet. There have been points in games when
>I've really not liked what I've had to do (say, in Trinity), but because
>it was part of the story I went ahead with it, but I wouldn't call it
>"offended."

Actually, when I read your first paragraph, Trinity was exactly what was on
my mind.  And I feel justified in taking offense, for the following
reasons:

A) The author went a long way out of his way to create a situation where I
had to do what I did.  That is, the eventual result that came from the
action had nothing to do with the action.

B) I read in this newsgroup at one time that the author deliberately wrote
this action in because he wanted everyone who finished the game to know
that they had deliberately performed a grossly immoral act in the name of
"going on with the game".

C) There was little foreshadowing in the game, or in it's packaging, that
one would be forced to leave one's scruples behind to get sufficient value
>from the game.

For this reason, I can feel slightly better about the questionable act in
Stationfall, which doesn't suffer from problem (A) and you know that you've
seen the whole game except the endgame so you don't miss much if you don't
type the words, as I chose not to.  So my Stationfall galaxy, instead of
being destroyed, went into stasis. :-)

One of the first scenes in Ultima VIII involves you (as the Avatar)
watching an unjust execution taking place, where you cannot intervene or
turn away.  It was a brutal and tactless way of indicating "you're not in
Britannia any more."

-Matthew
--
Matthew Daly       I feel that if a person has problems communicating
mwdaly@kodak.com   the very least he can do is to shut up - Tom Lehrer

My opinions are not necessarily those of my employer, of course.

--- Support the anti-Spam amendment!  Join at http://www.cauce.org ---


From jdblask@csd.uwm.edu Sun Sep  7 09:51:11 MET DST 1997
Article: 28268 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!cs.utexas.edu!uwm.edu!alpha3.csd.uwm.edu!jdblask
From: Wonder Boy <jdblask@csd.uwm.edu>
Newsgroups: rec.games.int-fiction,rec.arts.int-fiction
Subject: Re: Has anyone truly been *offended* by IF?
Date: Sat, 6 Sep 1997 23:35:06 -0500
Organization: University of Wisconsin - Milwaukee
Lines: 10
Message-ID: <Pine.OSF.3.96.970906233013.30167A-100000@alpha3.csd.uwm.edu>
References: <owls-0609971115370001@owls.vip.best.com>
NNTP-Posting-Host: 129.89.169.203
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
In-Reply-To: <owls-0609971115370001@owls.vip.best.com>
Xref: news.lth.se rec.games.int-fiction:25714 rec.arts.int-fiction:28268

	I found the "the Broken String" to be quite offensive.  Although
the text file that was in the zip states otherwise, I found the game to be
way too preachy about what the author felt was "punk".
	Also, I've found the banality of some of the low-brow, first-
attempt-text games somewhat offensive...
	But that's just me.
							-jon  





From radioguy@not.my.address Sun Sep  7 09:51:20 MET DST 1997
Article: 28269 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!newsfeed1-hme1!newsfeed.internetmci.com!198.60.253.107!skin02.micron.net.!news01.micron.net!not-for-mail
From: Jay Goemmer <radioguy@not.my.address>
Newsgroups: rec.games.int-fiction,rec.arts.int-fiction
Subject: Re: Has anyone truly been *offended* by IF?
Date: 7 Sep 1997 05:26:19 GMT
Organization: Societe pour je ne sais quois
Lines: 17
Message-ID: <5utdtr$ntp$2@news01.micron.net>
References: <owls-0609971115370001@owls.vip.best.com> <Pine.OSF.3.96.970906233013.30167A-100000@alpha3.csd.uwm.edu>
NNTP-Posting-Host: ctwf012p06.twf.micron.net
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 1.2N (Windows; I; 16bit)
Xref: news.lth.se rec.games.int-fiction:25715 rec.arts.int-fiction:28269

Wonder Boy <jdblask@csd.uwm.edu> wrote:

> I've found the banality of some of the low-brow, first-attempt-text > games somewhat offensive...

   I'll say the same thing about IF that I do about static fiction:

     "You *know* it's bad when you want to *edit* while you're 
*reading.*"


   Of course, this covers misspelling, bad grammar, poor punctuation, 
[insert your pet peeve here] . . .



--Jay Goemmer



From graham@gnelson.demon.co.uk Mon Sep  8 16:14:43 MET DST 1997
Article: 28315 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!news.grnet.gr!news-feed1.eu.concert.net!news-peer.bt.net!btnet!baron.netcom.net.uk!netcom.net.uk!dispatch.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: IF Authoring Systems
Date: Mon, 08 Sep 1997 12:00:27 +0100 (BST)
Organization: none
Message-ID: <ant081127345M+4%@gnelson.demon.co.uk>
References: <34139C3E.7609@juno.com> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 126
Xref: news.lth.se rec.arts.int-fiction:28315

In article <34139C3E.7609@juno.com>, Jeff Hatch
<URL:mailto:jeffhatch@juno.com> wrote:
> 
> What would a great interactive fiction authoring system be like?

Modesty forbids...

> I don't mean to rehash the "Tads vs. Inform vs. new systems" debate. 
> I've seen snippets of code from each of the most popular IF languages,
> and I know that Inform and Tads are both *good* languages.  No, I'm
> thinking of something else entirely.
> 
> Perhaps you'd rather create new works of art than improve your set of
> tools.  That's understandable, but in the long run a better IF authoring
> system will greatly benefit the genre.  Ancient Egyptians carved
> hieroglyphs in stone with chisels, but 19th-Century American authors had
> access to paper, ink, and quill pens.  It's no wonder fewer books were
> written in ancient Egypt!  Similarly, time-saving new tools for IF
> writers can only increase creative output.

You aren't the first to think so, and perhaps you're right.  It's
not quite so clearcut, though, and the history of IF authoring
systems has seen compilers like TADS triumph over simplistic
point-and-click database front-ends of the kind you propose.  The
comparison with chisels versus quill pens is not valid because it
implicitly assumes that the range of text writable with each
implement is the same.  This is not true: a cute, easy-to-get-started
adventure creation system (of which there are very many already)
is usually crippled by its inability to handle the full flexibility
needed by designers of even the most modest games.  When you have
that full flexibility, you pretty much have something as powerful
and necessarily complex as TADS.

Incidentally, books were written in ancient Egypt, and perhaps one
might point out that you could write the above backwards to lament
the decline of monumental inscriptions in 19th-century America
due to a culture of ephemerality based on wood-pulp...

> A great interactive fiction authoring system should be more than a
> compiler.  Even the "easy-to-use" IF systems obey the same paradigm:
> Edit text file. Recompile. Execute. Locate bug. Repeat.  Debuggers are
> perhaps the only modern innovation.  Why are we writing interactive
> fiction using the linear programming methods of the '60s?  Yes,
> algorithms have to be linear, but interactive fiction is more
> "object-oriented" than algorithm-oriented.  We make Rooms, Doors,
> Actors, Fuses, Items.  There should be a simple, fast way to create
> these things!

To some extent there is.  The present system is not so bad as it
sounds: modularity into objects makes it pretty easy to cut out,
e.g., the lantern from "Advent" and glue it directly into a
different game, where it will continue to work well.

Ask also: why are we writing procedural code using the programming
methods of the 60s?  (We are.)  Possibly because the basic techniques
of programming were perfected in the 60s, leaving out preferences in
the nuance of language design.

> A great interactive fiction authoring system should be a cross between a
> database and a compiler.  It should present the attributes of
> user-created "objects" in a table so they're easy to edit.

TADS, Inform, etc., all do this.  You seem to be proposing no more
than a pretty windowed text editor to be glued onto the side of
an existing system.  This has been done, too, but usually people have
found that it doesn't save as much time as the bother it causes.

The trouble is that a fancy text editor may indeed save a little time
on the routine jobs, such as typing in map connections, but these
jobs represent only a tiny proportion of the coding time involved in
getting a game up to releasable quality.  And the best authorship
systems are targetted toward producing quality, finished work,
not toward making it easy to set up primitive demos of adventure
games.

> Practical Example: Suppose your game has a Kitchen next to a Landing,
> and you want to put a door between them. Your library doesn't have a
> specific Door class, but it has "Obstacle" and "Openable" classes, which
> suffice. You define a "Door" object that's derived from the Openable and
> Obstacle classes.  A menu appears, and a portion of your screen shows
> this:
> 
> 		Obj  Door		COpenable CObstacle
> 		 Cls  COpenable		 CLocal
> 		  Var  Text
> 		  Var  Location		  0
> 		  Fnc  Commands
> 		  Var  IsOpen		  0
> 		 Cls  CObstacle
> 		  Var  Location2          0
> 
> In a matter of seconds, you move the cursor to the right-hand column and
> set Text = "door," Location = "Kitchen," and Location2 = "Landing." 
> Then you go set the Kitchen's West field and the Landing's East field to
> "Door."

As I said, this is no more than a text editor optimised to a language
like TADS or Inform.  Such text editors have not caught on, partly
because they're difficult to make universal across a broad range of
computer systems (whereas tools such as compilers are relatively easy
to spread).

> That's all--no compiling necessary!

You plan to interpret rather than compile?  This will have a significant
impact on the running speed of the final game.

> An experienced programmer
> would be able to do that in thirty seconds, and a beginner would find
> the database format much easier to understand than any compiler's text
> source file.

Are you really arguing that the textual example you give above is
easier to follow for beginners than the equivalent code in TADS?

> A great interactive fiction authoring system should be available in six
> to twelve months.

TADS and Inform each took about three years to reach their present
levels of maturity, and represent a very substantial wodge of
practical experience and contributions by many hands.  It just isn't
a trivial, or even a six-month problem.

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From graham@gnelson.demon.co.uk Mon Sep  8 16:15:34 MET DST 1997
Article: 28316 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!dispatch.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Models for handling NPCs
Date: Mon, 08 Sep 1997 11:44:03 +0100 (BST)
Organization: none
Message-ID: <ant081003d07M+4%@gnelson.demon.co.uk>
References: <01bcba19$41219580$aa9f72ce@hammurabi.blarg.net> <5upie2$9ds@nntp5.u.washington.edu> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 54
Xref: news.lth.se rec.arts.int-fiction:28316

In article <5upie2$9ds@nntp5.u.washington.edu>, Mary K. Kuhner
<URL:mailto:mkkuhner@phylo.genetics.washington.edu> wrote:
> 
> Inform (I can't comment on the others) has more support for rooms and
> objects than it does for people, or at least it feels that way.

I think this is a fair criticism, yes, but in mitigation I might
say that the Inform library uses a world model which is based on
an accepted idea of the state of the art -- that is, its basic
concepts are those which are understood as "basic" in the corpus
of existing adventure games of the 70s and 80s: containment,
supportability, enterability, vehicularity (oh dear: is there such
a word?), openability, etc.  There was and is much less common
ground on how NPCs "should" behave.  So crude is the standard
idea about NPCs that in Inform the same basic model handles animals
and people more or less equally.  That basic model -- which
appreciates that some actions make more sense in the case of
people, such as ordering about, kissing, giving or waking up --
is essentially a codification of how Infocom's games behaved.

It's difficult to decide how a next-level model would work.  The
various ideas about giving NPCs some kind of pseudo-random autonomy
(or even rudimentary abilities to plan ahead) are dangerous in an
IF authorship system such as Inform, because they need to be very
tightly under the designer's control, and because they would need
to "understand" the rest of the design in a quite sophisticated
way.  It's not easy to write a foolproof algorithm even for moving
a Zork I-style thief around a relatively uncluttered map, for
instance (see the Inform manual for example code).

An alternative which some people believe in, and which I think may
indeed have some merit, is to reassign one's way of thinking so
that the basic unit is not the actor but the script.  That is, to
provide facilities for running through sequences of events: we've
all played games where an NPC appears to exist in order to tell
us something, one paragraph per turn, and then to go away again.
If we interrupt the story, we're hissed at to keep quiet, or told
that the teller is inexplicably deaf, too caught up in his own
story, etc., etc.  Scripting can be made to be more sophisticated
than this, but whenever I've considered moving such facilities into
the Inform library, I've concluded that it's better to let
designers work such things out for themselves.  I may yet provide
an optional-extra library file, perhaps.

What should an NPC do?  It's not an easy question.  I think the
best NPC-led scene in the IF canon may be the prologue to
"Christminster", and this works well because the various NPCs are
interestingly different and interact with each other, not just with
the player.  It's the designer that makes all this happen, though.
A souped-up library wouldn't do it for you.

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From pissedoff@having.too.much.spam Mon Sep  8 21:33:09 MET DST 1997
Article: 28334 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!dispatch.news.demon.net!demon!arnod.demon.co.uk!pissedoff
From: Julian Arnold <pissedoff@having.too.much.spam>
Newsgroups: rec.arts.int-fiction
Subject: Re: Summary of Nonlinearity
Date: Mon, 08 Sep 1997 15:17:26 +0100 (BST)
Organization: Stephen Promulgation Society
Message-ID: <ant0814260b0c4bn@arnod.demon.co.uk>
References: <01bcb548$f6c96dc0$939f72ce@hammurabi.blarg.net> <19970830215100.RAA11608@ladder01.news.aol.com> <01bcb71a$2b5a4460$9a9f72ce@hammurabi.blarg.net> <ant041659b49c4bn@arnod.demon.co.uk> <01bcba18$c128f6c0$aa9f72ce@hammurabi.blarg.net> 
NNTP-Posting-Host: arnod.demon.co.uk
X-NNTP-Posting-Host: arnod.demon.co.uk [158.152.151.175]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=ISO-8859-1
Content-Transfer-Encoding: 8BIT
X-Newsreader: ANT RISCOS Marcel [ver 1.29]
Lines: 109
Xref: news.lth.se rec.arts.int-fiction:28334

In article <01bcba18$c128f6c0$aa9f72ce@hammurabi.blarg.net>, Brandon Van Every
<URL:mailto:vanevery@blarg.net> wrote:
> Julian Arnold <pissedoff@having.too.much.spam> wrote in article
> <ant041659b49c4bn@arnod.demon.co.uk>...
> > In article <01bcb71a$2b5a4460$9a9f72ce@hammurabi.blarg.net>, Brandon Van
> Every
> > <URL:mailto:vanevery@blarg.net> wrote:
> > 
> > Firstly, it sounds to me as if you're advocating "simulationist" IF over
> > "puzzle-/story-based" IF, and not really talking about (non)linearity. 
> > I did read the first few posts in this thread, where people questioned
> > your definition of (non)linearity actually, and as my own understanding
> > of this concept seems to differ slightly from other people's (see the
> > Linearity v Interactivity thread from a while back) I won't go on about
> > this.
> 
> I'll crank up DejaNews, but if in any event you'd be interested in going on
> about how you'd define "simulationist" IF, I'd be interested.

Essentially I would define "simulationist" IF in the same way as you
define "nonlinear" IF.  The main difference between this (theoretical)
IF and the (actual) IF which we have seen so far is that in "sim" IF the
story takes a back seat to the simulated environment and in fact their
need not be a coherent story at all (people have argued that the very
interactivity of IF precludes there being a story at all, but I say that
the dictionary definition of story is inadequate, though reasonably so
given the relative youngness and obscurity of IF).  This difference has
the knock on effect that in "sim" IF the illusion of freedom of choice
is and must be far less illusory.  Which is one reason why no-one does
or can write truly "simulationist," or "nonlinear" if you must, IF.

As you say in another message though, all these definitions are loose
and open to argument.  And yes I think that such argument has a useful
purpose.  OTOH you're use of the word "nonlinear" is a bit like deciding
that you personally will refer to fist-sized misshapen root vegetables
as bananas, despite what anyone else understands by the word.  It's
confusing and actually diverts attention away from the real discussion.

Of course, now everyone will claim that my vague definition of
"simulationist" IF is rubbish...

> > Secondly, I think you want such an extreme "simulationist" IF that it is
> > entirely open-ended and, yes, nonlinear.  Is your hypothetical Holy
> > Grail game from (4) an example of one of your nonlinear games?  If so,
> > how is (4) reconciled with (1)?  Surely finding the Holy Grail (or using
> > it in some way, or whatever) would constitute an endpoint?
> 
> Well, let's say you get the Grail.  And then you get a message that says
> "You've got the Grail."  Now what?  I think there's the possibility of
> anticlimax, as in... the wanting was better than the having.

This could be a fantastic ending to a Grail quest.  Wanting is better
than having... that's the way life works.  Frustrating, isn't it?

> Now that
> you've got the Grail, maybe you do something important like take it back to
> room #26 and pour yourself a beer in it.  Maybe the act of possessing the
> Grail could be trivialized even further by messages like "Your beer tastes
> really good, even better than if you didn't have a Grail!"

At this point I'd lose interest in the game.  I think having such a
loose/trivial/irrelevant/poor "ending" (or rather, no ending at all)
would only serve to detract from the story itself, and leave a bad taste
in theplayer's mouth (not that I generally ingest, or even taste, IF
games).

> > In (4) however,
> > your outline of a Holy Grail game has a certain implicit
> > structure--namely that the goal of the game is to find the Grail.
> 
> That's "a possible" goal of the game.  It is not "the" goal of the game. 
> It is important to have other things for people to do, so that the players
> can make choices about what they actually care about from moment to moment.

Agreed.  I just think you're taking this to such an extreme that cannot
be satisfied.

> > Finally, to summarize your summary, you seem to be calling for a game
> > where you can do anything at any point, a game that is in fact less
> > limited than Real Life (Star Trek's holodeck, right?).  Is this right? 
> > I don't think such a game will be possible (even text-based) for a very
> > long time, if ever.
> 
> I don't think so.  "Freedom of action and intention" is something of an
> approximation.  I think keeping this design goal firmly in mind is better
> than herding people towards the game's conclusion, yelling "You can't go
> that way" at every turn.  To be sure, this requires labor, and most of that
> labor must be done manually.

I refer the honourable gentleman to the answer I gave some moments ago.

> But still there are 2 questions: (1) can some
> of the labor be done automatically via a good theory of recombination, and
> (2) how can our tools amplify/exponentiate our ability to produce stuff
> manually?

Are you familiar with Jorn Barger's theorizing?  I think maybe you two
would agree about a lot.

I do agree that those two questions are important, but just how
important at the present time, I wouldn't like to say.

Jools
-- 
"For small erections may be finished by their first architects; grand
ones, true ones, ever leave the copestone to posterity.  God keep me
>from ever completing anything." -- Herman Melville, "Moby Dick"
             [ Please reply to jools@arnod.demon.co.uk ]



From s590501@tfh-berlin.de Mon Sep  8 21:33:44 MET DST 1997
Article: 28322 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-stkh.gip.net!news.gsl.net!gip.net!gsl-penn-ns.gsl.net!news-stock.gsl.net!news.gsl.net!gip.net!newsfeed.kornet.nm.kr!news.maxwell.syr.edu!newsfeed.nacamar.de!blackbush.xlink.net!fu-berlin.de!dip8085.tfh-berlin.DE!miron
From: "Miron Schmidt" <s590501@tfh-berlin.de>
Newsgroups: rec.arts.int-fiction
Subject: Re: I'm too stupid for I-0
Date: 8 Sep 97 09:54:36 +0200
Organization: TFH Berlin via PPP
Lines: 57
Message-ID: <3413CB5B.MD-0.198.s590501@tfh-berlin.de>
References: <01bcbb83$6e1dfc20$829f72ce@hammurabi.blarg.net>
NNTP-Posting-Host: dip8085.tfh-berlin.de (141.64.108.85)
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
NNTP-Posting-User: miron
X-Access: 16 1054 1055
X-Newsreader: MicroDot-II/AmigaOS 0.198
Xref: news.lth.se rec.arts.int-fiction:28322

Brandon Van Every <vanevery@blarg.net> wrote:
> Ok I've played I-0 for the 5th time now.  Since so much has been said about
> I-0 in reference to "nonlinearity," I felt that I needed to really dissect
> the thing.

I think you're making the mistake of regarding I-0 as the closest game yet to
fulfil your "nonlinearity" theory and dissecting it in regard to that.
I-0 is straying from the traditionally linear path and succeeding pretty well
in this. It does not *attempt* to truly snapshot a day in Tracy's life, with
clever answers to every possible action.

> I bring up these issues because they speak to the problems of nonlinearity.
>  In a linear story, if you can't guess the author's intent about a puzzle,
> or if you can't relate to the author's perspective on human behavior, you
> simply hit a brick wall.
> [...]
> Well, what if the player
> thought they were pretty thorough the 1st time around?  The player goes
> back, tries again, and doesn't even know where his blind spots are.  In a
> linear game, at least you know that the next game area is beyond the next
> locked-door-in-front-of-you, so you know what to work against.
> 
> I am wondering what solutions, if any, exist for this problem.
> [...]
> I see two possible solutions for this problem, which really are the same
> solution when viewed at a more abstract level.  One is to make a game with
> multiple characters, and let the player pick which character they're going
> to play with.
> [...]
> The other is
> to stick with a single character, but implement it as though it had
> multiple perspectives.

A third, obvious solution would be to accept the author's point of view and
play _his_ character in _his_ world.
As Adam said, if you don't like the character in a specific game, you're free
to try a different game.

One important part of characterization is, um, characterization. In a piece
of interactive *fiction*, it is usually not you you play (or any instance of
the character that resembles you the best), but the author's interpretation
of the protagonist. Many authors choose to not give the protagonist much
character at all, but that's still their decision.

Incidentally, a very simple way to find out if a certain game fulfils your
interpretation of "nonlinear" is to look at the game file size: if it's an
Inform game, it doesn't. If it's a TDAS game shorter than five megabytes, it
doesn't.
There just isn't enough room to plausibly react to every idea you might
follow.


--
Miron Schmidt <miron@comports.com>                       PGP key on request

WATCH TV... MARRY AND REPRODUCE... OBEY... PLAY INTERACTIVE FICTION...



From scythe@u.washington.edu Mon Sep  8 22:33:53 MET DST 1997
Article: 28346 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!newsfeed1-hme1!newsfeed.internetmci.com!192.220.251.22!netnews.nwnet.net!news.u.washington.edu!scythe
From: scythe@u.washington.edu (Dan Shiovitz)
Newsgroups: rec.arts.int-fiction
Subject: Re: Can't we all just get along?
Date: 8 Sep 1997 19:58:31 GMT
Organization: University of Washington, Seattle
Lines: 53
Message-ID: <5v1ld7$619@nntp5.u.washington.edu>
References: <01bcbc82$d1cbdc60$a69f72ce@hammurabi.blarg.net>
NNTP-Posting-Host: saul7.u.washington.edu
NNTP-Posting-User: scythe
Xref: news.lth.se rec.arts.int-fiction:28346

In article <01bcbc82$d1cbdc60$a69f72ce@hammurabi.blarg.net>,
Brandon Van Every <vanevery@blarg.net> wrote:
>Ah, what would we do without Rodney King.
[..]
>Having pursued DejaNews extensively, I know that this juncture has been
>crossed before.  Any words of wisdom on how to conduct critiques while
>avoiding flame wars, IN PRACTICE?  I've read someone's FAQ-or-other on
>addressing the idea, not the author, etc.  The problem as I see it, is that
>unsolicited criticism typically garners a negative response even if it is
>constructive.  Or at least, it risks an undefined interaction between
>author and critic.   It is tempting to ask permission before making any
>kind of critique upon anything.  Yet, we seem to have no problem radiating
>positive observations about an author's work on an ad hoc basis.  That's
>good for the soul and for the community, but does it not come at the
>expense of alternate views and potentially rewarding insights?  Of course,
>that depends on what one personally feels the "rewards" are, and we already
>know we don't hold the same standards as to what's important in IF theory
>or practice.

I don't think raif'ers are quite as leery of criticism as you say,
though it is pretty true that most people share relatively similar
expectations about IF and tend to flinch similarly when you wander too
far away from them. 

The real problem, though, is that IF is a combination of art and craft
and is also very new -- it's not an "understood" problem by any
means. Because of this, there's an IMO justified suspicion of people
who come in purporting to know what's Right, especially if they have
no experience putting it into practice. While a requirement of having
to "pay your dues" before being taken seriously does tend to lead to
an insular community to some extent (though there's nothing stopping
people *with* experience from doing weird stuff and having it be taken
seriously), it also screens out wildly impractical suggestions which
can not actually be done. And, after all, a major purpose of the
discussions here is to eventually do something with them and write an
actual thingy.

>I just have no desire/inclination to repeat the absurdities that I've read
>compared to the scathing abuses I've seen in past posts.

I think I speak for many people when I say I don't want to see
scathing abuse here; it gets old fast.

>Brandon J. Van Every  <vanevery@blarg.net>      DEC Commodity Graphics
>http://www.blarg.net/~vanevery                  Windows NT Alpha  OpenGL
--
Dan Shiovitz :: scythe@u.washington.edu :: shiov@cs.washington.edu
..................................................................
"Alas, I do not rule the world and that, I am afraid, is the story
of my life: always a godmother, never a God."     -- Fran Lebowitz
...http://weber.u.washington.edu/~scythe/home.html................




From lpsmith@rice.edu Mon Sep  8 23:20:54 MET DST 1997
Article: 28351 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news-peer.sprintlink.net!news.sprintlink.net!Sprint!howland.erols.net!newsxfer3.itd.umich.edu!gumby!newspump.wustl.edu!rice!ural.owlnet.rice.edu!lpsmith
From: lpsmith@rice.edu (Lucian Paul Smith)
Newsgroups: rec.arts.int-fiction
Subject: Re: Physical traversal of landscape
Date: 8 Sep 1997 20:15:24 GMT
Organization: Rice University, Houston, Texas
Lines: 57
Message-ID: <5v1mcs$9c6$1@joe.rice.edu>
References: <01bcbb05$2b27f200$9e9f72ce@hammurabi.blarg.net>
NNTP-Posting-Host: ural.owlnet.rice.edu
X-Newsreader: TIN [version 1.2 PL2]
Xref: news.lth.se rec.arts.int-fiction:28351


Brandon Van Every (vanevery@blarg.net) wrote:

: Let's say
: we're experiencing the story of Little Red Riding Hood.  The game's
: reaction to a statement like "I'M HUNGRY" might bring you to Grandma's
: house.  The game's reaction to "PULL OUT WOLF'S TEETH" might be to bring
: you to a place where the Wolf is likely to be (based on your character's
: current knowledge) and then wait around for awhile.

Actually, this isn't too far off from something I'd like to see in modern
IF.  Something like:

----------------
>PUT THE FROTZ IN THE SPORK.

Oops, you seem to have left the Frotz back in the Main Kitchen.  Go there
now? (Y/N)

>Y

[The game takes over at this point, taking you to the Main Kitchen]

...You see a Frotz here.

Do you still want the Frotz? (Y/N)

>Y

(Taken)

Go back to the Spork room?  (Y/N)

>Y

[The game takes you back across the landscape,...]

...There is a Spork here.

Do you still want to put the Frotz in the Spork? (Y/N)

>Y

The Spork doesn't seem to be able to contain things.

-----------

Well, the last line would be a little frustrating, but at least you'd have
been reminded about where the Frotz was, and let you get it with a minimal
amount of effort.

This would, to me, constitute a much friendlier parser response than
Inform's current (though generally excellent) response of 'You can see no
such thing.'  It has to do with a new concept of 'scope', one which I've
mentioned before here, and which I really do plan on trying to code up one
of these days,...

-Lucian "Lucian" SMith


From mol@bartlet.df.lth.se Tue Sep  9 00:14:25 MET DST 1997
Article: 28354 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Of IF, Jazz, and Why We Can't Seem to Get Along
Date: 8 Sep 1997 23:26:35 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 114
Message-ID: <5v1qib$fht$1@bartlet.df.lth.se>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:28354

Think of this newsgroup as a cafe; the kind of place where people
drop in not just for a cup of tea, but for a chat with
friends. Imagine, further, that this particular cafe is the locus for
a bunch of jazz musicians: people who like to sit around and discuss
their particular kind of music. The hard core of regulars are into
some rather esoteric, "arty" jazz forms, but the place is frequented
as well by trad jazz musicians, people into jazz/hip hop fusion, and
laymen who can't tell hard bop from cool, but still enjoy the music
and like to chat with the musicians.

The regulars tend to get a bit defensive at times. After all, their
particular form of music isn't very popular anymore - it had its
heydays in the sixties, but young people nowadays regard it as some
weird stone-age fossil. But most of the time, the atmosphere is
friendly and open-minded.

Occasionally, patrons venture to discuss country and western, or
even synthpop; sometimes, they quite vocally state that they prefer
these genres to the music of the regulars. They are generally
tolerated, because people here are all _musicians_ (or, at least,
they're all into music).

Every once in a while, some of the regulars fall out over something,
or join up against some hapless newcomer who just asked the same
cluless question the 28 newcomers before him asked, but feelings
quickly calm down again. At long intervals, a stone is thrown through
the window, with an attached banner "JA55 IS DEAD. DETH METL R00LZ,"
but these barbaric incursions are easily ignored (after all, virtual
glass is cheap to replace).


Then, one day, a newcomer walks up to the regulars' table. He's
clearly a sharp young man; fresh out of college, with well-honed
debating skills and lots of radical ideas. He's presented those ideas
before, in other cafes. Sure, his ideas always encounter some
opposition, but he is a bit polemic - intentionally so, the better to
shake up the convicitions of the conservative old-timers - and usually
he's appreciated for the clever fellow he is.

So, he walks up to the regulars' table, and presents his arguments:

Jazz music is dead.

Yeah, sure, everybody knows that it's at least marginalized, the
regulars grudgingly concede.

But our hero knows more than that: he knows the Reason, he has a
Theory. Jazz music is dead - OK, _marginalized_, you old farts still
insist on playing it - because it is improvised. And improvised music
is Wrong. Not only does our hero not like it, he has Objective Reasons
to believe that improvisation is the Wrong Way to Go.

Before the regulars have caught their breaths, he goes on:

In order to survive and evolve, Jazz music Must abandon
improvisation. The Future lies in pre-programmed synthesizers, with
their infinitely better precision and potential for
complexity. The audience doesn't want improvised music, can't
appreciate it, won't appreciate it; and, besides, there is no Artistic
Potential in it. And here our young hereo hauls out a bunch of
academic treatises (written by experts on all kinds of music _except_
jazz) to support his views.

The regulars start protesting that they are not *interested* in making
techno music, they prefer making jazz because they like it, and that
their artistic inclination is in that direction, audience be damned.

"And the artist always is right?" The irony is . The Audience
is right! Of course, our hero means the enlightened, refined audience
which shares his views, not the poor, deluded nostalgics who still
think they like jazz.

And the he pull out his most powerful argument:

"Look, if you conservatives didn't *insist* on sticking to your
antiquated ideas, you would surely see the Truth in my Ideas."

Some of the regulars start mumbling things about muddle-headed
theoreticians, while others bring out their flamethrowers. The young
hero starts loading his shotgun with verbal ammunition such as
"anti-intellectuals", "conservatives", "horticultural approach".

================

At this point, we leave the cafe, before things get too violent.

Let me just add three things:

1) The "young hero" above is *not* intended as a caricature of Brandon
Van Every, or any other person in particular. He's just some sort of
personification of what's happened a few times before on this
newsgroup (Brandon, let me stress that I'm referring to events from
before you joined the discussion).

2) Our "hero" has a very important advantage above certain people in
the IF debate: his alternative to jazz already exists and has shown
its potential. Some of the alternatives to "traditional IF" that have
been advocated with just as much fervour do *not* exist, and their
proponents have not managed to show that they are even *possible*
without strong AI.

3) Perhaps most important: our "hero" is not necessarily wrong in the
sense that his music is less worth playing than is jazz. The vehement
reaction from the regulars has not very much to do with this.



[ I can't believe I wrote this. I really can't believe I dared post it.
  I'm going to be flamed to a crisp. No, I'm going to be flamed into
  nonexistance. I'm crawling back under my rock to abide my curel fate. ]
-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From mol@bartlet.df.lth.se Tue Sep  9 11:55:41 MET DST 1997
Article: 28350 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: Can't we all just get along?
Date: 8 Sep 1997 22:33:06 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 50
Message-ID: <5v1ne2$5l1$1@bartlet.df.lth.se>
References: <01bcbc82$d1cbdc60$a69f72ce@hammurabi.blarg.net>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:28350

Well, I won't enter into the discussion between Adam and Brandon, 
just make some observations.

This group is, in general, remarkably civilized for Usenet. Voices are
raised in anger, sometimes, and people write things they really
shouldn't have written. However, we haven't seen any real flame wars
so far, the endless religious/political/Mac-vs-PC discussions that
plague many groups are missing, and - most important of all - even
when people do get excited and say silly things, the discussion
generally stays on topic and there generally remains some essence of
constcuttive dialogue.

If you feel roughly treated, Brandon - well, you *have* been sticking
your neck out a bit. You see, people here (me included) tend to be a
bit defensive about what they're doing, text-based games being a
(supposedly) dying art form under constant siege and all that. 

In fact, most people who even bother to care at all seem to think that
only strange nostalgics and total fools could think that a text-based
game has anything to offer today compared to graphic games.

On the other hand, traditional literature people tend to scorn
computer games as children's entertainment, and can't really think
that a game can have any artistic merit at all.

So, people are a bit sensitive about it. What I think has happened
here is that people have perceived you as yet another newcomer who
knows not only that what we're doing here is wrong, but also has a lot
of (untested, of course) ideas of how to do it. "Listen, guys, you've
been doing the wrong thing all the time; if you could just see things
as I do you'd be much happier". Sort of like if you went to a jazz
musicians' convention and said "Jazz is boring, you should all be
playing rock instead."

You will probably say that this is a blatant misrepresentation of your
intentions. I hope it is :-). However, it's easy to take your comments
that way. It's happened before - you're not the first person to come
in and *be perceived as* trying to ram your theories down our
collective throat. 

This said, I think you have interesting ideas to contribute, and I hope
the trigger-happines subsides. 




-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From alan@accessone.com Tue Sep  9 11:55:45 MET DST 1997
Article: 28382 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!news-peer.sprintlink.net!news-pull.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!198.68.191.16!news.accessone.com!not-for-mail
From: alan@accessone.com (Alan Conroy)
Newsgroups: rec.arts.int-fiction
Subject: Re: Can't we all just get along?
Date: Tue, 09 Sep 1997 06:41:56 GMT
Organization: AccessOne
Lines: 25
Message-ID: <3414ee50.17463132@news.accessone.com>
References: <01bcbc82$d1cbdc60$a69f72ce@hammurabi.blarg.net> <5v1ne2$5l1$1@bartlet.df.lth.se>
NNTP-Posting-Host: kirk04-17.accessone.com
X-Newsreader: Forte Free Agent 1.11/32.235
Xref: news.lth.se rec.arts.int-fiction:28382

On 8 Sep 1997 22:33:06 +0200, mol@bartlet.df.lth.se (Magnus Olsson)
wrote:

>This group is, in general, remarkably civilized for Usenet. Voices are
>raised in anger, sometimes, and people write things they really
>shouldn't have written. However, we haven't seen any real flame wars
>so far, the endless religious/political/Mac-vs-PC discussions that
>plague many groups are missing, and - most important of all - even
>when people do get excited and say silly things, the discussion
>generally stays on topic and there generally remains some essence of
>constcuttive dialogue.

I second that whole-heartedly.  I only monitor three news-groups
currently, though I've looked at others in the past.  But this news
group is where I come running after being bombarded with petty
nastiness, name-calling, and general disrespect (and that isn't
directed towards me - I just see it while I lurk).
rec.arts.in-fiction is a breath of fresh air after all that.  It's
been said here before, but bears repeating: this is quite a cordial
group.

(I've been meaning to make that comment for awhile, and this seemed a
good opportunity)...

          - Alan Conroy


From adamc@acpub.duke.edu Tue Sep  9 11:56:09 MET DST 1997
Article: 28375 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!howland.erols.net!agate!newsgate.duke.edu!usenet
From: Adam Cadre <adamc@acpub.duke.edu>
Newsgroups: rec.arts.int-fiction
Subject: Re: Can't we all just get along?
Date: Mon, 08 Sep 1997 23:35:51 -0400
Organization: Group 17
Lines: 39
Message-ID: <3414C417.21@acpub.duke.edu>
References: <01bcbc82$d1cbdc60$a69f72ce@hammurabi.blarg.net>
Reply-To: adamc@acpub.duke.edu
NNTP-Posting-Host: res-9-43.dorm.duke.edu
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01Gold (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:28375


Brandon Van Every wrote:
> So herein lies a classic problem.  I didn't "like" I-0.  I think my
> criticism has been constructive, but negative vibes have flown between
> myself and Adam anyways, possibly for differences of personality and
> discursive style if nothing else.  Epithets fly.  "Pithy smart-ass."
> "Theory-headed newcomer."

Well, first of all, I would never call you a "theory-headed newcomer."
It's not pithily smart-assed enough.

I have no problem with anyone not liking I-0.  For one, I'm not that
thin-skinned.  For another, it's not like it's my baby or anything --
it was just a little practice game that caught on to an extent I wasn't
expecting.  In any event, it's beside the point.  As I'm sure you're
aware, the problem came about long before you even addressed it.

At this point I would refer you to Deborah Tannen, YOU JUST DON'T
UNDERSTAND (New York: Ballantine Books, 1990), particularly chapters
three ("Rapport-talk and Report-talk") and five ("Lecturing and
Listening").  Tannen's observations on differences in discursive style
are good, but she makes a mistake in insisting on breaking them down
along gender lines; after all, we're the same sex, yet it irked me to
no end when you showed up out of the blue and, with little or no regard
for your audience, began declaiming.  Now, I'm well aware this probably
boils down to nothing more than a matter of style, and I'm not prepared
to argue that my style is better, though I obviously prefer it.  But
I have little patience for being talked at.  Once again, I urge you to
check out http://c2.com/cgi/wiki?IntFicMudDiscussion for a model of
theoretical discussing I'm much more comfortable with: one in which the
participants engage each other's ideas and each other as people, one in
which they're genuinely trying to collaboratively figure something out
rather than prove how smart they are.  And one in which a little joking
around is taken for granted, because it's fun.  IF is a hobby.  Hobbies
should be fun.  If you're not having fun, it's time to get a new hobby.

  -----
 Adam Cadre, Durham, NC
 http://www.duke.edu/~adamc
 http://www.retina.net/~grignr


From erkyrath@netcom.com Tue Sep  9 11:58:09 MET DST 1997
Article: 28360 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Physical traversal of landscape
Message-ID: <erkyrathEG7o0E.9qz@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <01bcbb05$2b27f200$9e9f72ce@hammurabi.blarg.net> <fake-mail-0709971320210001@van-52-0327.direct.ca> <01bcbc7e$fdd001a0$a69f72ce@hammurabi.blarg.net>
Date: Mon, 8 Sep 1997 22:33:50 GMT
Lines: 63
Sender: erkyrath@netcom22.netcom.com
Xref: news.lth.se rec.arts.int-fiction:28360

Brandon Van Every (vanevery@blarg.net) wrote:

> >  That doesn't mean that theoretical discussions aren't important. Just
> > that a concrete working example does far more to convince or interest or
> > whatever the crowd. If somebody (particularly a relative newcomer who
> > hasn't produced anything for public consumption) comes along

> Well this sort of ego tripping is really rather tiring, I have to say. 

You *choose* to say. (And that's all the snarking I will do in this post; 
honest.)

> I'm
> not going to play games about who's a newcomer and who isn't.  If you feel
> that threatened

... .

Here is a problem: in the past, people have popped up and said, lo, 
current IF is blindly charging into a dead end. It is all crap. Here is 
my theory, which will lift IF into a new realm of literary merit. 

And these people are -- understandably, if not necessarily fairly --
pounded up one side of r.a.if and down the other. This is not really
because they were proffering *theory*; it was because of the "you're all
idiots" part of the statement.

I do not believe that you, Brandon Van Every (any relation to Chris Van
Allsburg? Sorry.) have made such a statement. But you must understand that
there is a certain amount of built-up resentment, and a tendency to see
that kind of challenge in (literally) neutral questions like "Why are we
still using N, S, E, W?" 

There has also been a lot of interesting theorizing from people who have 
not in fact written code. And the poster quoted above is (literally) 
correct; a concrete working example has far more influence then a Usenet 
post. 

Now, I'm sure *this* Usenet post is going to have no influence whatsoever,
but nonetheless I suggest that the whole discussion be tabled for four
weeks. When the '97 competition entries are released, I am sure --
absolutely certain -- that we will see more new IF ideas than you or I can
possibly think up in a year. And *all* of them will be concrete working
examples. And then we will all go "holy shit" and commence thinking up 
yet another set of ideas.

(As my own personal ego trip, I will claim that a lot of this competition
wide-rangingness was inspired by my own "Space Under the Window" 
experiment. (Not that the games will *resemble* SUTWIN! They'll be all
over the map, or lack thereof. I just mean that I wrote SUTWIN as proof
that you *can* throw away huge chunks of the IF conventions of yore, and
the audience will be right there with you, judging the work on its own
merits.)

But you can ignore the ego trip, if you like. Previous paragraphs still 
hold.)

--Z

-- 

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


From namtlua@ten.ttocs Wed Sep 10 09:33:30 MET DST 1997
Article: 28423 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news-peer.sprintlink.net!news-sea-19.sprintlink.net!news-in-west.sprintlink.net!news.sprintlink.net!Sprint!204.181.147.3!news.scott.net!koala.scott.net!not-for-mail
From: namtlua@ten.ttocs
Newsgroups: rec.arts.int-fiction
Subject: Re: Can't we all just get along?
Date: 9 Sep 1997 00:40:57 -0500
Organization: Scott Network Services, Inc.
Lines: 59
Sender: aultman@koala.scott.net
Message-ID: <5v2nh9$3rk@koala.scott.net>
References: <01bcbc82$d1cbdc60$a69f72ce@hammurabi.blarg.net>
NNTP-Posting-Host: koala.scott.net
Xref: news.lth.se rec.arts.int-fiction:28423

In article <01bcbc82$d1cbdc60$a69f72ce@hammurabi.blarg.net>,
Brandon Van Every <vanevery@blarg.net> wrote:

>So herein lies a classic problem.  I didn't "like" I-0.  I think my
>criticism has been constructive, but negative vibes have flown between
>myself and Adam anyways, possibly for differences of personality and
>discursive style if nothing else.  Epithets fly.  "Pithy smart-ass." 
>"Theory-headed newcomer."  Etc. blah blah blah.  More people suck
>themselves into the vortex, resulting in a chain reaction which destroys
>the station.

The problem, Brandon, is really that you started this all off with
an implausible interpretation of "nonlinearity" and then based an
analysis of I-0 on that flawed (IMO) interpretation. I consider
I-0 a triumphant and damn refreshing break from the I-F mold.  It's fun
to play over and look for new things and new ways to get through.
I was thrilled when I figured out how to kill that rapist bastard.
Still, I didn't have to kill him, or even meet him, to progress
and finish the game (which I fully traversed without crying or
pouting once). Sounds nonlinear to me.

You explicitly claim disappointment that you couldn't have sex with
everyone and everything in I-0. I'm forced to wonder how we can have
characterization in your "nonlinear" ideal if that's the case.  You
ask the PC to perform every absurd and improbable action the player
tries.  What if it's a fact of the game world that Tracy doesn't want
filthy ugly guys to rape her?  It is necessary characterization that
Tracy must fight back.  It must be assumed the player will try to
escape.  Nothing else fits the character.

Additionally, you ask that NPC's act as simple mirrors to the player's
whims.  What if the cop doesn't want to screw Tracy on the side of the
road in the desert heat?  What if he is gay (see last week's controversy)
or a loyal family man?  What if he had his balls blown off in the 'Nam?
Characterization demands that NPC's do not comply with everything the
player wants to try.  They must be self-defined and self-motivated.

>Any words of wisdom on how to conduct critiques while
>avoiding flame wars, IN PRACTICE?

In practice you should judge a game by what it tries to do and not what
you think it should have tried to do.  It seems a highly popular
opinion that I-0 suceeded in its goals, none of which were to be
large or limitless.

While I don't doubt that you MEANT your posts to be constructively
critical, it's hard to read them that way when they more closely
resemble a continuation of your "linear" vs. "nonlinear" posts,
using I-0 (inappropriately) as an example of a failure to meet
your criteria.  In that light, your criticism takes on a very
self-serving flavor.  It's my observation that in this forum
any criticism that an author could not possibly use to improve
the very game in quesiton is not well received.

In that vein, I hope you can use this criticism to examine and improve
your vision of nonlinear I-F.  In its current form I find it undesirable
and nigh-impossible.

Joe


From erkyrath@netcom.com Wed Sep 10 18:11:31 MET DST 1997
Article: 28444 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!newscore.univie.ac.at!newsfeed.ecrc.net!howland.erols.net!news-peer.sprintlink.net!news.sprintlink.net!Sprint!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Summary of Nonlinearity
Message-ID: <erkyrathEGAuzz.240@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <01bcb548$f6c96dc0$939f72ce@hammurabi.blarg.net> <ant0814260b0c4bn@arnod.demon.co.uk> <01bcbc85$dba7a2c0$a69f72ce@hammurabi.blarg.net> <ant090010d07c4bn@arnod.demon.co.uk> <5v4d9s$k53$1@bartlet.df.lth.se> <01bcbdf5$ceee6b60$929f72ce@hammurabi.blarg.net>
Date: Wed, 10 Sep 1997 15:57:35 GMT
Lines: 29
Sender: erkyrath@netcom17.netcom.com
Xref: news.lth.se rec.arts.int-fiction:28444


Brandon Van Every (vanevery@blarg.net) wrote:
> Magnus Olsson <mol@bartlet.df.lth.se> wrote in article
> <5v4d9s$k53$1@bartlet.df.lth.se>...
> > In article <ant090010d07c4bn@arnod.demon.co.uk>,
> > 
> > To put it another way: Maybe people can't agree on exactly what is
> > meant by the "linear"/"non-linear" distinction, but most people seem
> > to agree on what it *doesn't* mean.

> This is non-rigorous and therefore, to my mind, pointless.  :-)

IF is art, and by definition, in Art, nothing interesting is rigorous. 
Heh. 

>  I'd rather argue the substance of the matter, than its sign.

To some extent, you bailed in on the ongoing argument over "linear" 
versus "nonlinear" game design. Or at least people thought so; I did. 
When it turned out that you were talking about different things than 
everyone else, we were disconcerted. More so when it turned out that you 
were talking about matters which were *also* an ongoing argument, but not 
a recently active one.

--Z

-- 

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


From erkyrath@netcom.com Wed Sep 10 19:56:35 MET DST 1997
Article: 28449 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news.algonet.se!news.maxwell.syr.edu!newsfeed.internetmci.com!newsfeed.internetmci.com!206.229.87.25!news-peer.sprintlink.net!news.sprintlink.net!Sprint!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform]: Technical/Programming Questions
Message-ID: <erkyrathEGAx3H.85r@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <34174199.1950@rz.uni-potsdam.de>
Date: Wed, 10 Sep 1997 16:42:53 GMT
Lines: 105
Sender: erkyrath@netcom17.netcom.com
Xref: news.lth.se rec.arts.int-fiction:28449

Michael Gaul (gaul@rz.uni-potsdam.de) wrote:

Woo hoo! Questions! 

> 1. Property limitations. The manual states that the number of individual
>    properties is unlimited. Is this true for every single object? I
> might
>    want to implement lots of classes and multiple inheritance, so I
> might
>    end up with objects having 150 properties or such (especially NPCs).
>    So, is there any limit to the number of properties that a single
> object
>    can have? Further, what is the most complex object that you ever
>    designed, and did you run into any limitation problems by doing that
>    (including processing speed)?

No, honest, the number of individual properties is not limited, except by 
the size of an Inform integer (16 bits.) You're more likely to run out of 
memory, since the storage for all property data has to fit in the 64K RAM 
space. 

There is a small delay associated with property lookup, which I think is 
proportional to the size of the property table. But I've never seen this 
become a problem.

> 2. The same questions for classes: What was the most complex class
> system
>    that you ever developed, and did you run into any limitations
>    (including processing speed)?

Classes are a compile-time construct; a convenient way to get several 
properties into an object, from different places in the source code. 
There aren't any limitations associated directly with classes.

> 3. When I inherit demons from several classes, only one demon will
> actually
>    appear, because the demon property is not additive. Can I simply
> declare
>    it as additive to circumvent that or will that cause problems with
> the
>    rest of the library?

Mmm. I think it would work; the library calls RunRoutines(obj, daemon);
and that goes through added properties, right? However, remember that if 
any of the daemon routines returns true, no later ones will be called. 
(The same that happens with added "before" or "after" properties. It's 
usually not important what a daemon routine returns, exactly because it's 
not additive.)

> (By the way, what is the reason for "demon"
> being not additive?)

Lack of specificity, I guess. You're constrained to turning all the 
inherited daemons on and off at the same time. And you can get the same 
effect by having each daemon routine end with
  self.ParentClass::daemon();
In other words, chain them yourself.

> 4. I always felt that actions like "look" or "inventory" should not take
> up
>    game time. In a real situation, you know where you are without
> looking
>    around, and if you do, it takes very little time. Can I simply
> declare
>    ##look and ##inv as "meta" or will that cause problems with the rest
> of
>    the library?

Meta verbs bypass the before/after system, so this will limit you in some 
ways. It may be better to have some time-keeping system which isn't 
strictly one turn at a time. This sort of thing is only relevant in a 
small part of a game, yes? (Barring absurdities like "Weather". :-)

A more useful hack, which I have never implemented, would be to abstract 
the end-turn code (daemons, timers, and each_turn) into a routine and 
have the library call it N times after each command. N is 1 by default, 
of course, but an action routine could set it to 0 or a higher number.

> 5. How does Inform distinguish between rooms and objects (in the sense
> of
>    "things")? Many explanations in the manual suggest that it does,

It doesn't.

If an object is the outermost object surrounding you (the parent of your
parent of your parent, etc) then the library treats it as a room. 
Otherwise, the library treats it as an object. 

I recommend never confusing the two situations; some properties behave
differently in the room or object treatments. (For example, "initial". And
"description" is the result of an examine command on objects, but a look
command in rooms.  And "before"/"after" are called with noun=self for
objects, but with noun=something else for rooms; note that "before" on a
room is similar to "react_before" on an object, but not to "before" on an
object. 

Really, I don't know *any* situation when you'd want a single Inform
object to be a room and an object at different times. 

--Z

-- 

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


From mol@bartlet.df.lth.se Thu Sep 11 09:51:23 MET DST 1997
Article: 28487 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.games.int-fiction,rec.arts.int-fiction
Subject: Re: An irrelevant Esperanto comment that grew
Date: 11 Sep 1997 09:49:19 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 103
Message-ID: <5v87pv$3bu$1@bartlet.df.lth.se>
References: <19970806171001.NAA05520@ladder01.news.aol.com> <wkraaxurrh.fsf@turangalila.harmonixmusic.com> <5v6fnq$1gl$1@bartlet.df.lth.se> <5v6lep$e7g$1@vixen.cso.uiuc.edu>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.games.int-fiction:25835 rec.arts.int-fiction:28487

In article <5v6lep$e7g$1@vixen.cso.uiuc.edu>,
Jonathan Badger <badger@aquarius.scs.uiuc.edu> wrote:
>[While this is irrelevant to the topic of IF in Esperanto or
>otherwise, Magnus Olsson, a regular here, did post on the subject, so
>I guess it's kosher to respond]

It's OK to respond, but the Powers That Be will deduct 2.5 megapoints
>from your Karma Score. :-)

>>That's true today. However, I was wondering what would happen if a
>>large population of people were to speak Esperanto in their daily lives
>>for, say, 500 years - would the pronunciation change so much that the
>>various words derived from the same stem wouldn't be distinct? And how
>>would the language evolve in that case? 
>
>I think you are missing the point of what Esperanto is for. It isn't
>meant to be used like an ethnic language for only talking to one's
>neighbors, but as a secondary language like Latin was in the medieval
>and Renniasance periods, for communicating orally and in written form
>with people who might a a very different ethnic language.

Well, I'm aware that Zamenhof's intention was to create an auxilliary
language to help overcome language barriers. However, I'm not sure that
not some Esperantists want it to become a living language - there are,
after all, people writing novels in Esperanto and such things.

And regardless of intention: how do you force people to pronounce
Esperanto in the correct way, or to conform to the correct grammar?
("correct" == according to the official language definition, or
whatever).

Besides, I was merely speculating: *what if* Esperanto started
evolving like any living language? Would it still be as regular after
500 years of daily use?

>Among the peasants, Latin did evolve into the Romance languages, but
>this was because 1) After the fall of the Roman Empire, people were
>isolated and had little or no contact with other Latin speaking
>peoples,

Actually, the differentiation of Vulgar Latin seems to have started already
during the Empire. Which isn't surprising, given the huge distances involved
and the different substrata of the various pre-Roman languages.

> and 2) Most people were illiterate, and thus had no access to
>Latin writers of the past, who could supply good models of Latin
>usage.

A literary tradition can certainly help preserve a language - as long
as people view the old writers as canon. In most written languages
today, people don't view writers more than a hundred years old or so
as models of good usage. English is one exception (along with, for
example, Arabic).

>But among scholars, Latin remained alive and well. Almost all
>scholarly research was written in it and it was the standard language
>used in university lectures. While certainly new words were coined,
>the language remained pretty much the same, and any educated person
>who could read Copernicus' scientific works could read Cicero just as
>easily. And when Bruno came to England from Italy, his lectures in
>Latin were understood by his English colleagues. This sort of thing
>went on for hundreds of years without significant grammatical changes
>in the language.

I think you're idealizing things a bit. As I've heard it, Latin
continued to be a living language among educated people well into
medieval times, and it *did* change grammatically. Not so much that
people couldn't understand the classical writers, but an 11th century
Italian would *not* write the same kind of Latin as Cicero did, no
more than a modern Englishman would write like Shakespeare.

In the early Rennaissance, the humanists wanted to revive the "pure"
classical Latin in favour of the "degenerate" medieval Latin. They
succeeded to a degree. At the same time, Latin became a dead langauge
- whether it was despite or because of the humanist revival is
debatable.

>Today Esperanto is used in much the same way. I subscribe to an
>Esperanto magazine from China, and have read books in the language
>from Chinese, Hungarian, Polish, Croatian, and French writers. The
>fact that these people are writing for an international audience tends
>to limit the formation of any non-standard dialects of Esperanto.

Do you notice any influences from their native languages on the
Esperanto written by all these nationalities? For example, can you
tell if a certain Esperanto text is written by a Chinese or by a
Frenchwoman?

Besides, even if the Esperanto language is kept homogeneous by the
constant international exchange, it could still evolve. For example, a
purely hypothetical case: Suppose native speakers of language X don't
distinguish between the vowels 'a' and 'o' in their own language, and
hence tend to avoid Esperanto constructs that depend on this
distinction. Suppose further that such a person writes something very
influential in Esperanto, that causes Esperantists of other
nationalities to imitate her. Unless there were an "Esperanto Police"
that corrected all these "deviations", the language could very well
start to change. 

-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From jarobe@po.eecs.berkeley.edu Thu Sep 11 10:35:02 MET DST 1997
Article: 28485 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!news1.best.com!nntp1.ba.best.com!not-for-mail
From: Jake Roberts <jarobe@po.eecs.berkeley.edu>
Newsgroups: rec.arts.int-fiction
Subject: Re: Physical traversal of landscape
Date: Thu, 11 Sep 1997 00:12:36 -0700
Organization: No
Lines: 106
Message-ID: <341799E3.2185@po.eecs.berkeley.edu>
References: <01bcbb05$2b27f200$9e9f72ce@hammurabi.blarg.net> <fake-mail-0709971320210001@van-52-0327.direct.ca> <01bcbc7e$fdd001a0$a69f72ce@hammurabi.blarg.net>
Reply-To: jarobe@po.eecs.berkeley.edu
NNTP-Posting-Host: dynamic23.pm04.wc.best.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.0Gold (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:28485

Brandon Van Every wrote:
> 
> Neil K. <fake-mail@anti-spam.address> wrote in article
> <fake-mail-0709971320210001@van-52-0327.direct.ca>...
> > In article <01bcbb05$2b27f200$9e9f72ce@hammurabi.blarg.net>, "Brandon Van
> > Every" <vanevery@blarg.net> wrote:
...
> to you and say "Go ahead.  Talk about whatever you wish.  (Or not.)"  Over
> dozens of high-powered intellectual newsgroups, I've consistently observed
> that people are interested in talking about the things that I tend to talk
> about, and I never have any shortage of people who want to participate and
> contribute.
 
I've lurked on this group for quite a while.  It's the theoretical
discussions that I usally poke into, skipping over questions about
minute subtleties of Inform or TADS coding.  But even the theoretical
discussions touch base frequently with practical issues of
implementation.  I think this is in many ways a nuts-and-bolts crowd. 
After all, one of the most attractive things about IF is that a single
person can complete an entire work on his/her own, working only in
his/her spare time.  It's quite remarkable, when you think about it.
This precious quality, "finishableness", is what makes amateur
IF-writing feasable for most people who do it.  People want things that
work, now.  That (in addition, I think, to a perceived tone of contempt
on your part for the games people have humbly been producing so far) is
why folks are suspicious of what sound like remote fanciful goals for
the future of IF.

Still, I'd be surprised if this kind of discussion didn't have an
impact, perhaps subtle but still signifigant, on the way some future
games are written.  Authors want to please their audience, and it pains
them when a member of that audience (you) vocally criticizes the style
in which they work.  So I think many authors (including me as soon as I
legitimately become an author) will keep your remarks at the back of
their/our minds.     

> And it's easy to spread negative energy where none is required.  To some
> degree one has to be cautious about spreading it, although it will
> inevitably happen.  And when one receives it, one can hopefully remember to
> have a slightly thicker skin, so as not to escalate it.  High morale is the
> only thing that gets volunteer projects done.  In fact, high morale is
> pretty important to getting commercial projects done as well, $$$$ will
> only take you so far.

Do I correctly perceive (from your allusion to the Impressionist
movement, for one thing) that you'd like there to be a movement in the
IF community, cooperation, common focusing of energy, in order to
achieve the sort of sweeping, long-term paradigm-movement that you've
been suggesting?  Is this why you're talking about morale?

If so, I think it's important to consider why people like traditional IF
before trying to persuade them to look beyond it's sad old predictable
devices.  In my very basic analasys, traditional IF games are, broadly
speaking, three things to the people who enjoy them:

1. A story
2. A puzzle
3. A place

Although of course the relative importance of these three ingredients
varies from game to game.  By "a place", I mean a world whose structure
is comprehensible, whose character, landmarks, and layout becomes
familiar, a volume of space that you can wonder around in, go in circles
if you want to.

My point is that in proposing the idea of a non-linear, non
goal-oriented design, you stepped on numbers 1 and 2, and in proposing
trancendental, logic-of-the-heart style navigation commands, you've now
trodden on number 3 as well.  And although some things need stepping on
>from time to time, insensitivity to these issues won't help you win
support for your ideas.  I, for example, don't ever want to see any of
these three qualities lost by the body interactive fiction as a whole.

But I'm being a little unfair.  For one thing, your suggestion that "I'm
hungry" should beam you to grandmother's house, can, as you've
suggested, happily coexist with a regular old cartesian-mapped,
compass-navigated route to grandmother's place.  The game is simply
playable on more than one level, in more than one style.  And this, I
think, is the key to success.  Ideas like the ones you propose are best
eased into, in such a way that everyone can stop easing along with you
at whatever point they choose.  If you're going to write anything, I
would start with a traditionally structured game overlayed, enriched,
with elements of your new approach.  A game which must be entirely
navigated with commands like "I'm cold and lonely" will either be hell
to write, hell to play, or both, while a mostly straight-laced game that
responds sympathetically to such desperate pleas, as well as
understanding the usual commands, could be quite powerful and thought
provoking.  And of course, finishable.  You would also be exploring the
possibilities of your ideas in circumscribed, miniature form, where I
think you have a better chance of initial success.  I'd quite like to
see this, and if you don't write it then maybe, somewhere down the line,
I will.

What I'm saying is, I think you should take the suggestion that you
produce a working model *as* a suggestion, rather than just a petulant
complaint.  

- Jake

> Cheers,
> --
> Brandon J. Van Every  <vanevery@blarg.net>      DEC Commodity Graphics
> http://www.blarg.net/~vanevery                  Windows NT Alpha  OpenGL
> ------------------------------------------------------------------------
> The anvil upon which you hammer another's words is as hard or as soft
> as you care to make it.  Wherein lies insight?


From dicks@math.ohio-state.NO.SPAM.edu Thu Sep 11 17:21:38 MET DST 1997
Article: 28498 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!howland.erols.net!math.ohio-state.edu!not-for-mail
From: dicks@math.ohio-state.NO.SPAM.edu (Ethan Dicks)
Newsgroups: rec.arts.int-fiction
Subject: Re: Scott Adams + Others
Date: 11 Sep 1997 13:06:09 GMT
Organization: Department of Mathematics, The Ohio State University
Lines: 50
Message-ID: <5v8qc1$ps3$1@mathserv.mps.ohio-state.edu>
References: <5v8jt5$tfe@inferno.mpx.com.au>
NNTP-Posting-Host: math.mps.ohio-state.edu
X-Newsreader: trn 4.0-test58 (13 May 97)
Originator: dicks@math.ohio-state.edu (Ethan Dicks)
Xref: news.lth.se rec.arts.int-fiction:28498


In article <5v8jt5$tfe@inferno.mpx.com.au>,
Stuart George <entropy@mpx.com.au> wrote:
>
>mm i was just playing a C64 game called "Castle Adventure"
>and it happened to b e in basic and so listable.
>what struck me was how familiar the "data" looked.
>it was very much like the scott adams data files.

There's a reason for that.

>anyone know (highly unlikely:) who wrote said adventure?

David Malmberg (sp?)

>i might see if i can port the c64 version data to
>scott-free... hmmm strange.

Perhaps.  There is *one* difference between David Malmberg's adventures and
Scott Adams' - In the original engine, *all* words that make sense to the
parser must be in the vocabulary table.  In David's modified parser, if the
noun is not in the noun list, the parser checks the item array for carryable
items that match the input noun.  If *both* lists turn up negative, the
parser tells the user that it doesn't recognize the word.

This was written up in an old Creative Computing article that was referenced
in a comment in the Castle Adventure listing.  I originally ran into this
program on the PET.  Perhaps someone removed the comments when it was moved
to the C-64.

Back when I was a tot (read: 13), I had a 32K PET and the 1st two SA 
adventures.  I was fascinated by them and wanted to see how they worked.  I
took apart that engine line by line, and even started writing adventures
of my own.  My first big assembly project was to hand-assemble a SA engine
using only the machine language monitor.  In the middle of the project, I 
got my hands on the fancy , add-on monitor with the line-at-a-time assembler;
I didn't have to calculate my own branch offsets anymore.  Yay!  Given that,
when I ran across Castle Adventure, I was able to modify my program to
handle its data file.  If you want to do the same for Scott Free, in the
language of the Implementors, Feel Free.  Just remember the vocabulary 
difference, and you'll be fine.

Enjoy,

-ethan

-- 
Ethan Dicks                                      http://www.infinet.com/~erd/
(dicks) at (math) . (ohio-state) . (edu)       sellto: postmaster@[127.0.0.1]

harvestbot fodder: president@whitehouse.gov fccinfo@fcc.gov root@[127.0.0.1]


From thoni@softlab.se Thu Sep 11 20:42:05 MET DST 1997
Article: 28509 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!in2p3.fr!oleane!newsxfer3.itd.umich.edu!newsfeed.internetmci.com!newsfeed.internetmci.com!194.22.194.4!masternews.telia.net!news.lejonet.se!not-for-mail
From: Thomas Nilsson <thoni@softlab.se>
Newsgroups: rec.arts.int-fiction
Subject: Re: IF Authoring Systems
Date: Mon, 08 Sep 1997 14:55:28 +0200
Organization: SoftLab
Lines: 52
Message-ID: <3413F5BF.F63E9126@softlab.se>
References: <34139C3E.7609@juno.com>
Reply-To: thoni@softlab.se
NNTP-Posting-Host: trias.softlab.lejonet.se
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
X-Mailer: Mozilla 4.02 [en] (X11; I; SunOS 5.5 sun4m)
Xref: news.lth.se rec.arts.int-fiction:28509

Jeff Hatch wrote:
> 
> What would a great interactive fiction authoring system be like?
> 
<good motivation for improved tool support deleted>

> A great interactive fiction authoring system should be a cross between a
> database and a compiler.  It should present the attributes of
> user-created "objects" in a table so they're easy to edit.

Although a more visual environment is a very good thing I think this can
only solve part of the problem. Why? Because this can only be used to
define the simple things such as the static environment. Have you tried
to imagine what programming the simple Dam Control panel in Zork/Dungeon
would be like?

To do anything more complex your back at "typing" or "building"
expressions, sequences of statements etc. even though they would be in a
tool where you could find your way around more easily.

So my theory is that these types of systems have not emerged for IF (nor
in great numbers for any computer language really) because the benefits
does not currently greately overcome the trouble it is to develop them.

Another problem that has to be resolved is how to handle to flow of the
story. Most IF has a beginning, a middle and an end. A simple-minded
implementation using a database would be cumbersome to use because you
need to set and reset the "state of the world" to test if your story
plays as you want it to. As an example, I use many short initial, start
sections that set up different situations so that I can test them.
Something similar has to be available.

This is *not* to say I wouldn't want such a system, au contaire. This is
of course the next step forward. I am just saying watch your step, there
might be a large hole there...

	Thomas

PS. I have been looking for many years for a system or tool which would
allow one to *define* such an environment, much like the parser and
compiler generators now part of every software engineers toolset, a
programming environment generator, if you will. I have yet to find such
a system, but when I do (or build) I will have hundreds of applications
for it! DS.

-- 
"Little languages go a long way..." (ThoNi of SoftLab in 1985)
------------------------------------------------------------------------
Thomas Nilsson                   Phone Int:(+46)13235704 Nat:013-235704
SoftLab                          Mobile: 070-5617541
Datalinjen 1, S-583 30 LINKPING E-mail: thoni@softlab.se
SWEDEN                           http: www.softlab.se, www.rational.com


From gruffin@netcom.com Sun Sep 14 10:01:23 MET DST 1997
Article: 28639 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!ix.netcom.com!gruffin
From: gruffin@netcom.com (Giovanni Ruffini)
Subject: Re: Zork: Grand Inquisitor
Message-ID: <gruffinEGH23s.3C3@netcom.com>
Organization: Netsnort
References: <340fbbd9.51621232@news.supernews.com> <341719b1.2769748@netnews.worldnet.att.net> <34173805.164777937@newsserver.rdcs.kodak.com> <lynnjEGEF4z.3Ct@netcom.com>
Date: Sun, 14 Sep 1997 00:16:40 GMT
Lines: 18
Sender: gruffin@netcom15.netcom.com
Xref: news.lth.se rec.arts.int-fiction:28639

A partial corrective to what's been said above. The "Zork" entry in the 
RTZ Encyclopedia Frobozzica was only partially a quote. The second half 
of the entry is an independent description of what happened in the Zork 
Trilogy. I oughtta know... I was the guy that copied... er... wrote it.

Yes, the RTZ Encyclopedia uses the Z0 entries verbatim wherever possible, 
but the sources of information for the RTZ version are much more 
expansive, including, in several cases, old Infocom newsletters, some of 
the old Zork books, etc.

Speaking of game packaging -- hey Laird, is Activision going to do 
anything with that History of the Zork World it bought from me? I've got 
three words for you: Grand Inquisitor Packaging. :)

Yers,

Nino Ruffini



From dicks@math.ohio-state.NO.SPAM.edu Sun Sep 14 10:03:39 MET DST 1997
Article: 28638 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!wnfeed!worldnet.att.net!205.252.116.205!howland.erols.net!math.ohio-state.edu!not-for-mail
From: dicks@math.ohio-state.NO.SPAM.edu (Ethan Dicks)
Newsgroups: rec.arts.int-fiction,rec.games.int-fiction
Subject: Re: Hugo Zork
Date: 13 Sep 1997 23:35:54 GMT
Organization: Department of Mathematics, The Ohio State University
Lines: 27
Message-ID: <5vf80q$15l$1@mathserv.mps.ohio-state.edu>
References: <3410A9C6.4577@pixi.com> <34189900.1AD3@earthlink.net> <ant122158b49c4bn@arnod.demon.co.uk> <MPG.e84a8c5b1b6b95f989ee2@news.alt.net>
NNTP-Posting-Host: math.mps.ohio-state.edu
X-Newsreader: trn 4.0-test58 (13 May 97)
Originator: dicks@math.ohio-state.edu (Ethan Dicks)
Xref: news.lth.se rec.arts.int-fiction:28638 rec.games.int-fiction:25910


In article <MPG.e84a8c5b1b6b95f989ee2@news.alt.net>,
Fortran Dragon <xyzzy@ponyexpress.net> wrote:
>My glass typewriter shows Julian Arnold saying...
>[Snip]
>> Technically I believe yes, _Hugo Zork_ does infringe Activision's
>> copyright.  OTOH _Zork_ has a (continuing) hazy history concerning it's
>> copyright status--the public domain _Dungeon_ incarnation, Activision's
>> recentish release of _Zork_ as freeware.
>
>	I don't know if Dungeon is really in the public domain.  I thought 
>it was copyrighted by someone.

Yes.  The comments in the MDL source code say the copyright holder is MIT.
The comments in the FORTAN source code say the copyright holder is Infocom.
>From qan old e-mail discussion with one of the Implementors, I was told that
when Infocom formed, they were told that MIT had no financial interest in
their works, so they could freely take them commercial, even though MIT
still held the copyright.

How's _that_ to muddle the issue?

-ethan
-- 
Ethan Dicks                                      http://www.infinet.com/~erd/
(dicks) at (math) . (ohio-state) . (edu)       sellto: postmaster@[127.0.0.1]

harvestbot fodder: president@whitehouse.gov fccinfo@fcc.gov root@[127.0.0.1]


From graham@gnelson.demon.co.uk Wed Sep 17 18:20:53 MET DST 1997
Article: 28812 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!dispatch.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction,rec.games.int-fiction
Subject: Re: Hugo Zork
Date: Tue, 16 Sep 1997 19:05:41 +0100 (BST)
Organization: none
Message-ID: <ant161841868M+4%@gnelson.demon.co.uk>
References: <3410A9C6.4577@pixi.com> <5vgrg3$ikm$1@bartlet.df.lth.se> <5vijve$5ie@joker.rz.hu-berlin.de> <ant151812bc8c4bn@arnod.demon.co.uk> <5vkajr$fj2$1@bartlet.df.lth.se> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 27
Xref: news.lth.se rec.arts.int-fiction:28812 rec.games.int-fiction:25985

In article <5vkajr$fj2$1@bartlet.df.lth.se>, Magnus Olsson
<URL:mailto:mol@bartlet.df.lth.se> wrote:
> Well, we can argue this until we're blue, without knowing what a real
> judge would say about it if it were ever brought to court. The best
> thing you can do is to point people at the various copyright law
> resources on the net. And it's better to be cautious: some companies
> don't like fans infringing on their copyrights, no matter how harmelss
> it may seem, and some of them are apparently prepared to throw
> considerable resources at defending their copyrights. You don't want
> to be crushed by a legal juggernaught. (Not that Activision has shown
> any such tendencies so far).

I don't wish to give any specific details, but I know of a case
in which the very same Activision has been sending threatening
lawyers' letters to a small company (which couldn't possibly stand
up to them in court) over a frankly doubtful claim of copyright
violation.  It's not what the law says.  It's how much money you
have, compared to how much Activision has.  And do not confuse
the nice people who design Zork: The Grand Inquisitor with their
colleagues in slightly sharper suits down the corridor.

Besides, "Dungeon" is at least as good a name, and has some historical
roots.

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From graham@gnelson.demon.co.uk Thu Sep 18 12:30:49 MET DST 1997
Article: 28891 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!news-kar1.dfn.de!news-was.dfn.de!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!dispatch.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Inform library 6/7 now out
Date: Thu, 18 Sep 1997 09:50:59 +0100 (BST)
Organization: none
Message-ID: <ant1808590b0M+4%@gnelson.demon.co.uk>
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 137
Xref: news.lth.se rec.arts.int-fiction:28891


Inform Library 6/7 now released
===============================

I would not normally release a new library exactly a month after the
last one (6/6, of course): except that the competition deadline is
imminent and a great many games will probably be compiled in the
next fortnight.  The bugs fixed are minor but cosmetically
worthwhile (since some text messages were ungrammatical in 6/6).

As a temporary measure, you can get hold of the new library from the
Inform 6 Home Page:

    http://www.gnelson.demon.co.uk/inform.html

(in the form of a tar archive containing all 8 of the files
making up the library).  It will be filed in its proper place at
ftp.gmd.de shortly, and then the temporary uploads page will be
taken out.

As usual, some browsers may experience slight delays before the
new files become visible from my Internet site.

I should like to thank everybody who emailed me over the last
week or so in response to my plea for bug reports.

                                                  Graham Nelson
                                                  18 September 1997

-----------------------------------------------------------------------------

Extracted from the 6-series library modification history, which
resides on the World Wide Web at:

   http://www.gnelson.demon.com/lhistory.html

-------------------------

248. When the parser asks a "Which do you mean...?" question, it normally
    ignores the answer if the answer begins with a verb.  A new exception
    to this is that if that verb is one of...

       long  short  normal  full  brief  verbose

    then it'll be treated as an answer after all, because it's probably
    meant as an adjective.

249. The DarkToDark routine wasn't being called in library 6/6.

250. Capitalised "They're ..." and "Those ..." library messages were
    being garbled (this applied only to objects given the pluralname
    attribute).

251. An incorrect linkage importation has been taken out of "linklv.h":
    this makes no difference unless you're linking the library files
    and you're using Inform 6.14 or later (which is able to detect the
    incorrectness and gives an error, unlike Inform 6.13 and earlier).

252. The implementation of turning scripting on and off has been
    rewritten, so that it no longer gets confused by restarts of the
    game in mid-transcript.  (For instance, the variable "transcript_mode"
    has the correct value when the game begins.)  Also, the verb
    "transcript" has been added to the standard grammar as being
    equivalent to "script": thus, "transcript on" is now a legal
    command.  (Two LibraryMessages have been added: ScriptOn 3 and
    ScriptOff 3.)

253. Library messages referring to "that", "those", "him", etc.,
    further refined to print better English when a person is the
    object in question.  For instance,

       > put me in me
       You need to be holding yourself before you can put yourself
       into something else.
       > set george to 5
       No, you can't set him to anything.
       > turn me on
       You're not something you can switch.

    (Apologies to those hoping for a racier response.)

254. "Floating objects" are now correctly placed when a ChangePlayer
    call occurs.  Furthermore, the requirement that player objects
    must provide a "number" property is hereby withdrawn, and the
    same change means that the library is no longer confused if
    a change back occurs to a previous player-object which has been
    moved between light and darkness since its previous incarnation.

255. "Floating objects" are now removed if their found_in properties
    or routines indicate that they should not be present in a
    particular location.  This ensures that if the found_in test
    gives different results over time, the object is actually taken
    away from any room which should no longer have it.  It doesn't
    apply if the object is given "absent", as this would possibly
    interfere dangerously with the designer's own code, or with a
    use of the same object as an ordinary game item.
    (I'm very grateful to Paul Mikell for sending me detailed and
    accurate diagnoses of the bugs behind changes 254 and 255.)

256. If the constant MANUAL_PRONOUNS is defined, then the parser
    won't automatically update pronoun settings to follow items
    mentioned in inventory listings or room descriptions, so that
    pronouns will only reflect what the player actually types.
    (I don't think there's ever going to be consensus on what the
    pronouns ought to do: people seem split roughly half-and-half:
    so now you can make Inform behave either way.)

257. If the constant DEATH_MENTION_UNDO is defined, then the game
    will offer "UNDO the last move" as one of the options when the
    game is over.

258. Score changes (if any) are notified immediately after a
    reincarnation (if any) resulting from a designer's AfterLife()
    routine (if any), rather than being notified a turn later.

259. A bug in some versions of the "Zip" interpreter and its
    derivatives (such as "MaxZip") caused code in the parser, to do
    with reading the answers to questions like "What do you want
    to take?", to be misinterpreted: the result was that the
    answers to such questions might be misunderstood if a story
    file were played on such an interpreter.  Details of the bug have
    been circulated to interpreter-maintainers, but in the mean time
    I've altered the Inform parser so that the issue no longer
    arises.

260. The text describing the standard compliance number of the
    interpreter (if any) has been moved from the main game banner to
    the output from the "Version" command (which is also printed at
    the start of any recorded transcripts).  This is a little less
    cluttered; the Standard is well-enough established now to need
    less advertisement.

-----------------------------------------------------------------------------

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From marsh@freenet.tlh.fl.us Thu Sep 18 18:02:45 MET DST 1997
Article: 28897 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!in2p3.fr!oleane!news-raspail.gip.net!news.gsl.net!gip.net!news-peer.gsl.net!news.gsl.net!gip.net!news-peer.sprintlink.net!news-sea-19.sprintlink.net!news-in-west.sprintlink.net!news.sprintlink.net!Sprint!204.117.192.21!news.utelfla.com!nntp2.sprintans.net!news
From: marsh@freenet.tlh.fl.us (Steven Marsh)
Newsgroups: rec.games.int-fiction,rec.arts.int-fiction
Subject: Re: Ending of The Pawn?!?
Date: Thu, 18 Sep 1997 14:10:49 GMT
Organization: Sprint ANS
Lines: 80
Message-ID: <342228a1.4997186@news.nettally.com>
References: <3419dffd.342951@news.nettally.com> <341EE07F.5915@df.lth.se>
NNTP-Posting-Host: m4-124.ppp.nettally.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Newsreader: Forte Agent .99g/16.339
Xref: news.lth.se rec.games.int-fiction:26039 rec.arts.int-fiction:28897

	Just to warn y'all, this kinda turns into a min-rant about
Magnetic Scrolls.  Also, mini-spoilers on Guild of Thieves and The
Pawn, again.  Finally, it turns into a quasi-theoretical question on
the nature of text adventure endings themselves (hense the cross-post
to rec.arts.int-fiction).  And it does windows.

On Tue, 16 Sep 1997 21:39:43 +0200, Darius Katz <katz@df.lth.se>
wrote:

>Steven Marsh wrote:
>> 
>>         I just finished the Magnetic Scrolls game "The Pawn", and I
>> have to say...
>> 
>>         [WARNING:  SPOILER ALERT FOR "THE PAWN" AND (MINORLY)
>> "WONDERLAND"]
>> 
>>         THAT'S IT?!?!
>>         Did they just forget to end it?  I mean, I don't recall any
>> congratulatory text, nothing that informed me that the game had
>> ended... did I miss something?
>
>As for as I remember you could enter a special room at the end of the
>game. In there a lot of people, programmers if I'm not wrong (that must
>mean the Magnetic Scrolls team), gave you a round of applause and cheers
>to celebrate your completing the game.
>
>Then you entered a special debug mode and could explore the game a bit
>more (and maybe rescue the princess?)

	True enough.  I can't remember the exact words, but it was
along the lines of, " 'He removed the bracelet.  Let's go get lunch. '
They hand you some code."  I mean, it certainly wasn't a -big-
indicator that the game had ended.  I also mean that in the sense of
the story of the plot (such as it was).  Best as I can tell, the idea
was to remove the bracelet so you could... return home?  Go where you
belong?  Move beyond the red dotted line boundary?  If you move beyond
the boundary, you're stuck (near as I can tell... is there anything
you can do with the waterfall?).
	Pawn was the most guilty of the Magnetic Scrolls games for
it's (non-) ending, but they all seem to not end very satisfying.  I
mean, I finished Guild of Thieves yesterday.  It was one of the
hardest text adventures I've ever played (mostly because the game
would forget to tell you something was there, or you'd just have to
-guess- that the desrciptionless rock was a -valuable- rock), but it
was also pretty satisfying (the various dice puzzles were quite good).
But the ending was basically, "Congratulations.  You're now in the
Guild of Thieves."  I mean, it was that short!
	Now, I'm not expecting a novella when I complete a game, but I
-am- looking for -some- sense of accomplishment... hopefully somewhat
proportional to the amount of effort it took to solve the game.  I
don't expect wondrous prose from, say, Scott Adams, 'cause the games
are pretty limited, and I can solve one in an evening (In an absurdist
way, I particularly like the ending of Mission Impossible (#3): "You
defuse the bomb  You win."  Dump back to DOS.)  But in Zork Zero (a
game where, even if you know -exactly- what you have to do and have a
walkthrough, it would -still- take you about three hours) I expected,
and received, a very satisfying ending.
	In fact, most of the Infocom games had very good endings.  A
Mind Forever Voyaging, Deadline, Zork III, Wishbringer, Spellbreaker,
Stationfall, Beyond Zork, Infidel... these are all games that spring
to mind as having been satisfying to complete (even if the games
themselves may be lacking).  Likewise, the endings of most fan-created
text adventures have been fulfilling.
	But the Magnetic Scrolls games have all been disappointing to
win.  They're a joy to play, and I suppose that's the most important
thing, but I admit to playing a game for the end... it's the biggest
reason that I strive to do the fiddly bits that -aren't- as satisfying
to play (many of the "run around like a chicken" puzzles in Zork Zero
spring to mind).  And I've found myself not pushing in Fish! as much
as I could, 'cause I just know I'm going to be disappointed.
	So... how important is the ending?  Obviously the fans of text
adventures -don't- need the CD-ROM 30-minute full-motion video
conclusion, or else they wouldn't be playing text adventures.  But
beyond that, how much do we need to be rewarded by winning?
	Just food for thought.

Steven Marsh
marsh@nettally.com
	"Retro is just nostalgia with a nosering."  -- Frank Miller


From gkw@pobox.com Fri Sep 19 09:55:55 MET DST 1997
Article: 28947 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!newscore.univie.ac.at!newsfeed.ecrc.net!cpk-news-hub1.bbnplanet.com!cam-news-hub1.bbnplanet.com!news.bbnplanet.com!prodigy.com!nntp.earthlink.net!usenet
From: gkw@pobox.com (Gerry Kevin Wilson)
Newsgroups: rec.arts.int-fiction
Subject: Re: [Hype]  Runtime modification in new system
Date: 19 Sep 1997 05:20:39 GMT
Organization: None
Lines: 43
Distribution: world
Message-ID: <5vt237$cn2@suriname.earthlink.net>
References: <3420989C.5DF2@juno.com> <5vqmv1$lks$1@crux.cs.su.oz.au> <wk90wugs96.fsf@turangalila.harmonixmusic.com> <3421FE03.5254@juno.com>
NNTP-Posting-Host: 207.217.141.95
X-Newsreader: News for Windows NT X1.0-93
Xref: news.lth.se rec.arts.int-fiction:28947

 Dan Schmidt wrote:
 
> Just like everyone, I wonder just how much of all the stuff Jeff's
> promising is going to exist and at what level of quality, not to
> mention when the thing will actually be released (I've co-written a
> few big published programs, and the last 20% took a LONG time; heck,
> look at Avalon).

Grr.  Here I am, haven't posted anything in a long time about Avalon,
running this here contest thingy, and handing SPAG off to Magnus so it
actually gets done, and still, STILL I am the butt of late jokes.  :P
I tell ya, I get no respect, no respect.
 
> But I see no reason not to give him the benefit of the doubt, instead
> of making sarcastic remarks about a program that he's seen and we
> haven't.

Likewise, wait until the corpse is cold before moving in to feast, sheesh.
 
> I guess one reason is his tone.  For some reason, humility seems to be
> a common characteristic among IF designers (I'm not sure why; perhaps
> it has something to do with working in a slightly antiquated medium?),

Hey, I'm not humile, and yet you guys put up with me for years.

[Hmm, now, what odds am I laid that someone will correct the above
sentence...I think I've got a safe bet here, being the Internet and all...

> and Jeff seems to be a little, um, humility-poor.  So I understand why
> people could get irritated by Jeff's arrogance (the negative side of
> it; the positive way to put it would be "self-confidence"), which is
> not a great combination with naivete.  But I think we should all back
> off a bit until it's done.  In three years :)

Four.  But maybe Jeff can meet deadlines, unlike your other friendly
neighborhood non-humile I-F author.  I put out Underoos on time, didn't
I?  And ZTUU got finished, and let's not forget Avalon 0, the prequel.
Didn't I put that out just like I said I would?

Just seeing if you've read down this far....
 
G. Kevin Wilson: Freelance Writer and Game Designer.  Resumes on demand.



From max@alcyone.com Mon Sep 22 15:06:13 MET DST 1997
Article: 28998 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!news.maxwell.syr.edu!news.he.net!nixon.area.com!not-for-mail
From: Erik Max Francis <max@alcyone.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: Hugo Zork
Date: Sat, 20 Sep 1997 01:07:26 -0700
Organization: Alcyone Systems
Lines: 48
Message-ID: <3423843E.1ED10079@alcyone.com>
References: <3410A9C6.4577@pixi.com> <ant122158b49c4bn@arnod.demon.co.uk> <MPG.e84a8c5b1b6b95f989ee2@news.alt.net> <ant132307868c4bn@arnod.demon.co.uk> <5vgrg3$ikm$1@bartlet.df.lth.se> <341C2107.3836@pixi.com> <874330011.1410@dejanews.com>
NNTP-Posting-Host: newton.alcyone.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01 (X11; I; Linux 2.0.29 i686)
Xref: news.lth.se rec.arts.int-fiction:28998

tessman@cibc.ca wrote:

> As far as the legality of the issue goes (and here I pull out
> the "I am not a lawyer" rubber stamp, too), we're in the territory of
> a
> non-commercial reproduction--and that's probably the right word--of a
> non-commercial piece of work.

It would be copyright infringement, plain and simple.  It clearly
constitutes a derivative work, and as such is protected under copyright.

> Any "damages" would relate to lost revenue
> from Zork itself--and in this case there isn't any--or potential loss
> of
> revenue from future "Zork: ______" games due to John's having so fully
> dragged the good Zork name through the mud--which he didn't, of
> course:
> Hugo Zork is quite excellelntly done.

Nonsense.  There are two types of damages in a copyright case; the one
you are talking about is actual damages, where the copyright holder
calculates how much revenue was lost due to the infringement.  This is
not hard and fast; one could even imagine a good lawyer estimating the
number of users who have a copy, multiplying that by some number to
indicate a cost per unit, and demanding that amount, even though this
wasn't being sold.

Beyond that, however, are statutory damages, which are awardable to the
infringed party in the event that the copyright was registered within
three months of publication, and are awardable up to $100,000.  Yes, you
read that right.

> Zork is obviously a special case; hopefully Activision, if asked,
> would
> recognize it as such.

Why?  If they own the copyright, there's nothing special about it.

The bottom line here is:  Not making a profit off your infringement is
not just a poor defense, it is _no defense_.

-- 
          Erik Max Francis, &tSftDotIotE / mailto:max@alcyone.com
                        Alcyone Systems / http://www.alcyone.com/max/
   San Jose, California, United States / icbm://37.20.07n/121.53.38w
                                      \
                  "War is like love; / it always finds a way."
                                    / Bertolt Brecht


From jcompton@typhoon.xnet.com Mon Sep 22 18:12:34 MET DST 1997
Article: 29107 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!newscore.univie.ac.at!newsfeed.nacamar.de!howland.erols.net!news-peer.sprintlink.net!news-pull.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!198.147.221.37!news.xnet.com!typhoon.xnet.com!not-for-mail
From: jcompton@typhoon.xnet.com (Jason Compton)
Newsgroups: rec.arts.int-fiction
Subject: Re: The problem with the IF Competition
Date: 22 Sep 1997 14:04:46 GMT
Organization: XNet - The Midwest's Leading Network Service Provider - (630) 983-6064
Lines: 23
Message-ID: <605ttu$lcq$2@flood.xnet.com>
References: <60416o$h00$1@news.nyu.edu> <erkyrathEGwBJ3.2q8@netcom.com>
NNTP-Posting-Host: typhoon.xnet.com
X-Trace: flood.xnet.com 874937086 21914 (None) 198.147.221.70
X-Complaints-To: news-admin@news.xnet.com
X-Newsreader: TIN [UNIX 1.3 950824BETA PL0]
Xref: news.lth.se rec.arts.int-fiction:29107

Andrew Plotkin (erkyrath@netcom.com) wrote:
: The competition concentrates effort, but it does it so *well*. If you'd 
: asked me last August whether seventy IF games would be released in 1997, 
: I would have giggled. If you'd asked me two months after that, when the 
: thirty-odd competition games appeared, I would have said "Nahh." Anything 
: which elicits this much creativity from people -- I'm not messing with it!

Well, a statement saying "Yeah, I'll do one" and an entry are two
different things.  My feeling is that some of those people who committed
will not complete an entry.  (And I do believe that part of the reason was
when Wilson told us how many there were--and then listed them.)  Even
without that, there's always some cancellation in any major undertaking.

And you earlier asserted that the competition entries were by definition
not long games, but I seem to remember last year there being a lot of
discussion over games that most certainly required more than two hours of
play time.

-- 
Jason Compton                                jcompton@xnet.com
Editor-in-Chief, Amiga Report Magazine       Anchor, Amiga Legacy
http://www.cucug.org/ar/                     http://www.xnet.com/~jcompton/
There are only dreams...                     ...like any other.


From erkyrath@netcom.com Mon Sep 22 18:46:35 MET DST 1997
Article: 29021 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!news-fra1.dfn.de!news-ber1.dfn.de!news-lei1.dfn.de!news-nue1.dfn.de!wuff.mayn.de!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!howland.erols.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Author's Motives
Message-ID: <erkyrathEGrr32.B9q@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <34206490.21754917@news.demon.co.uk> <fake-mail-1809970245420001@van-52-1024.direct.ca> <erkyrathEGpx5C.13L@netcom.com> <5vsajl$irc$1@stalker.oem.net>
Date: Fri, 19 Sep 1997 18:52:14 GMT
Lines: 46
Sender: erkyrath@netcom14.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29021

Neil deMause (neild@fcl-us.net) wrote:
> Andrew Plotkin (erkyrath@netcom.com) wrote:
> : Neil K. (fake-mail@anti-spam.address) wrote:
> : >  I wish I were a good writer. Like most people, I'm not. But I want to
> : > tell stories. The thing about IF is that you can get away with being a
> : > ho-hum writer and storyteller in ways you couldn't when it comes to
> : > regular fiction.

> : I think this is becoming less and less true, as more IF is released which 
> : has good writing and storytelling. (Without any particular modesty, I 
> : include my works in that group. :)

> Yes and no. IF writing is improving, absolutely. But it's still a genre
> that 1) is smaller

Granted.

> 2) has lower expectations

Only because it hasn't improved enough yet.

> and 3) requires different skills than static fiction.

I've always thought that any skill useful in short-story writing is 
useful in IF writing. I might even argue that the reverse is true (well, 
not *any* skill -- being able to decipher Graham's indentation style is 
probably not very useful for short static fiction.)

Does IF require different skills? No, I don't think so. It's easier to 
cover for lack of experience, because the gaming aspect is an automatic 
hook. 

There's also plain old specialization. If I'd spent as much time writing
short stories in the past 2.5 years as I've spent on IF-related
programming, I'd probably be a whole lot better at story stories. But then
you folks wouldn't have heard of me. As it is, I've gotten in the
equivalent of two stories' writing experience, plus a lot of programming
experience. And I don't spend any time at writers' workshops, or 
whatever it is those folks do to try to improve. So it's slower going. 

--Z

-- 

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


From adamc@acpub.duke.edu Mon Sep 22 18:49:37 MET DST 1997
Article: 28981 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news.kth.se!nntp.uio.no!nntp.uib.no!nntp-bergen.UNINETT.no!uninett.no!sn.no!Norway.EU.net!EU.net!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!newsxfer3.itd.umich.edu!agate!newsgate.duke.edu!usenet
From: Adam Cadre <adamc@acpub.duke.edu>
Newsgroups: rec.arts.int-fiction
Subject: IF and early film (was Author's Motives)
Date: Fri, 19 Sep 1997 21:05:30 -0400
Organization: Group 17
Lines: 68
Message-ID: <3423215A.7EF8@acpub.duke.edu>
References: <34206490.21754917@news.demon.co.uk> <EGq68n.3qB@undergrad.math.uwaterloo.ca>
Reply-To: adamc@acpub.duke.edu
NNTP-Posting-Host: res-9-43.dorm.duke.edu
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01Gold (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:28981

Stephen van Egmond wrote:
> I think it depends on what period the IF you played comes from.
> There are many parallels with film, for instance.

I've long thought the same thing.  I don't think they can help us
predict or even understand anything, but they're fun.

> The early years of film were plagued with gimmickry: a person
> walking, a train travelling, etc.  No story at all.

I'm not sure I'd call this "gimmickry."  The very early cinema was
indeed a cinema of attractions: people wanted to see a moving picture.
The gimmickry comes in with people like Melies, who provided more than
just a moving picture: the Lumiere films were the documents of trains
arriving at stations and such, but Melies shot films in which people
seemed to disappear into thin air, or walk around underwater breathing
comfortably (this accompished by simply placing an aquarium between the
camera and the people being filmed.)  This kind of thing had never been
seen before the way, say, people leaving a factory had.

> The concepts of characterization, theme, and plot were left for books
> and theatre.  Sound like the text adventures of the 70's?

If Adventure had simply been about wandering around a cave, it sounds
like Lumiere; throw in XYZZY, and you've got Melies.

> Once film moved beyond the gimmick stage, many productions were plays,
> but filmed.  A stationary camera with an actors' stage, with the
> script more or less unchanged.

True.  There was also a more camera-conscious type of film which
remained concerned with movement more than narrative: in the 1903
production of Uncle Tom's Cabin, for instance, any kind of narrative
thread is discarded in favor of scenes of slaves dancing, with
everyone's arms and legs flailing wildly about for no reason other
than that movement is cool.

> This was thankfully brief in film, and as far as I can tell, IF never
> went through this -- though Infocomics come to mind.
> 
> Some early film pioneers (whose names escape me) finally brought
> storytelling to film, in a way that only film could do.

DW Griffith is probably the most important of those names.

> They used the camera's frame as a part of the story: showing you some
> things, ignoring others.

Film editing was the big advance here: it allowed montage, the close-up,
cross-cutting, all sorts of things the stage couldn't do.

> I haven't looked at a chronology of Infocom's releases, but it is in
> their time that the story-telling and characterization aspect of IF
> began to be explored.  AMFV and Trinity come to mind.

That seems to be where we still are.  You might want to compare the
rise of graphical games with that of talkies, which drastically changed
the direction of film -- just as silent films more or less disappeared
after the advent of sync sound, graphical games wiped out text
adventures, and the only people making silent films were people with
home movie cameras.  But I'm just making connections for the sake of
making connections here.  Like I said, it's fun to think about, but I'm
not sure there's much to it.

  -----
 Adam Cadre, Durham, NC
 http://www.duke.edu/~adamc
 http://www.retina.net/~grignr


From neild@fcl-us.net Tue Sep 23 15:11:11 MET DST 1997
Article: 29056 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news.kth.se!nntp.uio.no!news.uoregon.edu!newsxfer3.itd.umich.edu!newsfeed.internetmci.com!192.48.96.123!in1.uu.net!208.192.192.27!stalker.oem.net!retina.net!neild
From: neild@fcl-us.net (Neil deMause)
Newsgroups: rec.arts.int-fiction
Subject: Re: Author's Motives
Date: 21 Sep 1997 04:28:34 GMT
Organization: I need to put my ORGANIZATION here.
Lines: 32
Message-ID: <6027pi$k2m$1@stalker.oem.net>
References: <34206490.21754917@news.demon.co.uk> <fake-mail-1809970245420001@van-52-1024.direct.ca> <erkyrathEGpx5C.13L@netcom.com> <5vsajl$irc$1@stalker.oem.net> <erkyrathEGrr32.B9q@netcom.com>
NNTP-Posting-Host: ns1.fcl-us.net
X-Newsreader: TIN [version 1.2 PL2]
Xref: news.lth.se rec.arts.int-fiction:29056

Andrew Plotkin (erkyrath@netcom.com) wrote:
: Neil deMause (neild@fcl-us.net) wrote:

: > and 3) requires different skills than static fiction.

: I've always thought that any skill useful in short-story writing is 
: useful in IF writing. I might even argue that the reverse is true (well, 
: not *any* skill -- being able to decipher Graham's indentation style is 
: probably not very useful for short static fiction.)

: Does IF require different skills? No, I don't think so. It's easier to 
: cover for lack of experience, because the gaming aspect is an automatic 
: hook. 

I didn't mean different *levels* of skills, though that may also be true
for the time being. I meant different skills. In learning how to write I-F
I've had to learn how to write very concise descriptions of places and
objects that won't be boring if you read them over and over, how to write
plots that will make sense no matter what order you read them in, how to
create dialog that sounds natural even when the player keeps interjecting
nonsense like ASK WAITER ABOUT BOSNIA-HERCEGOVNIA, and so on.

These are not skills I would need in writing static fiction, just like in
writing I-F I don't particularly need the ability to create a well-paced
plot, motivate my main character to move on to new things, or write lots
of transitional prose, because those things aren't as important for the
author to provide in I-F.

*Different* skills. And seeing that I feel like I'm better at the former
than the latter, I stick with I-F writing. (For fiction, anyway.)

Neil


From erkyrath@netcom.com Tue Sep 23 15:18:49 MET DST 1997
Article: 29059 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Author's Motives
Message-ID: <erkyrathEGvCE7.FK@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <34206490.21754917@news.demon.co.uk> <fake-mail-1809970245420001@van-52-1024.direct.ca> <erkyrathEGpx5C.13L@netcom.com> <5vsajl$irc$1@stalker.oem.net> <erkyrathEGrr32.B9q@netcom.com> <6027pi$k2m$1@stalker.oem.net>
Date: Sun, 21 Sep 1997 17:25:19 GMT
Lines: 58
Sender: erkyrath@netcom21.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29059

Neil deMause (neild@fcl-us.net) wrote:
> Andrew Plotkin (erkyrath@netcom.com) wrote:
> : Neil deMause (neild@fcl-us.net) wrote:

> : > and 3) requires different skills than static fiction.

> : I've always thought that any skill useful in short-story writing is 
> : useful in IF writing. I might even argue that the reverse is true (well, 
> : not *any* skill -- being able to decipher Graham's indentation style is 
> : probably not very useful for short static fiction.)

> : Does IF require different skills? No, I don't think so. It's easier to 
> : cover for lack of experience, because the gaming aspect is an automatic 
> : hook. 

> I didn't mean different *levels* of skills, though that may also be true
> for the time being. I meant different skills.

I also meant different skills. I'm not sure how to rephrase what I meant 
so that it's clearer.

> In learning how to write I-F
> I've had to learn how to write very concise descriptions of places and
> objects that won't be boring if you read them over and over, how to write
> plots that will make sense no matter what order you read them in, how to
> create dialog that sounds natural even when the player keeps interjecting
> nonsense like ASK WAITER ABOUT BOSNIA-HERCEGOVNIA, and so on.

Mmm, I guess I'm thinking of a different definition of "skills". The 
stuff you talk about, I think of as more like "subskills" -- meaning the 
sort of task which *goes along* with general writing skills. 

For example, being able to write a concise description is desperately
important in any kind of writing. Being able to write an IF room
description does have other constraints. As you say, it will be read more
than once; it can't have actions or thoughts in it ("You trip over the
carpet as you enter" -- the Barringer Mistake. :-) But it's the sort of 
thing you can figure out once you *can* write a description. It's not a 
stumper, it's a new application of your skill.

> These are not skills I would need in writing static fiction, just like in
> writing I-F I don't particularly need the ability to create a well-paced
> plot, motivate my main character to move on to new things, or write lots
> of transitional prose, because those things aren't as important for the
> author to provide in I-F.

Ooo, see, that's exactly what I mean. I think those three things are 
equally important in static and interactive fiction. I think we haven't 
seen them as brilliantly applied as they are in the best novels, which is 
what I meant by "lower expectation". And yes, the mechanisms for (say) 
plot pacing in IF are different. But I consider it the same skill.

--Z

-- 

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


From erkyrath@netcom.com Wed Sep 24 13:16:03 MET DST 1997
Article: 29194 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!newscore.univie.ac.at!newsfeed.ecrc.net!.NOMATCH!europa.clark.net!205.252.116.205!howland.erols.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Concept -> Story -> Game
Message-ID: <erkyrathEGz3BB.6Hq@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <606914$fe3@eve.enteract.com> <erkyrathEGy3rK.Dx9@netcom.com> <608k1b$onp@eve.enteract.com>
Date: Tue, 23 Sep 1997 17:59:35 GMT
Lines: 40
Sender: erkyrath@netcom2.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29194

NoJunkMail_sharvey@spamfree.enteract.com wrote:

> : If you're starting with a story, and your intent is for the player to 
> : witness it, then by definition you're shoving the plot in the player's 
> : face. The original IF trick was to make the player think that *he's* 
> : digging up the plot and winkling it out of *you*, or rather out of the 
> : universe. This is a fine old classic trick and I use it all the time.


> (*WARNING: Enchanter spoilers below*)

> I suppose what's confounding me is this: Enchanter, being my fondest
> memory to date of any work of IF, really didn't have much of a story. 

This is true. It's true of a lot of early games. It's also true of some 
recent games.

> I perceive that "modern" IF is striving to be more than an exploration of
> a setting and puzzle-solving, while discovering a relatively simple plot.
> I'm reading discussions of character development, morality, philosophy -
> concepts much more advanced than just making a fun game.

There is no "modern" IF. There's me, and Graham Nelson, and Gareth Rees,
and so on and so on. I can tell you what *I'm* striving to do -- although
I'd rather show you -- but the only overall movement these days is to try
everything possible. 

> What's more important to you, as an IF writer: writing an important and
> well-done piece of fiction that people interact with, or writing a
> literate and entertaining diversion?

I can't separate those two categories, since to me a well-done piece of 
fiction *is* a diversion. That's what I read for.

--Z

-- 

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


From svanegmo@undergrad.math.uwaterloo.ca Wed Sep 24 13:21:15 MET DST 1997
Article: 29244 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!feed1.news.erols.com!newsfeed.internetmci.com!199.60.229.5!newsfeed.direct.ca!logbridge.uoregon.edu!news.bc.net!torn!kwon!watserv3.uwaterloo.ca!undergrad.math.uwaterloo.ca!svanegmo
From: svanegmo@undergrad.math.uwaterloo.ca (Stephen van Egmond)
Subject: Re: IF and early film (was Author's Motives)
Sender: news@undergrad.math.uwaterloo.ca (news spool owner)
Message-ID: <EGxCuz.nGL@undergrad.math.uwaterloo.ca>
Date: Mon, 22 Sep 1997 19:30:35 GMT
References: <34206490.21754917@news.demon.co.uk> <EGq68n.3qB@undergrad.math.uwaterloo.ca> <3423215A.7EF8@acpub.duke.edu>
Nntp-Posting-Host: cantor.math.uwaterloo.ca
Organization: University of Waterloo
Lines: 43
Xref: news.lth.se rec.arts.int-fiction:29244

Adam Cadre  <adamc@acpub.duke.edu> wrote:
>I've long thought the same thing.  I don't think they can help us
>predict or even understand anything, but they're fun.

I think they are useful for predicting trends, and understanding media is 
certainly possible, and in fact it is nice to see a few of McLuhan's 
theories in operation here.


>DW Griffith is probably the most important of those names.

That's the one.

>That seems to be where we still are.  You might want to compare the
>rise of graphical games with that of talkies, which drastically changed
>the direction of film -- just as silent films more or less disappeared
>after the advent of sync sound, graphical games wiped out text
>adventures, and the only people making silent films were people with
>home movie cameras.  But I'm just making connections for the sake of
>making connections here.  Like I said, it's fun to think about, but I'm
>not sure there's much to it.

Possibly.  Consider the device in Hitch Hiker's in which the game itself 
gives you a hard time about entering the Improbability Drive chamber, 
that is the sort of effect which I think we're going to see more of, 
where people experiment with the medium itself, changing its nature.

Analogy: Kurt Vonnegut, in _Bluebeard_, used the technique of marking with an
asterisk (*) the names of any people who were going to be dying shortly.  
He claimed, in the meta-prose he broke into after the first asterisk, 
that this was to eliminate the whole business of suspense over who was 
going to be dying and who was going to live.  It was surprising how it 
changed the story.

FWIW, my current idea (and it is little more than that) is to develop a 
story where the computer is outright unreliable and may, from time to 
time, lie to you in various ways.  I think of it as a parallel for the 
way our minds can work, highlighting some input and ignoring others.

It's like the Unreliable Narrator in film, but, I hope, will turn out to 
be more personal in its effect.

/Steve


From erkyrath@netcom.com Wed Sep 24 19:22:39 MET DST 1997
Article: 29269 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!news-ge.switch.ch!newscore.univie.ac.at!newsfeed.ecrc.net!.NOMATCH!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!wnfeed!204.127.130.5!worldnet.att.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: To whom it may concern
Message-ID: <erkyrathEH0uon.1BJ@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <ant2413140b0c4bn@arnod.demon.co.uk> <slrn62iann.eu7.spatula@kona.javanet.com>
Date: Wed, 24 Sep 1997 16:48:22 GMT
Lines: 25
Sender: erkyrath@netcom23.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29269

tv's Spatch (spatula@javanet.com.andbacon) wrote:
>  Julian Arnold: 
> > In time-honoured tradition I've withdrawn my name from the competition.
> > Ho-hum.

> I'd like at this moment to state as well that I, too, am withdrawing my
> entry from the competition.  The fact that I didn't have an entry this
> year doesn't stop me from hopping on the bandwagon and withdrawing it as
> well.

Heh.

Anyway, for all of you that *planned* to enter but will not be able to, I 
hope you follow Magnus's lead and keep working on your game. Whenever it 
appears, it will be played.

We may not have 71 games appearing in the competition, but I'll bet we 
get pretty close to 71 games released in the next six months.

--Z

-- 

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


From graham@gnelson.demon.co.uk Wed Sep 24 20:57:10 MET DST 1997
Article: 29277 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter0!news.primenet.com!nntp.primenet.com!newsfeed.nacamar.de!newscore.univie.ac.at!news-ge.switch.ch!news-zh.switch.ch!surfnet.nl!newsgate.unisource.nl!bullseye.news.demon.net!demon!dispose.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Parsing, then and now (Was: Re: Using Smalltalk for IF (also space habitats and sound adventures))
Date: Wed, 24 Sep 1997 11:41:47 +0100 (BST)
Organization: none
Message-ID: <ant2410470b0M+4%@gnelson.demon.co.uk>
References: <34274C61.3476@netins.net> <608kkv$8pn@decgate.bridgewater.ne.hcc.com> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 78
Xref: news.lth.se rec.arts.int-fiction:29277

In article <608kkv$8pn@decgate.bridgewater.ne.hcc.com>,
<URL:mailto:jkennedy@himail.hcc.com> wrote:
> The Inform library has faults here and there, but by now the parser
> is almost as sophisticated as Infocom's V1-V5 parsers...

  Begging your pardon, m'lud, but it's vastly more sophisticated
than the Infocom parser was: Infocom hand-wrote a different version
of the parser for every single game, in a really hacky way.  There's
no philosophy of "here's a library to use" about it: the Implementors
more or less tore the parser code out of Zork I and stuck on
whatever features they felt like.  In the case of, e.g., "Spellbreaker"
this took a lot of messing about.  Of course partly this indulged
the passion felt by some of the Implementors for writing better
and better parsers from scratch.  I forget which one of them claimed
to have written something like 50?

  Name one thing the Infocom parser (in so far as there is one
single Infocom parser) could handle that the standard, unenhanced,
unprogrammed Inform parser can't!  And I'll probably be glad to add
it.  I don't believe any of the Infocom games contain proper
parsing of plurals ("get three of the toads"), groups of
indistinguishable objects, etc.: and most of the nicer features, like
"undo", are only present in the later games.  Here's the final
release of Zork II:

    Inside the Barrow
    You are inside an ancient barrow hidden deep within a dark forest. The
    barrow opens into a narrow tunnel at its southern end. You can see
    a faint glow at the far end.
    A strangely familiar brass lantern is lying on the ground.
    A sword of Elvish workmanship is on the ground.

    >x it
    I don't know the word "x".

    >examine it
    There's nothing special about the .

    >pronouns
    I don't know the word "pronouns".

    >oops sword
    I don't know the word "oops".

    >dragon sword
    There was no verb in that sentence!

    [Note that it doesn't respond "I don't know the word "dragon".",
    so it has inadvertently revealed the existence of a dragon to us.
    Likewise, trying "get gold" and then "get silver" demonstrate
    that the game contains the former and not the latter.]

Sophisticated stuff!  And compare that to

    You can see a plastic fork, knife and spoon, three hats (a fez, a
    Panama and a sombrero), the letters X, Y, Z, P, Q and R from a
    Scrabble set, a defrosting Black Forest gateau, Punch magazine,
    a recent issue of the Spectator, a die and eight stars (four silver,
    one bronze and three gold) here.

    >get three silver stars
    silver star: Taken.
    silver star: Taken.
    silver star: Taken.

    >get a hat
    (the fez)
    Taken.

We shouldn't allow the golden memory of Infocom to blind our critical
faculties altogether.  Their work was brilliant and pioneering, but
(on the implementation side at least) it is no longer the state of
the art, and their early- and middle-period parsers were looking
quite rickety as early as 1988 or so.

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From erkyrath@netcom.com Fri Sep 26 13:29:17 MET DST 1997
Article: 29378 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: A digression on gems
Message-ID: <erkyrathEH3rMJ.935@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
Date: Fri, 26 Sep 1997 06:35:07 GMT
Lines: 57
Sender: erkyrath@netcom8.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29378

This has some sort of IF relevance, but I'm not going to let that stop me.

I went to the Smithsonian Natural History Museum on Monday, and they've 
just opened their new Hall of Neat Rocks. (That's not what they call it. 
Pity, though. If *I* were putting the Hope Diamond on display, I'd 
definitely call the room the Hall of Neat Rocks.)

Anyway, wandering through this room full of immensely valuable gems, I 
started thinking that this *is* the payoff of Colossal Cave and Zork 1. 
When a game describes an "emerald the size of a plover's egg", it means 
something like *this*. 

Well, once again I've forgotten how big a plover's egg is. But the museum
has an emerald about an inch square. And that's *really impressive*. I
mean, it's huge. It's easy to put "a big diamond" in your game, but you
really ought to convey what it's like to be staring at a stone which was 
stolen from a French king, passed around the black market, traded, 
bought, and generally lusted after for *four hundred years*. 

It's the Monty Haul problem from D&D; the player grows calluses where his
sense of wonder should be. Because whenever anyone wants four treasures,
they put in "a big diamond", "a big ruby", "a big emerald", and "a big
sapphire". 

So what I'm saying is -- be gentle. A diamond the size of your smallest 
fingernail is pretty damn stunning. Make sure the player knows it.

...Actually, the museum does have a faceted topaz the size of an *emu*
egg.  That's a stone that passes the Flathead test: "If you dropped this
gem on your foot, would you yell 'ow'?" 

And of course the crystal ball. A ten-inch sphere of flawless clear
quartz. When you think of crystal balls, think of this. I think it's the 
biggest (flawless) one in existence.

I really, really wish I could bring my pen-laser in there and just shine
it around for a few minutes. I mean, without being hauled away and 
strip-searched by the guards.

Oh well.

While I'm at it, I shouldn't forget all the other mineral samples they
have on display. It's great, just as an example of what can grow naturally
in a cave. Many science museums have similar displays; the one in the
Carnegie Museum in Pittsburgh (near Pitt U.) is particularly nice.  Next
time you want to write a cave crawl game, go check out the one nearest
you. 

--Z




-- 

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


From graham@gnelson.demon.co.uk Sat Sep 27 09:34:58 MET DST 1997
Article: 29418 of rec.arts.int-fiction
Path: news.lth.se!c-1996!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!news-peer.gsl.net!news.gsl.net!gip.net!news.idt.net!dispose.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: Off-topic optics rambling (was Re: A digression on gems)
Date: Fri, 26 Sep 1997 19:30:33 +0100 (BST)
Organization: none
Message-ID: <ant261833d07M+4%@gnelson.demon.co.uk>
References: <erkyrathEH3rMJ.935@netcom.com> <342B5C34.1B477490@alcyone.com> <Pine.SUN.3.91.970926081754.24986B-100000@nebula.phy.duke.edu> <erkyrathEH4K27.3v9@netcom.com> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 19
Xref: news.lth.se rec.arts.int-fiction:29418

In article <erkyrathEH4K27.3v9@netcom.com>, Andrew Plotkin
<URL:mailto:erkyrath@netcom.com> wrote:
>
> Everyone should get a laser.

Whose signature is it that collects absurd sentences written on
rec.arts.int-fiction?

("Don't get a life, get a laser... hours of fun with total internal
reflection... and that's not all!  You also get thin film
interference ABSOLUTELY FREE!  Worried what to do with all those
dull workaday household parabolic mirrors, sugar-filled aquaria
and dry ice machines?  Well worry no more.  Have we got an excited
ionisation trail for you!  Yes, it's the noble gas that you don't
have to be noble to pass electric currents through! ...")

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From daryl@cogentex.com Sat Sep 27 09:35:14 MET DST 1997
Article: 29408 of rec.arts.int-fiction
Path: news.lth.se!c-1996!www.nntp.primenet.com!globalcenter0!news.primenet.com!nntp.primenet.com!newsfeed.internetmci.com!205.185.79.5!zdc!szdc!newsp.zippo.com!zdrn
From: daryl@cogentex.com (Daryl McCullough)
Newsgroups: rec.arts.int-fiction
Subject: Re: A digression on gems (and caves)
Date: 26 Sep 1997 10:42:30 -0700
Organization: CoGenTex, Inc.
Lines: 20
Message-ID: <60gs66$fv1@drn.zippo.com>
References: <erkyrathEH3rMJ.935@netcom.com> <342BADBD.7D00@iquest.net.nospam>
NNTP-Posting-Host: mali.cogentex.com
Xref: news.lth.se rec.arts.int-fiction:29408

In article <342BADBD.7D00@iquest.net.nospam>, Carl says...

>Some of the most impressive caves in the world, including Mammoth Cave,
>are just a few hours south of here.  I'm trying to get some friends and
>family together to go check them out next month, but I think it may end
>up being too late in the year.  I haven't been down there since I was
>just a little kid.  I'd really like to see if I can recognize any bits
>of Mammoth Cave that were used in Adventure, but I don't know if you can
>get to those places without serious spelunking.  There's even an
>underground boat ride through one set of caves, which I think sounds
>like a lot of fun.  Anyway, it seems to me like a trip like this would
>also be quite inspirational.

I heard that the underground boat rides were cancelled, because the lights
of the boats were harmful to the species of water creatures that evolved
(only in Mammoth Cave) to live in absolute darkness.

Daryl McCullough
CoGenTex, Inc.
Ithaca, NY


From sothoth@xxx.usa.net Sat Sep 27 17:38:39 MET DST 1997
Article: 29035 of rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news01.sunet.se!sunic!mn6.swip.net!nntp.uio.no!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.ecrc.net!newscore.univie.ac.at!03-newsfeed.univie.ac.at!alijku02!not-for-mail
From: "Gunther Schmidl" <sothoth@xxx.usa.net>
Newsgroups: rec.arts.int-fiction
Subject: Re: Darkness
Date: 21 Sep 1997 09:14:59 GMT
Organization: privat
Lines: 30
Message-ID: <01bcc5a3$95fda360$LocalHost@async.edvz.uni-linz.ac.at>
References: <34206D22.9AFD0694@uiuc.edu> <34209CBF.4A2E@bigfoot.com> <5vqm37$99d$1@bartlet.df.lth.se>
NNTP-Posting-Host: async13.edvz.uni-linz.ac.at
X-Newsreader: Microsoft Internet News 4.70.1161
Xref: news.lth.se rec.arts.int-fiction:29035

> "Aayela", BTW, was written in TADS, which has a *much* more flexible
> way of handling darkness than Inform has. I was going to write an
> article about the implementation of darkness in "Aayela" for
> XYZZYnews; however, due to the negative reception of the game, I lost
> interest in it and everything that had to do with it. (Yes, I know
> that one or two people liked "Aayela", but most people seem to have
> regarded the darkness as nothing special, or just an annoyance.)

What? What? 
Heck, Aayela was one of the best games I played this year; interestingly
though, I thought of putting just such a kind of thing into my game
_before_ I played it - just to have the player use more senses (mine's
really a dark and moody game) and I must say that I was *very* reluctant to
even enter the cave without a working lantern, but then was amazed at the
skill with which you incorporated darkness into your game without the rooms
being really "dark" in the usual style ("...pitch blank...grue...yadda
yadda...")
-- 

+------------------------+----------------------------------------------+
+ Gunther Schmidl        + "I couldn't help it. I can resist everything +
+ Ferd.-Markl-Str. 39/16 +  except temptation"           -- Oscar Wilde +
+ A-4040 LINZ            +----------------------------------------------+
+ Tel: 0732 25 28 57     + http://oberon.home.ml.org - german homepage! +
+------------------------+---+------------------------------------------+
+ sothoth (at) usa (dot) net + please remove the "xxx." before replying +
+----------------------------+------------------------------------------+





From cklutzke@iquest.net.nospam Sat Sep 27 17:38:50 MET DST 1997
Article: 29322 of rec.arts.int-fiction
Path: news.lth.se!solace!mn6.swip.net!nntp.uio.no!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.internetmci.com!4.1.16.34!cpk-news-hub1.bbnplanet.com!cpk-news-feed4.bbnplanet.com!cpk-news-feed1.bbnplanet.com!news.bbnplanet.com!news.iquest.net!not-for-mail
From: Carl Klutzke <cklutzke@iquest.net.nospam>
Newsgroups: rec.arts.int-fiction
Subject: Re: Darkness
Date: Wed, 24 Sep 1997 22:26:22 -0500
Organization: IQuest Internet, Inc.
Lines: 27
Message-ID: <3429D9DE.6D91@iquest.net.nospam>
References: <34206D22.9AFD0694@uiuc.edu> <34209CBF.4A2E@bigfoot.com> <5vqm37$99d$1@bartlet.df.lth.se> <01bcc5a3$95fda360$LocalHost@async.edvz.uni-linz.ac.at>
Reply-To: cklutzke@iquest.net.nospam
NNTP-Posting-Host: laf-0000-23.iquest.net
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.0Gold (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:29322

Gunther Schmidl wrote:
> 
> > "Aayela", BTW, was written in TADS, which has a *much* more flexible
> > way of handling darkness than Inform has. I was going to write an
> > article about the implementation of darkness in "Aayela" for
> > XYZZYnews; however, due to the negative reception of the game, I lost
> > interest in it and everything that had to do with it. (Yes, I know
> > that one or two people liked "Aayela", but most people seem to have
> > regarded the darkness as nothing special, or just an annoyance.)
> 
> What? What?
> Heck, Aayela was one of the best games I played this year; interestingly
> though, I thought of putting just such a kind of thing into my game
> _before_ I played it - just to have the player use more senses (mine's
> really a dark and moody game) and I must say that I was *very* reluctant to
> even enter the cave without a working lantern, but then was amazed at the
> skill with which you incorporated darkness into your game without the rooms
> being really "dark" in the usual style ("...pitch blank...grue...yadda
> yadda...")

I also was quite impressed with Aayela.  I've never had a game make me
feel
quite so claustrophobic before.  Playing Aayela was one of the things
that
made me decide to put more senses in my game than just sight.

Carl


From graham@gnelson.demon.co.uk Sun Sep 28 10:56:56 MET DST 1997
Article: 29496 of rec.arts.int-fiction
Path: news.lth.se!c-1996!feed1.news.erols.com!howland.erols.net!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!dispose.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: A digression on gems
Date: Sun, 28 Sep 1997 00:24:08 +0100 (BST)
Organization: none
Message-ID: <ant2723081cbM+4%@gnelson.demon.co.uk>
References: <erkyrathEH3rMJ.935@netcom.com> <60j3md$jeq$1@bartlet.df.lth.se> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 19
Xref: news.lth.se rec.arts.int-fiction:29496

In article <60j3md$jeq$1@bartlet.df.lth.se>, Magnus Olsson
<URL:mailto:mol@bartlet.df.lth.se> wrote:
> 
> This reminds me of another treasure in "Advent": the "platinum
> pyramid, eight inches to a side". If that thing were solid, it would
> weigh almost forty kilograms, and be worth a million dollars or so.

Somebody did a nice piece in the newspapers, one Christmas, about
how the relative worths of gold, frankincense and myrhh have done
since Biblical times.  Gold is still the one to have, of course,
but frankincense (still used in perfumes) is quite dear, rather
the way saffron is: myrhh, though, costs almost too little to
bother weighing it out.  It's mass-produced and used in salves
and balms for mouth ulcers, a bit of a come-down from anointing
kings.

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From mol@bartlet.df.lth.se Sun Sep 28 14:28:59 MET DST 1997
Article: 29508 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: A good book ->Re: z-code interpreter for t.v. remote control
Date: 28 Sep 1997 14:23:40 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 57
Message-ID: <60li8c$7ag$1@bartlet.df.lth.se>
References: <19970925172001.NAA24423@ladder02.news.aol.com> <17BF4D883S85.I1700004@VM.SC.EDU> <342C5BB4.5B6C7E4D@ix.netcom.com> <17BF513603S85.I1700004@VM.SC.EDU>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:29508

In article <17BF513603S85.I1700004@VM.SC.EDU>, Phase <look@my.sig> wrote:
>Russell Glasser <rglasser@ix.netcom.com> writes:
>>        Even a simple game of chess contains many more possible positions than
>>there are estimated atoms in the universe; you honestly think Zork would
>>be smaller or easier?
> 
>Dunno, I don't know how large and complex Zork is.  And while chess
>may have an incomprehensible number of possible moves, there are
>only so many configurations the board can be in at any one instance,
>right?  Or is this number comparable?  A state is a state, you don't
>have to remember how you got there.

That reduces the complexity somewhat, but not enough. Lets see: you
have 32 pieces and 64 squares to put them on. That makes for
64*63*...*32 = 64!/32! states if all pieces were distinguishable, but
they aren't, which reduces the number of states by a further factor of
(8! 2!^3)^2 (8! for the pawns of each colour, 2! for the rooks,
bishops and knights, and square it all for the two colours.

This still leaves us with 4634726695587809641192045982323285670400000
states.  Many of these are illegal (for example, the kings may not
stand next to each other, there may not be any pawns on the last row,
and so on), but that still doesn't reduce the complexity to manageable
proportions.

>>        So I'm not just saying it's not practical (which is obvious); it
>>couldn't be done no matter how good the technology or how bored the
>>programmer.
> 
>Oh I knew it wasn't practical, but I still thought it could be done.
>You don't need a page (or several) for each state, btw, you could rely
>on the player to keep track of some states.

I don't think you realize just how complex a game like "Zork" is, in
the terms of states. For example, take a hypothetical game with just
ten locations and ten objects. If any object can be in any location,
then just that gives you 10^10 (ten billion) states. Large IF games
tend to have hundreds of locations and scores of objects, many of
which have different "intrinsic" states (closed/open, on/off, etc).

You *could*, of course, require the player to keep track of the state
and location of every item he's found, and then you're beginning
to approach a managable number of "external" states, but then your
multiple-choice questions would become impossibly detailed:

"If the platinum pyramid is currently in a dark room, and your
flashlight is on, and you're wearing the red cap, and the dilithium
crystals are in a closed container, and the Mad Professor is on
the fifth floor of the building, and the moon is waxing, then turn to
page 712316".



-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From mol@bartlet.df.lth.se Sun Sep 28 15:17:26 MET DST 1997
Article: 29509 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: Physical traversal of landscape
Date: 28 Sep 1997 14:49:56 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 83
Message-ID: <60ljpk$9nr$1@bartlet.df.lth.se>
References: <5vcmfk$sio$1@charm.magnus.acs.ohio-state.edu> <3420B6DF.4735093B@ix.netcom.com> <erkyrathEGpxot.1nD@netcom.com> <5vsc39$ila$1@nntp5.u.washington.edu>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:29509

In article <5vsc39$ila$1@nntp5.u.washington.edu>,
Dan Shiovitz <scythe@u.washington.edu> wrote:
>
>In article <erkyrathEGpxot.1nD@netcom.com>,
>Andrew Plotkin <erkyrath@netcom.com> wrote:
>[..]
>>The "standard IF room" was *never* a concession to crude or simplistic 
>>programming. 
>
>I disagree.

Hmmm. I was about to post that I agreed with everything Andrew had
written, and that I couldn't have expressed it better myself, etc,
but I think I see Dan's point as well. To me, it seems like you disagree
because you're discussing different things.

To start with, the *origin* of the "standard IF room". Having read all
the interesting "history of _Colossal Cave_" docs, I think it's pretty
clear that Crowther's made a conscious choice of world model. Cavers
seem to think of caves as "rooms" connected by twisty passages, not
just by convention, but because real caves often look that way.

And "Colossal Cave" wasn't written in the stone age - while computer
modelling has certainly advanced a lot since then, there were surely
other ways of modelling a cave available in 1972. A coordinate based
grid, for example, would have been feasible on the PDP-10 (which was
quite a large computer with virtual memory, so memory constraints
wouldn't have been much of a problem).

(I can see a disturbing tendency, BTW, of people who, wanting to
change the IF conventions in one way or another, claim that the old
conventions are due to incompetence or lack of imagination, and that
present-day IF authors cling to them out of sheer conservativism. It's
a facile way of making your proposed "paradigm shift" look more
attractive, but in most cases it's simply not true, and an insult to
boot.)

But of course Dan is right in the sense that, like all conventions,
the discrete-rooms-connected-by-pathways has been misapplied to cases
where it's less than appropriate. In some cases this may be due to
sloppy design (sloppy programming is something different: that's what
leads to bugs :-) ). In most of the cases, however, I think it's a
case of not knowing the limits of your tools. 

The Inform library by itself is, for example, not very suitable for
handling rooms that consist of many locations where you can see
objects in the other locations (if you're in the west part of the
living room, shouldn't you be able to see everything in the east part
as well?). In such a case, you either have to extend your tools, or
lower your ambitions and change the design.

Another case: The AGT game "Mop and Murder" takes place almost
entirely in one room, the office of a CIA agent. However, when I
played it, I got the feeling that the room really ought to have been
divided into sub-locations. One puzzle concerned a bookshelf; when I
was trying to solve that, I mentally placed myself in front of the
bookshelf, and was just distracted by the room description describing
everything else. In short, the game state didn't reflect what I was
doing to a sufficient degree.

Another shortcoming of the current convention is that there is no
concept of what is behind and in front of the player. A room
description sort of supposes that you look around you and note
everything you see. This is OK in 99% of the cases, but suppose you
have a puzzle involving someone sneaking up behind you? How do you
implement "look behind me" in a consistent way, *and* make your world
modelling acknowledge the fact that you 

1) are always facing one way or another

and

2) can't see things behind your back?

And remember that this modelling has to be consistent, or the player
will be confused: you can't cheat and just have the notion of "behind
me" when a single puzzle is concerned.


-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From graham@gnelson.demon.co.uk Sun Sep 28 18:29:30 MET DST 1997
Article: 29515 of rec.arts.int-fiction
Path: news.lth.se!c-1996!news-stkh.gip.net!news.gsl.net!gip.net!gsl-penn-ns.gsl.net!news-dc.gsl.net!news-peer.gsl.net!news.gsl.net!gip.net!howland.erols.net!newsfeed.internetmci.com!194.72.7.126!news-peer.bt.net!btnet!dispose.news.demon.net!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: What is an Acorn?
Date: Sun, 28 Sep 1997 12:24:39 +0100 (BST)
Organization: none
Message-ID: <ant281139b49M+4%@gnelson.demon.co.uk>
References: <3415BFF0.4346@juno.com> <19970912094801.FAA13352@ladder02.news.aol.com> <ant122121868c4bn@arnod.demon.co.uk> <60hivn$35d$1@prometheus.acsu.buffalo.edu> <fake-mail-2709970128420001@van-52-0303.direct.ca> <342D54D2.1EDA189@alcyone.com> <fake-mail-2709971255510001@van-52-0236.direct.ca> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 20
Xref: news.lth.se rec.arts.int-fiction:29515

In article <fake-mail-2709971255510001@van-52-0236.direct.ca>, Neil K.
<URL:mailto:fake-mail@anti-spam.address> wrote:
>  - Graham, as Inform's creator, has been quietly evangelical about
> promoting Inform with Acorn users. (eg: including Inform and Inform games
> on Acorn CD collections, etc.)

Quietly?  QUIETLY?  Inform 6.14 is on the Acorn User cover CD
for December, incidentally, complete with HTML Designer's Manual.

In fact, more is true: Inform would not exist at all if TADS had
been ported to the Acorn (at that time, 1993, it hadn't been and
source code was somewhat restricted-access, as an entirely
understandable measure of caution on the part of TADS's authors --
let's not start _that_ row again).  I had to write Inform in order
to write Curses, and I had to write Curses to make good use of
Inform, and...

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From crs1583@inforamp.net Mon Sep 29 09:44:06 MET DST 1997
Article: 29530 of rec.arts.int-fiction
Path: news.lth.se!c-1996!feed1.news.erols.com!newsfeed.internetmci.com!204.191.217.2!News1.Toronto.iSTAR.net!news.istar.net!news1.istar.ca!not-for-mail
From: crs1583@inforamp.net (Gord Jeoffroy)
Newsgroups: rec.arts.int-fiction
Subject: Re: What is an Acorn?
Date: Mon, 29 Sep 1997 04:24:21 GMT
Organization: iSTAR Internet Incorporated
Lines: 18
Message-ID: <34352ccf.4383277@news.istar.ca>
References: <3415BFF0.4346@juno.com> <19970912094801.FAA13352@ladder02.news.aol.com> <ant122121868c4bn@arnod.demon.co.uk> <60hivn$35d$1@prometheus.acsu.buffalo.edu>
NNTP-Posting-Host: ts14-11.tor.istar.ca
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Newsreader: Forte Agent 1.01/32.397
Xref: news.lth.se rec.arts.int-fiction:29530

On 27 Sep 1997 00:11:35 GMT, goetz@cs.buffalo.edu (Phil Goetz) wrote:

>I assumed Acorns were some old, obsolete, and dying
>computer from the early 1980's.

I'm normally loath to contribute to off-topic threads but....

One of my first computers was an Acorn Atom (yes, back in the early
80s). I learned more about computers from that machine than from any
other. It was immensely fun and flexible. I even wired up an RS-232
interface for a remote controlled toy robot and had the thing
following hour-long sequences of commands.

If it weren't for Acorn computers, I don't think I'd be the incredibly
talented, fabulously wealthy, eligible bachelor (are you reading,
ladies?) that I am today.

--Gord, waxing nostalgic...


From mol@bartlet.df.lth.se Mon Sep 29 17:29:29 MET DST 1997
Article: 29556 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: Parsing: a new beginning
Date: 29 Sep 1997 17:28:35 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 72
Message-ID: <60ohf3$bdr$1@bartlet.df.lth.se>
References: <ant2902100b0c4bn@arnod.demon.co.uk>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:29556

In article <ant2902100b0c4bn@arnod.demon.co.uk>,
Julian Arnold  <jools@arnod.demon.co.uk> wrote:
>> I wrote something above about "particles" as well. That would be words
>> like the "of" in "tin of cat food". The adventure parser should
>> recognize them as parts of the name, but they should never be
>> recognized on their own (which adjectives may be). Because otherwise
>> we could get strange things like this:
>> 
>> > take of  [ really, a misspelling of "take off" ]
>> 
>> Which of do you want to take, the tin of cat food, the wand of death,
>> or the Akond of Swat?
>
>Good point. A while back I suggested to Kent that it might be useful to
>have a property to contain what I termed compound nouns. For instance:
>  object wand_of_death
>  {
>    nouns "wand", "death"
>    compound_noun "wand of death"
>  }
>
>As Magnus says, the parser should ideally recognize "of" (and like
>words) as part of a name, but never alone. Compound nouns addresses this
>from the opposite side to particle--decreasing resolution rther than
>increasing it (or is it the other way round? :). I believe there are
>some practical problems with compound nouns, but I'm not clear what they
>are.

Ideally, compound nouns would be trivial to add: we'd just make the parser
recognize spaces inside of dictionary words. 

However, it's not as simple as that, at least not with Inform, because
the parser breaks up the input sentence at word limits before starting
the "real" parsing, and spaces count as word limits...

[ Another linguistic note: most Germanic languages tend to use compound
nouns a lot more than English does (perhaps because English is heavily
influenced by the non-Germanic French). For example, the Swedish word
for "cat food" is "kattmat", literally "catfood". ]

My proposal to add particles was based on the desire to avoid spaces
inside dictionary words.

IIRC, TADS has some notion of "particles"; words declared as prepositions
can work in that way. For example, 'of' is a preposition in TADS, and
the way you make a wand of death in TADS is to declare 'wand' as the noun
and 'death' as the adjective, and "wand of death" then is recognized
as a synonym for "death wand" - which also means that "bird of yellow"
is a synonym for "yellow bird". This said with the qualification that my
memory of TADS is somewhat obscured by recent Inform-use.

>This makes a further point, which maybe Magnus hasn't brought up 'cos
>it's obvious, I don't know if it is or not. In Hugo the player's input
>may only contain one noun for each object being referred to, but
>multiple adjectives. Nouns must go after any and all adjectives. For
>these reasons, "tin" should be an adjective in the above example, even
>though it might be thought of as a noun.

Hmm. The reason	I didn't bring this up was that I didn't know of it.
But it sounds rather limited to me. How does Hugo handle synonyms?
I'm talking about true synonyms, where you really want several nouns
to refer to the same object (though probably not at the same time).
Suppose you have an NPC that is 15-year-old Lucy; you'd probably
want to be able to refer to her as 'Lucy', 'girl' or 'woman', all of
which should be parsed as nouns.



-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From jools@arnod.demon.co.uk Tue Sep 30 09:03:11 MET DST 1997
Article: 29586 of rec.arts.int-fiction
Path: news.lth.se!c-1996!news.kth.se!news.stupi.se!newsfeed.ecrc.net!128.230.129.106.MISMATCH!news.maxwell.syr.edu!dispose.news.demon.net!demon!news.demon.co.uk!arnod.demon.co.uk!jools
From: Julian Arnold <jools@arnod.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: Parsing: a new beginning
Date: Mon, 29 Sep 1997 20:05:29 +0100 (BST)
Organization: None, absolutely none
Message-ID: <ant2919291cbc4bn@arnod.demon.co.uk>
References: <ant2902100b0c4bn@arnod.demon.co.uk> <60ohf3$bdr$1@bartlet.df.lth.se> 
NNTP-Posting-Host: arnod.demon.co.uk
X-NNTP-Posting-Host: arnod.demon.co.uk [158.152.151.175]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.29]
Lines: 80
Xref: news.lth.se rec.arts.int-fiction:29586

In article <60ohf3$bdr$1@bartlet.df.lth.se>, Magnus Olsson
<URL:mailto:mol@bartlet.df.lth.se> wrote:
> In article <ant2902100b0c4bn@arnod.demon.co.uk>,
> Julian Arnold  <jools@arnod.demon.co.uk> wrote:
> Ideally, compound nouns would be trivial to add: we'd just make the parser
> recognize spaces inside of dictionary words. 
> 
> However, it's not as simple as that, at least not with Inform, because
> the parser breaks up the input sentence at word limits before starting
> the "real" parsing, and spaces count as word limits...

Hugo also. Maybe some concept of "hard spaces" could be added?

> My proposal to add particles was based on the desire to avoid spaces
> inside dictionary words.

Yep, and as things stand, that would be the practical thing to do, I
think.

> >This makes a further point, which maybe Magnus hasn't brought up 'cos
> >it's obvious, I don't know if it is or not. In Hugo the player's input
> >may only contain one noun for each object being referred to, but
> >multiple adjectives. Nouns must go after any and all adjectives. For
> >these reasons, "tin" should be an adjective in the above example, even
> >though it might be thought of as a noun.
> 
> Hmm. The reason I didn't bring this up was that I didn't know of it.

Sorry, my first sentence probably came across the wrong way. Of course
you didn't know about this.

> But it sounds rather limited to me. How does Hugo handle synonyms?
> I'm talking about true synonyms, where you really want several nouns
> to refer to the same object (though probably not at the same time).
> Suppose you have an NPC that is 15-year-old Lucy; you'd probably
> want to be able to refer to her as 'Lucy', 'girl' or 'woman', all of
> which should be parsed as nouns.

Well, there's the thing: "you really want several nouns to refer to the
same object (though probably not at the same time)."

Hugo adjectives are exactly the same as Inform names. It would be
perfectly valid to have the lucy object with
  adjectives "lucy", "girl", "woman"

and no nouns at all. This would mean that, as in Inform, the player
could refer to "lucy girl girl woman lucy," etc.

If instead you had
  nouns "lucy", "girl", "woman"

and no adjectives, you could only refer to "lucy," *or* "girl," *or*
"woman," as you can't use multiple nouns to refer to any object.

More likely, though, you'd use both adjectives and nouns (and this isn't
really a very good example, but it should make the point)
  adjective "lucy"
  nouns "girl", "woman

Now you could refer to "lucy," or "girl," or "woman," or "lucy girl," or
"lucy woman" (and in any one you could use "lucy" many times, but only
before you used "girl" or "woman"). You couldn't refer to, for example,
"girl lucy" or "woman girl."

As I said, not a great example, but I don't think this is more limiting
than the Inform method (as you can always simulate that by just using
adjectives). Well, I suppose it does impose a certain "limitation" on
the player--to use reasonably sensible names, but this isn't a bad thing
IMO. OTOH there is certainly room for improvement, by using particles,
or compound nouns (actually, I think "composite noun" would be a better
phrase) and "hard spaces."

Anyone else?

Jools
-- 
"For small erections may be finished by their first architects; grand
ones, true ones, ever leave the copestone to posterity.  God keep me
>from ever completing anything." -- Herman Melville, "Moby Dick"



From erkyrath@netcom.com Tue Sep 30 10:15:54 MET DST 1997
Article: 29574 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!c-1996!feed1.news.erols.com!howland.erols.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] A question of punctuation.
Message-ID: <erkyrathEHAAxz.6vo@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <60h5ns$kr0$1@joe.rice.edu> <60on9c$sjj$1@joe.rice.edu> <60ots2$pdn$1@bartlet.df.lth.se>
Date: Mon, 29 Sep 1997 19:17:58 GMT
Lines: 71
Sender: erkyrath@netcom3.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29574

Magnus Olsson (mol@bartlet.df.lth.se) wrote:
> In article <60on9c$sjj$1@joe.rice.edu>,
> Lucian Paul Smith <lpsmith@rice.edu> wrote:
> >Okay, I managed to figure this out myself--it turns out that there is no
> >way (in the current library) to get NextWord() or NextWordStopped() to
> >give you any useful feedback about punctuation other than the comma and
> >the period.  Any punctuation encountered by NextWord is returned as a
> >zero, exactly the same as unknown words.  In my game, however, it was
> >important to distinguish between:
> >
> >say "pop" to man
> >
> >and 
> >
> >say fizzle pop pow to man

> I had a similar problem a few months ago, but decided on modifying
> the grammar rather than the Library. The difference is that my solution
> kept the default behaviour for all the verbs except for "say".

I hate to bring up potential problems -- but when you write this sort of 
code, be careful to watch for 'then' or commas or periods.

That is, once you've tested

> say "pop" to man
> say fizzle

try

> say "pop" to man. jump
> say fizzle. jump

If you want the last example to be interpreted as "say 'fizzle . jump'",
that's fine; that's what will happen. However, if you want it to be "say
fizzle" followed by "jump", you have check for 'then' and other
punctuation. 

This is a little tricky. When you have a loop like

>     do {
> 	if (WordIsQuote(wn)) {
> 	    ++wn;
> 	    quote = false;
> 	}
>         w = NextWordStopped();
> 	if (w == -1)
> 	    break;
>     } until (w == 'to' && ~~quote);

(From Magnus's example), you would want

  if (w == -1 or THEN1__WD or THEN2__WD or THEN3__WD or comma_word)

(The case of a period is converted to THEN1__WD by NextWordStopped(), so 
you can ignore it.)

However, there is something weird with the way wn is set in such a loop. 
I had to put

  wn--;

at the end of the parse routine, before returning 0, or else the
two-command input still wouldn't work right. I'm not sure why. 

--Z

-- 

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


From erkyrath@netcom.com Tue Sep 30 13:03:02 MET DST 1997
Article: 29574 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!c-1996!feed1.news.erols.com!howland.erols.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] A question of punctuation.
Message-ID: <erkyrathEHAAxz.6vo@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <60h5ns$kr0$1@joe.rice.edu> <60on9c$sjj$1@joe.rice.edu> <60ots2$pdn$1@bartlet.df.lth.se>
Date: Mon, 29 Sep 1997 19:17:58 GMT
Lines: 71
Sender: erkyrath@netcom3.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29574

Magnus Olsson (mol@bartlet.df.lth.se) wrote:
> In article <60on9c$sjj$1@joe.rice.edu>,
> Lucian Paul Smith <lpsmith@rice.edu> wrote:
> >Okay, I managed to figure this out myself--it turns out that there is no
> >way (in the current library) to get NextWord() or NextWordStopped() to
> >give you any useful feedback about punctuation other than the comma and
> >the period.  Any punctuation encountered by NextWord is returned as a
> >zero, exactly the same as unknown words.  In my game, however, it was
> >important to distinguish between:
> >
> >say "pop" to man
> >
> >and 
> >
> >say fizzle pop pow to man

> I had a similar problem a few months ago, but decided on modifying
> the grammar rather than the Library. The difference is that my solution
> kept the default behaviour for all the verbs except for "say".

I hate to bring up potential problems -- but when you write this sort of 
code, be careful to watch for 'then' or commas or periods.

That is, once you've tested

> say "pop" to man
> say fizzle

try

> say "pop" to man. jump
> say fizzle. jump

If you want the last example to be interpreted as "say 'fizzle . jump'",
that's fine; that's what will happen. However, if you want it to be "say
fizzle" followed by "jump", you have check for 'then' and other
punctuation. 

This is a little tricky. When you have a loop like

>     do {
> 	if (WordIsQuote(wn)) {
> 	    ++wn;
> 	    quote = false;
> 	}
>         w = NextWordStopped();
> 	if (w == -1)
> 	    break;
>     } until (w == 'to' && ~~quote);

(From Magnus's example), you would want

  if (w == -1 or THEN1__WD or THEN2__WD or THEN3__WD or comma_word)

(The case of a period is converted to THEN1__WD by NextWordStopped(), so 
you can ignore it.)

However, there is something weird with the way wn is set in such a loop. 
I had to put

  wn--;

at the end of the parse routine, before returning 0, or else the
two-command input still wouldn't work right. I'm not sure why. 

--Z

-- 

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


From erkyrath@netcom.com Tue Sep 30 15:56:29 MET DST 1997
Article: 29574 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!c-1996!feed1.news.erols.com!howland.erols.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] A question of punctuation.
Message-ID: <erkyrathEHAAxz.6vo@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <60h5ns$kr0$1@joe.rice.edu> <60on9c$sjj$1@joe.rice.edu> <60ots2$pdn$1@bartlet.df.lth.se>
Date: Mon, 29 Sep 1997 19:17:58 GMT
Lines: 71
Sender: erkyrath@netcom3.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29574

Magnus Olsson (mol@bartlet.df.lth.se) wrote:
> In article <60on9c$sjj$1@joe.rice.edu>,
> Lucian Paul Smith <lpsmith@rice.edu> wrote:
> >Okay, I managed to figure this out myself--it turns out that there is no
> >way (in the current library) to get NextWord() or NextWordStopped() to
> >give you any useful feedback about punctuation other than the comma and
> >the period.  Any punctuation encountered by NextWord is returned as a
> >zero, exactly the same as unknown words.  In my game, however, it was
> >important to distinguish between:
> >
> >say "pop" to man
> >
> >and 
> >
> >say fizzle pop pow to man

> I had a similar problem a few months ago, but decided on modifying
> the grammar rather than the Library. The difference is that my solution
> kept the default behaviour for all the verbs except for "say".

I hate to bring up potential problems -- but when you write this sort of 
code, be careful to watch for 'then' or commas or periods.

That is, once you've tested

> say "pop" to man
> say fizzle

try

> say "pop" to man. jump
> say fizzle. jump

If you want the last example to be interpreted as "say 'fizzle . jump'",
that's fine; that's what will happen. However, if you want it to be "say
fizzle" followed by "jump", you have check for 'then' and other
punctuation. 

This is a little tricky. When you have a loop like

>     do {
> 	if (WordIsQuote(wn)) {
> 	    ++wn;
> 	    quote = false;
> 	}
>         w = NextWordStopped();
> 	if (w == -1)
> 	    break;
>     } until (w == 'to' && ~~quote);

(From Magnus's example), you would want

  if (w == -1 or THEN1__WD or THEN2__WD or THEN3__WD or comma_word)

(The case of a period is converted to THEN1__WD by NextWordStopped(), so 
you can ignore it.)

However, there is something weird with the way wn is set in such a loop. 
I had to put

  wn--;

at the end of the parse routine, before returning 0, or else the
two-command input still wouldn't work right. I'm not sure why. 

--Z

-- 

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


From mol@bartlet.df.lth.se Tue Sep 30 19:07:12 MET DST 1997
Article: 29624 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: What is an Acorn?
Date: 30 Sep 1997 19:06:15 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 48
Message-ID: <60rbi7$t8b$1@bartlet.df.lth.se>
References: <3415BFF0.4346@juno.com> <60i77l$k2a$1@neko.syix.com> <60j41f$kb5$1@bartlet.df.lth.se> <ant272342b49M+4%@gnelson.demon.co.uk>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:29624

In article <ant272342b49M+4%@gnelson.demon.co.uk>,
Graham Nelson  <graham@gnelson.demon.co.uk> wrote:
>In article <60j41f$kb5$1@bartlet.df.lth.se>, Magnus Olsson
><URL:mailto:mol@bartlet.df.lth.se> wrote:
>> 
>> I bought an Acorn Atom kit computer by mail order from them in
>> 1982. That was a small, 6502 based machine with a whopping 12 Kb of
>> RAM. The current Archimedes is a RISC machine with a little more power
>> than that... AFAIK, they still make them.
>
>I had an Atom, too -- a bottom of the range one, with only 2K,
>but after a year or two I'd managed to upgrade it to 12.  I had
>the circuit diagram pinned to my wall (hey, that was how much fun
>it was being a geek teenager in the 80s) and it was the last
>machine I ever owned where you could actually see exactly how it
>worked, and how it would go wrong if you cut any particular track
>on the PCB -- there were no mysterious black boxes called just "ULA"
>in those days.

Apart from the fact that there was no documentation of the internals
(at least none I knew of), it was a very nice and "open" machine for
hacking.  Something that was made easier by the (BCPL-inspired)
low-level operators in Basic (much more convenient than PEEK and POKE)
and the in-line assembler. I bought an extension ROM from Watford
Electronics (after reading an advert in some British computer
magazine), which provided me with a disassembler and some other tools
that made for great fun exploring the ROMs.

>You'll be pleased to know the basic operating system design of
>the Acorn Atom lives on in the Risc PC, though.  Commands are
>entered at the same * prompt: *LOAD, *SAVE, *CAT, they're all
>still in there, as are the OSWRCH and OSRDCH vectors, and...

Ah, nostalgia...

In case anyone wonders, this is relevant to the newsgroup, since Imade
my first acquaintances with text adventures on the Atom: the Acornsoft
adventures. They weren't really IF, but combat-based games in a
room-based world (like a single-user MUD).

I alsow wrote my first own adventure on the Atom - see the file
atomia.zoo on the IF archive.


-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From erkyrath@netcom.com Mon Oct  6 16:41:24 MET DST 1997
Article: 29785 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!news.idt.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [INFORM] Order of described objects
Message-ID: <erkyrathEHJEE9.HLx@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <615r2n$1u$1@news1.fast.net>
Date: Sat, 4 Oct 1997 17:10:56 GMT
Lines: 75
Sender: erkyrath@netcom15.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29785

ravipind@fast.net wrote:
> Just when you thought it was safe to go coding...

> *sigh*

> Here we go again. :)

> OK.  I've got an object that appears in several locations (I'm using found_in) 
> and I need it's 'initial' property to be the last one described in the room.

> In other words, in one of the rooms there are several other objects that have 
> "initial".  I want the found_in object's description to be printed last.

> For some reason, it's printed first.

Very simple reason -- when you insert an object in a linked list, it 
appears at the top of the chain.

The order of objects in a container is an icky problem, and there are a
few ways to deal with it. 

1: Clever and ugly code. Assume the order of objects in a room is 
arbitrary. Give everything an "initial" property which checks for the 
other object's existence, and displays it under certain circumstances. 
Come to think of it, this is not a general solution; I've used it when I 
had exactly two objects and it was important that one appear first. I 
just had one check for the other's existence, and remain silent, and the 
other one printed messages for both.

2: Ugly brute force.

Object shiftercache;
[ ShiftToEnd obj   par ix;
  if (child(shiftercache))
    print_ret "ShiftToEnd: shiftercache isn't empty! [BUG]";
  par = parent(obj);
  if (par == nothing)
    print_ret "ShiftToEnd: ", (object) obj, " is in nothing! [BUG]";
  while (child(par)) {
    ix = child(par);
    if (ix == obj)
      remove ix;
    else
      move ix to shiftercache;
  }
  move obj to par;
  while (child(shiftercache)) {
    ix = child(shiftercache);
    move ix to par;
  }
];

Calling ShiftToEnd(obj) will move obj to the end of whatever it's in. 

3: Ugly clever hacks.

Object thingie
  with
    describe [;
      give self general; 
      rtrue; 
    ],
    each_turn [; 
      if (self has general) {
        print "^A thingie floats by.^";
        give self ~general;
      }
    ];

--Z

-- 

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


From erkyrath@netcom.com Mon Oct  6 16:41:26 MET DST 1997
Article: 29792 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!wuff.mayn.de!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [INFORM] Order of described objects
Message-ID: <erkyrathEHK9CH.1M7@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <615r2n$1u$1@news1.fast.net> <erkyrathEHJEE9.HLx@netcom.com> <E63VbHAfcqN0Ewit@highmount.demon.co.uk>
Date: Sun, 5 Oct 1997 04:19:28 GMT
Lines: 40
Sender: erkyrath@netcom21.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29792

Neil Brown (neil@this.address.is.fake) wrote:
> At 17:10:56 on Sat, 4 Oct 1997, Andrew Plotkin wrote:
> >3: Ugly clever hacks.
> >
> >Object thingie
> >  with
> >    describe [;
> >      give self general; 
> >      rtrue; 
> >    ],
> >    each_turn [; 
> >      if (self has general) {
> >        print "^A thingie floats by.^";
> >        give self ~general;
> >      }
> >    ];

> In certain circumstances this may not work. For instance, if this object
> was in the very first location at the beginning of the game, each_turn
> wouldn't execute before the end of the first turn, which means that it
> wouldn't appear right at the start. Er, I think.

True. For this very reason I sometimes hack the library to call
end_turn_sequence before the first turn -- so that each_turns and daemons
will show up before the first prompt. 

That's actually the only case I think this will fail. 

Hm. If a daemon kills you or yanks you away after your move/look
command, the each_turn won't run. That's another. There are probably 
others less likely.

--Z



-- 

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


From erkyrath@netcom.com Tue Oct  7 09:55:01 MET DST 1997
Article: 29857 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!wuff.mayn.de!Cabal.CESspool!bofh.vszbr.cz!eerie.fr!news.maxwell.syr.edu!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] Frotz problem with articles
Message-ID: <erkyrathEHo1EK.J3K@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <1997Oct7.085803@vaxc.cc.monash.edu.au>
Date: Tue, 7 Oct 1997 05:18:20 GMT
Lines: 58
Sender: erkyrath@netcom8.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29857

David Dyte (ddyte@vaxc.cc.monash.edu.au) wrote:

> There is an item in my game which begins with a vowel but is pronounced
> in such a way that "a" is the correct article. I have given it article "a"
> accordingly. When I run my game under jzip, no problem, works a treat.
> However Frotz for DOS is giving me "an" in all my inventory lists and
> room descriptions.

As you note in a later post, this is a Jzip bug, or rather a part of Jzip 
which has not been updated to the latest Z-Machine standard. (And you 
forgot to give it the article "a" property.)

Actually, in a sense it's not a bug *or* old code, but *correct* behavior 
on the part of JZip. Let's see... 

The most modern Inform library attempts to determine the correct article 
for a word by printing it to an array, and seeing if it starts with a 
vowel. However, it only does this if the interpreter announces itself as 
compliant with some post-Infocom version of the Z-Spec (ie, one of 
Graham's versions, 0.2 or later.)

There is a good reason for this: On Infocom's interpreters, the
output-to-array operation will screw up if it occurs nested inside another
output-to-array. That means that if you divert the output stream to an
array, and then "print (a) object", and the library tries to infer the
article using print-to-array, bad things will happen. 

The modern Z-Spec says that nested output-to-array operations should be
handled correctly. Therefore, the library doesn't try to infer articles
unless the interpreter says it is modern. 

Currently, Frotz says it is modern, and (I believe) all the ZIP 
derivatives don't.

The upshot of this is that articles behave differently on Frotz and on 
ZIP derivatives. 

If you don't like this, just be sure to provide an article whenever
the object's name begins with a vowel. The article property always 
overrides this foofaraw.

You can also disable the article-inferring mechanism in the library. Go
into parserm, look at the PrefaceByArticle function, and rip out the
six-line block that starts

  if (standard_interpreter ~= 0 && findout) ...

I find this to be safer, but as I've said, I'm a professional paranoid.

--Z




-- 

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


From erkyrath@netcom.com Fri Oct 10 19:19:43 MET DST 1997
Article: 29991 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!mn6.swip.net!nntp.uio.no!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] How do you prevent the player from entering?
Message-ID: <erkyrathEHtrAH.5EJ@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <61jes9$o8j$1@mathserv.mps.ohio-state.edu>
Date: Fri, 10 Oct 1997 07:25:28 GMT
Lines: 105
Sender: erkyrath@netcom15.netcom.com
Xref: news.lth.se rec.arts.int-fiction:29991

Ethan Dicks (dicks@math.ohio-state.NO.SPAM.edu) wrote:

> OK.  I've been over the manual, I've been over the libraries, I've been
> over sample adventures... I can't figure out how to block the player from
> entering arbitrary rooms.  It's trivial to implement a Star Trek-style
> force field with doors, but I need to keep the player out of a room,
> no matter what the adjacent rooms are; think of a poison cloud, or a
> roving actor that the player is required to avoid.

>  What's not obvious to me is the sequence of events that
> happen when the player moves from room to room.  Do I need to re-write
> PlayerTo()?  MovePlayer()?

Gosh, you picked an ugly problem. 

There is no standard way of doing this.

There is no simple library hack, either, because there is no single 
bottleneck for player movement. 

If I was faced with this, I would really honestly put up force-field doors
on every single door and passage in the affected area. This doesn't have
to be code-heavy. Write a single function

[ MaybeGo dest;
  if (dest has unenterable)
    "You bounce off a poison cloud and stagger back.";
  else
    return dest;
];

and then have rooms with directions like

with
  n_to [; return MaybeGo(RoundRoom); ],
  e_to [; return MaybeGo(LongCanyon); ];

If you really want the general solution... (Warning: deep Inform library 
spelunking ahead.)

There are four places in the code where the player is actually moved: 
PlayerTo(), EnterSub(), ExitSub(), GoSub(). If you're talking about 
unenterable rooms, as opposed to unenterable chairs or bathtubs, you can 
*probably* ignore EnterSub() and ExitSub(). 

So you change PlayerTo() to check if its argument is unenterable, and if 
it is, print the appropriate message. But you probably want to be careful 
now when you're calling PlayerTo(), because there's now the possibility 
that it will fail. You'll have to resolve this sort of conflict yourself. 
(What happens if the player says XYZZY but the other teleport room is 
occupied by the poison cloud?) Fortunately, the library never calls 
PlayerTo(), so you only have to worry about this possibility in your own 
code.

Then look at GoSub(). Then find some aspirin, and look again. I *think* 
the appropriate place to check is right before the line

  if (movewith==0) move player to j; else move movewith to j;

You'd stick in something like

  if (j has unenterable) 
    "You bounce off a poison cloud and stagger back.";

before that line. That should take care of riding in vehicles as well as 
walking.

But now there's a problem; the library calls <Go dir> in a few places, 
and doesn't take into account the possibility that the movement will 
fail. (It also calls <<Go dir>>, but that's not a big deal, since it will 
return immediately whether the movement succeeds or fails.)

This occurs in AllowPushDir() and MovePlayer(). In both cases you'd want 
to change 

  <Go dir>;

to something like

  oldloc = location;
  <Go dir>;
  if (location == oldloc)
    rtrue;

but this may require a lot of fiddling. (Or, ignore the problem by not 
calling AllowPushDir() or MovePlayer() anywhere.)

Calling ChangePlayer() makes this several dozen times worse, since now 
you have to worry about off-duty player objects staying out of the poison 
cloud as well as the current player.

--Z









-- 

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


From spatula@javanet.com.andbacon Sat Oct 11 19:30:11 MET DST 1997
Article: 30004 of rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!luth.se!news-ge.switch.ch!newscore.univie.ac.at!newsfeed.Austria.EU.net!EU.net!news.maxwell.syr.edu!newsfeed.internetmci.com!208.134.60.8!news.javanet.com!spatula
From: spatula@javanet.com.andbacon (tv's Spatch)
Newsgroups: rec.arts.int-fiction
Subject: Re: [IF-COMP 97] Games are up.
Date: 10 Oct 1997 14:31:37 GMT
Organization: The Group W Bench
Lines: 113
Message-ID: <slrn63sf20.p9c.spatula@kona.javanet.com>
References: <61gosu$p7n@argentina.earthlink.net> <343c3650.2300980@news.demon.co.uk> <61hmg2$b4$1@calvin.skidmore.edu> <343cb502.4359503@news.demon.co.uk> <3440e7f2.64966484@newsserver.rdcs.kodak.com> <343ced86.18829969@news.demon.co.uk> <61j5df$9qh@drn.zippo.com> <343d9844.9191132@news.accessone.com>
NNTP-Posting-Host: kona.javanet.com
X-Newsreader: slrn (0.9.4.3 UNIX)
Xref: news.lth.se rec.arts.int-fiction:30004

Wormer, he's a dead man!  Marmalard: dead!  Alan Conroy: 
> >When you post a message, it goes spreading out around the world at
> >the speed of electricity. Does "cancel" release some kind of
> >electronic worm to hunt down every copy of a message and destroy
> >it?
> 
> Sounds like a good plot for a new game ;-)

     Late in the 20th century, the USENET CORPORATION advanced newsgroup
     evolution into the WRECKS-US phase -- a spam post virtually identical
     to a human's -- known as a Spamford.

     The WRECKS-US 6 Spamfords were superior in content and propagation,
     and at least equal in some intelligence, to the exploitative
     programmers who created them.

     Spamfords were used Off-alt.* as idiot labor, in the hazardous
     exploration of advertising newsgroups and Web promotion.

     After a bloody mutiny by a WRECKS-US 6 advertising agency in
     biz.ads.cool-web-pages, Spamfords were declared illegal on Usenet -
     under penalty of death.  Special netcop squads -- CANCELBOT UNITS --
     had orders to post to cancel, upon detection, any trespassing
     Spamfords.

     This was not called censorship.
     It was called retroactive moderation.

     ALT.MIDNIGHT.CITY.STUMBLE.LURK.SKULK
     OCTOBER, 1997

DO SPAMMERS DREAM OF ELECTRIC PORK
A not-so-interactive newsgroup demonstration
Written half-heartedly in the middle of tech calls by R. Noyes
Release 1 / Serial number 971010 / rec.arts.int-fiction

Administering the Lewis-Furr Test
You fidget in your hardwood chair, staring uncomfortably at the gray
surroundings of the testing cubicle.  For all the USENET Corporation's
untold wealth and advanced technology, you'd think they could afford to
hire decent interior decorators.  But no.  All is utilitarian, all is
gray, all is Spartan -- except for the cute poster of the cat hanging off
a branch you brought in one day to cheer things up a little.  

Your laptop obediently sits on the table, awaiting a command.

> EXAMINE LAPTOP

The latest in the ThinkForYourselfPad line of fun laptops for corporate
use.  As it is your only link to the outside world, use of it is
monitored constantly by the USENET Corporation -- which is a drag because
you've been itching to play Stiffy Makane 5 for a long time now.  

Currently your laptop screen displays the standard Post Review procedure.
A post to alt.games.heist,no.flaming blinks serenely at you.  It is up to
you to decide whether or not it is a Spamford.  Day in, day out.  For
about 5 minutes every day.  The rest you just spend here in that damn
chair.

Ceiling fans revolve fruitlessly overhead.

> POST, HELLO

"Hi! I'm new to the newsgroup. Wanna see a real cool homepage?"

> BEGIN THE TEST

The all-too-familiar L-F Test Interface now shares the laptop screen with
the Post.  A list of questions numbered from one to ten appears on the
left. 

A searchlight sweeps across the night, lighting up your room temporarily
for no apparent reason.

> ASK QUESTION ONE

You click on the Question One icon.
"Post, you are walking down the road and you come across a turtle who
is trying to cross the road.  It is a hot day and the turtle really wants
to cross but it's so slow, it can't cross.  What do you do?"

"Has it been to see my webpage?"

Little statistics start scrolling in a small window by the question
interface.  You don't know what they mean, and you don't plan on learning
any time soon.

> ASK QUESTION TWO
You click on the Question Two icon.
"Post, if a train is leaving Phoenix heading east at 60mph at 2 PM, and a
train is leaving South Portland heading west into the Pacific Ocean some
time around lunch, why is pie-switching so popular?"

The post hesitates.
"It's a really great webpage, full of lots of neat things and chat rooms
and links to naughty sites too!"

> ASK POST ABOUT ITS MOTHER

That won't be necessary in this parody.

> ASK POST ABOUT ITS WEBPAGE

"Webpage?  I'll tell you about my webpage..."

Internal error(0) logic flaw in parody
END OF SESSION

-- 
spatula@javanet.com.andbacon is but one of many Spatula tentacles. Rar.
          More fun than a Promise Keepers convention -- PUTPBAD!
Let http://www.javanet.com/~spatula/booth.html show you how it's done.
                                         "MERV GRIFFIN!" - Milk & Cheese


From kjfair@midway.uchicago.edu.REMOVEME Sun Oct 19 14:12:31 MET DST 1997
Article: 30328 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!newsfeed.direct.ca!iagnet.net!condor.cns.iit.edu!uchinews!kjfair
From: kjfair@midway.uchicago.edu.REMOVEME (Kenneth Fair)
Subject: [Inform] Sitting in a default chair
X-Nntp-Posting-Host: ntcs-ip306.uchicago.edu
Content-Type: text/plain; charset=ISO-8859-1
Message-ID: <kjfair-ya02408000R1910970141480001@uchinews.uchicago.edu>
Sender: news@midway.uchicago.edu (News Administrator)
Content-Transfer-Encoding: 8bit
Organization: University of Chicago School of Law
X-Newsreader: Yet Another NewsWatcher 2.4.0
Mime-Version: 1.0
Date: Sun, 19 Oct 1997 06:41:48 GMT
Lines: 92
Xref: news.lth.se rec.arts.int-fiction:30328


This isn't a question, since I managed to figure this out on my own.
However, I thought I'd give a heads up to any new Informers out there,
and maybe some old ones, on a little trick I came up with that perhaps
they can use or improve.

I am coding an office which has two chairs in it.  One is the chair of
the person who's office it is; the other is a client chair that the
player can sit in:

   Object -> "wooden client chair"
      with  name "wooden" "client" "chair",
            description "A wooden chair for clients to sit in.",
      has   supporter enterable;

   Object -> "black leather chair"
      with  name "black" "leather" "chair",
            description "This is Bob's chair, covered with black leather.",
            before [; Enter: "Bob's sitting in that chair."; ],
      has   supporter enterable;

Okay, nice and simple.  But note what happens:

   >SIT IN CHAIR
   Which chair do you mean, the wooden client chair or the black leather 
   chair?

I thought I'd like to give the player a break.  Shouldn't the game just
assume that the player is referring to the client chair unless otherwise
specified?

So here's what I did.  I created a Chair class, with a special parse_name
routine:

   Class Chair
      with  name "chair",
            parse_name [; return -1; ],
      has   supporter enterable;

And then the chairs:

   Chair -> "wooden client chair"
      with  name "wooden" "client",
            description "A wooden chair for clients to sit in.";

   Chair -> "black leather chair"
      with  name "black" "leather",
            description "This is Bob's chair, covered with black leather.",
            before [; Enter: "Bob's sitting in that chair."; ];

The name property is additive, so "chair" is added to both chairs' name list.

What happens with the parse_name routine?  By returning -1, the parse_name
routine tells the parser that it doesn't want to parse the name.  The
parser then goes ahead and checks the name list like normal.  No big deal.
So for example:

   >EXAMINE BLACK CHAIR
   This is Bob's chair, covered with black leather.

   >EXAMINE WOODEN CHAIR
   A wooden chair for clients to sit in.

However, the trick comes if the player just uses "chair" and doesn't
specify which one.  The parser attempts to select an object, but it can't
choose between the two chairs. Why not?  Because they have the same
parse_name routine, which they inherited from the Chair class.  So the
parser calls the parse_name routine for each a second time, this time with
parser_action set to ##TheSame.  Normally, a parse_name routine can use
this to check that two objects are different-- as would normally be the
case here--rather than indistinguishable nouns like coins or dollar bills. 
BUT, we tell the parser (by returning -1 again) that the two chairs
actually ARE the same.  What does the parser do then? It just picks the
first object defined, namely the wooden client chair. If I had defined the
black leather chair above the wooden client chair, it would pick that chair
instead.  See Chapters 24 and 25 of the Designer's Manual for more details.

So the result is:

   >SIT IN BLACK LEATHER CHAIR
   Bob's sitting in that chair.

   >SIT IN CHAIR
   You get on the wooden client chair.

Neat, huh?

-- 
KEN FAIR - U. Chicago Law  | <http://student-www.uchicago.edu/users/kjfair>
Of Counsel, U. of Ediacara | Power Mac! | CABAL(tm) | I'm w/in McQ - R U?

  The Internet was not created for companies to make money from.


From nestvold@metro.net Sun Oct 19 14:13:08 MET DST 1997
Article: 30332 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!news-fra.maz.net!newsfeed.ecrc.net!newsfeed.metronet.de!news.metronet.de!not-for-mail
From: nestvold@metro.net (Ruth Nestvold)
Newsgroups: rec.arts.int-fiction
Subject: IF and narrative theory
Date: Sun, 19 Oct 1997 11:41:53 GMT
Organization: Metronet
Lines: 42
Message-ID: <62cr8r$n2g$1@news.metronet.de>
NNTP-Posting-Host: stuttgart2.pop.metronet.de
X-Newsreader: Forte Free Agent 1.0.82
Xref: news.lth.se rec.arts.int-fiction:30332

Once again, I will try to respond to Adam's very thoughtful post on my article.

>(1) Are you familiar with the work of Seymour Chatman, by any chance?

Yes. For this paper I was more interested in using the theories which dealt with
the second person narrative situation in detail. 

>(2) In the section on the specificity of the protagonist, it might be
>    productive to explore the difference between specificity inherent
>    in the character at the beginning and specificity added later due
>    to the player's commands.  

Some _very_ interesting thoughts in this section. Are there that many games
which allow the reader / player to reach the assigned goal using different
methods? It's an honest question - my experience is limited. I'll take a look at
the games you mentioned. Trinity I've started, but I'm not familiar with the
other two.

>(3) You reflect on a definition of fiction that would insist on print
>    as a medium, which reminded me that, at least in my experience,
>    most of the recent work that has been done on narrative has been
>    in the field of film studies

Part of what I wanted to do in this paper was to examine the way standard
narrative theories based on print media are inadequate when confronted with the
written word conveyed electronically. Also, I was limited by the demands of the
conference - I gave the paper at an international convention on literary
semantics. The spatial metaphor for the text being so much mored pronounced in
electronic fictional forms, however, I should eventually draw on film theory in
my research too. :/ 

>(4) A technical point: "That's not a verb I recognize" is a product of
>    the Inform language, not the Frotz interpreter.

Thanks. You and half a dozen others. :) 

Ruth

Ruth Nestvold, rmn@lit-arts.com
Cutting Edges: Or, a Web of Women
www.lit-arts.com/cutting_edges



From bifurcate@geocities.com Sun Oct 19 14:13:29 MET DST 1997
Article: 30090 of rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!news.edu.sollentuna.se!news.oru.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.internetmci.com!204.238.120.130!jump.net!grunt.dejanews.com!not-for-mail
Date: Sun, 12 Oct 1997 21:11:42 -0600
From: Rick Dague <bifurcate@geocities.com>
Subject: [INFORM] stopping autotake
Newsgroups: rec.arts.int-fiction
Message-ID: <876708395.26900@dejanews.com>
Organization: Deja News Posting Service
X-Article-Creation-Date: Mon Oct 13 02:06:36 1997 GMT
X-Originating-IP-Addr: 166.55.23.87 (usr18-dialup23.mix1.Bloomington.mci.net)
X-Http-User-Agent: Mozilla/3.01 (Win16; I)
X-Authenticated-Sender: Rick Dague <bifurcate@geocities.com>
Lines: 33
Xref: news.lth.se rec.arts.int-fiction:30090

I figured out how to stop auto-take on certain items. I'm sorry I'm
starting a new thread instead of replying to the old one, but I'm having
newsgroup posting problems. :)

The message "(first taking the object)" is printed _before_ auto-take, and
the manual shows how to replace the message by defining the object
LibraryMessages. You can add the following lines in your source code
between the Include "Parser"; and Include "Verblib" lines.

  Object LibraryMessages
  with
    before [;
      Miscellany:
        if(lm_n == 26 && not_holding == carrot)
          "You're not holding ",(the)not_holding,".";
    ];

And after that's printed the before property of carrot will be called with
the variables action == ##take and onotheld_mode == true.

  Object carrot "carrot"
  with
    name 'carrot',
    before [;
      take: return onotheld_mode;
    ],
  has edible;

I hope that's not too unclear. :)
-- Rick

-------------------==== Posted via Deja News ====-----------------------
      http://www.dejanews.com/     Search, Read, Post to Usenet


From kjfair@midway.uchicago.edu.REMOVEME Sun Oct 19 14:13:58 MET DST 1997
Article: 30058 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news01.sunet.se!sunic!mn6.swip.net!nntp.uio.no!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.internetmci.com!192.52.106.6!ncar!uchinews!kjfair
From: kjfair@midway.uchicago.edu.REMOVEME (Kenneth Fair)
Subject: Re: [Inform] How to nix implicit taking
X-Nntp-Posting-Host: ntcs-ip52.uchicago.edu
Content-Type: text/plain; charset=ISO-8859-1
Message-ID: <kjfair-ya02408000R1110971922390001@uchinews.uchicago.edu>
Sender: news@midway.uchicago.edu (News Administrator)
Content-Transfer-Encoding: 8bit
Organization: University of Chicago School of Law
X-Newsreader: Yet Another NewsWatcher 2.4.0
References: <kjfair-ya02408000R0910970200590001@uchinews.uchicago.edu> <343F07B9.2340@-.->
Mime-Version: 1.0
Date: Sun, 12 Oct 1997 00:22:38 GMT
Lines: 69
Xref: news.lth.se rec.arts.int-fiction:30058


Recap: I have an object (representing various plants in a room) that I
want to provide some amusing responses for various actions (Take, Eat, etc.)
Unfortunately, actions such as Eat, Disrobe, and Unlock--all actions in
the Grammar with the "held" or "multiheld" token--generate an "implicit
take" for the object, i.e., they send a Take action to the object if the
player is not holding it.  This causes two problems: first, it generates a
"(first taking the <object>)" message; and second, it moves directly to
a Take action, thereby returning my amusing Take response rather than
my amusing Eat response.

I think I've generated a solution for this problem, but I'd like to see
if someone who is more knowledgeable about the parser can shoot a hole in it.

The solution revolves around the variable "action_to_be."  The parser stores
the action actually called for by the player here, rather than in the "action"
variable, which can contain other actions generated by the parser.

So to prevent the Take rule from stealing the Eat action, one can do this:

Object -> foliage "foliage"
  with  name "planter" "flowers" "flower" "planters"
          "plant" "plants" "foliage" "tree" "trees",
        description  "You admire the foliage and take a whiff of 
            the flowers.  Mmm, they smell lovely.  All of the foliage
            is certainly well kept.";
        react_before [;
          if (self == noun) {
            switch (action_to_be) {
              ##Eat: "You absently-mindedly snap off a leaf and pop it in 
                  your mouth, a move you regret instantly.  Blecch!  You
                  spit the leaf out and toss it back.";
                  }
               }
        ],
        before [;
          Smell: "You take a deep whiff of the plants.  The airy floral
              scent transports you for an instant back to happier days
              at home.";
          Take: "What, are you going to become a gardener now?";
       ],
   has scenery edible;   

If there is an implicit take, action_to_be will have the actual action,
in this case Eat, while action has the value Take.  Since the parser
has generated a <Take foliage> action, noun is set to foliage.  Thus
(self == noun) and the react_before rule traps only those actions 
affecting the foliage object.  Note that Smell works fine as before
since Smell does not generate an implicit take.


The first problem of the "(first taking the <object>)" message is easily
solved by a LibraryMessages object:

Object  LibraryMessages
  with  before [;
           Miscellany: if (lm_n == 26) {
               if (lm_o == foliage) rtrue; ! This keeps the (first taking)
                                           ! message from appearing
             }
        ];

See Section 21 of the Designer's Manual.

-- 
KEN FAIR - U. Chicago Law  | <http://student-www.uchicago.edu/users/kjfair>
Of Counsel, U. of Ediacara | Power Mac! | CABAL(tm) | I'm w/in McQ - R U?
 "My vulva is wide opened, a cave of darkness where the people . . . hang out 
  at night . . . .  There is music playing in my Womb."  - Doctress Neutopia


From erkyrath@netcom.com Sun Oct 19 14:14:00 MET DST 1997
Article: 30128 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!news.edu.sollentuna.se!newsfeed.pop.de!newsfeed.ecrc.net!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] How to nix implicit taking
Message-ID: <erkyrathEI0oDn.97z@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <kjfair-ya02408000R0910970200590001@uchinews.uchicago.edu> <343F07B9.2340@-.-> <kjfair-ya02408000R1110971922390001@uchinews.uchicago.edu>
Date: Tue, 14 Oct 1997 01:05:47 GMT
Lines: 65
Sender: erkyrath@netcom4.netcom.com
Xref: news.lth.se rec.arts.int-fiction:30128

Kenneth Fair (kjfair@midway.uchicago.edu.REMOVEME) wrote:

> Recap: I have an object (representing various plants in a room) that I
> want to provide some amusing responses for various actions (Take, Eat, etc.)
> Unfortunately, actions such as Eat, Disrobe, and Unlock--all actions in
> the Grammar with the "held" or "multiheld" token--generate an "implicit
> take" for the object, i.e., they send a Take action to the object if the
> player is not holding it.  This causes two problems: first, it generates a
> "(first taking the <object>)" message; and second, it moves directly to
> a Take action, thereby returning my amusing Take response rather than
> my amusing Eat response.

>         react_before [;
>           if (self == noun) {
>             switch (action_to_be) {
>               ##Eat: "You absently-mindedly snap off a leaf and pop it in 
>                   your mouth, a move you regret instantly.  Blecch!  You
>                   spit the leaf out and toss it back.";
>                   }
>                }
>         ],
>         before [;
>           Smell: "You take a deep whiff of the plants.  The airy floral
>               scent transports you for an instant back to happier days
>               at home.";
>           Take: "What, are you going to become a gardener now?";
>        ],

It's not necessary to use react_before. You can put it all in the before 
clause, like this:

  before [;
    Take:
      switch (action_to_be) {
        ##Eat: "You eat a leaf.";
        default: "What, are you going to become a gardener now?";
      }
  ];

In fact, you can even do this with a takeable object:

  before [;
    Take:
      switch (action_to_be) {
        ##Eat: "You eat a leaf.";
        default: rfalse;
      }
  ];

will produce an object which you can take normally, and which will 
auto-take normally for most actions; but if it's on the ground and you 
"eat" it, you'll see "You eat a leaf." and it will *not* be taken.

> The first problem of the "(first taking the <object>)" message is easily
> solved by a LibraryMessages object:

Thanks for thinking of this. It's a major pain and I've never come up 
with a good solution.

--Z

-- 

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


From kjfair@midway.uchicago.edu.REMOVEME Sun Oct 19 14:14:02 MET DST 1997
Article: 30147 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!eru.mt.luth.se!newsfeed2.luth.se!luth.se!news.algonet.se!news.maxwell.syr.edu!news-peer.sprintlink.net!news.sprintlink.net!Sprint!newsfeed.internetmci.com!192.52.106.6!ncar!uchinews!kjfair
From: kjfair@midway.uchicago.edu.REMOVEME (Kenneth Fair)
Subject: Re: [Inform] How to nix implicit taking
X-Nntp-Posting-Host: ntcs-ip188.uchicago.edu
Content-Type: text/plain; charset=ISO-8859-1
Message-ID: <kjfair-ya02408000R1410971057030001@uchinews.uchicago.edu>
Sender: news@midway.uchicago.edu (News Administrator)
Content-Transfer-Encoding: 8bit
Organization: University of Chicago School of Law
X-Newsreader: Yet Another NewsWatcher 2.4.0
References: <kjfair-ya02408000R0910970200590001@uchinews.uchicago.edu> <343F07B9.2340@-.-> <kjfair-ya02408000R1110971922390001@uchinews.uchicago.edu> <erkyrathEI0oDn.97z@netcom.com>
Mime-Version: 1.0
Date: Tue, 14 Oct 1997 15:57:03 GMT
Lines: 37
Xref: news.lth.se rec.arts.int-fiction:30147

In article <erkyrathEI0oDn.97z@netcom.com>, erkyrath@netcom.com (Andrew
Plotkin) wrote:


>It's not necessary to use react_before. You can put it all in the before 
>clause, like this:
>
>  before [;
>    Take:
>      switch (action_to_be) {
>        ##Eat: "You eat a leaf.";
>        default: "What, are you going to become a gardener now?";
>      }
>  ];
>
>In fact, you can even do this with a takeable object:
>
>  before [;
>    Take:
>      switch (action_to_be) {
>        ##Eat: "You eat a leaf.";
>        default: rfalse;
>      }
>  ];
>
>will produce an object which you can take normally, and which will 
>auto-take normally for most actions; but if it's on the ground and you 
>"eat" it, you'll see "You eat a leaf." and it will *not* be taken.

There we go.  For some reason I didn't think of putting it inside the 
Take rule.  Beautiful.

-- 
KEN FAIR - U. Chicago Law  | <http://student-www.uchicago.edu/users/kjfair>
Of Counsel, U. of Ediacara | Power Mac! | CABAL(tm) | I'm w/in McQ - R U?
      "Your grasp of science lacks opposable thumbs."
         - Ben Waggoner


From mol@bartlet.df.lth.se Tue Oct 21 15:54:46 MET DST 1997
Article: 30410 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Re: [INFORM] Skipping disambiguation
Date: 21 Oct 1997 15:54:09 +0200
Organization: Societas Datori Universitatis Lundensis
Lines: 52
Message-ID: <62ic61$o9s$1@bartlet.df.lth.se>
References: <62e0ov$us$1@bartlet.df.lth.se> <erkyrathEIBtr8.8M5@netcom.com> <62f139$2ra$1@bartlet.df.lth.se>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:30410

In article <62f139$2ra$1@bartlet.df.lth.se>,
Magnus Olsson <mol@bartlet.df.lth.se> wrote:
>In article <erkyrathEIBtr8.8M5@netcom.com>,
>Andrew Plotkin <erkyrath@netcom.com> wrote:
>>This is normally better handled through the library's identical-objects 
>>system. That is, get the parse_name routine to say that all the relevant 
>>objects are indistinguishable.
>
>That would be rather awkward in this particular case, since I want to
>be able to "foo" every object in the game (OK, I can just as well tell
>you: we're talking about a new implementation of the "ask" verb, one
>that draws on the power of the parser rather than insisting on working
>around it :-) ), and objects that are to be treated as
>indistinguishable must inherit their parse_name routines from the same
>base class. Well, I suppose letting everything inherit froma common
>base class isn't too much trouble, so I'll keep this solution in mind.

On second thought, I think this is not just "rather awkward", but very
awkaward indeed, nearly impossible.  If I've understood the Library
corectly, the parser considers two objects "possibly identical" if

1) They both match the given input.
2) Their parse_name properties point to the same routine (or to nothing).

Note that 2) is not the same thing as requiring that their parse_name
routines be identical, but is a stricter condition, normally
requiring that both objects inherit their parse_name from the same
base class.

If the parser finds that several such objects match the given input,
it will call their parse_name routines to give them a chance to say
"no, we're not identical, treat us as individuals". If they don't
object, the parser just picks one of them, rather than asking "Which
coin do you mean, the coin, the coin or the coin?".


Now, my new "ask" grammar line will in effect want to treat *all*
objects in the game as "possibly identical". So this means that as
soon as two objects share a common noun, they'd have to inherit their
parse_name routines from a common ancestor. Suppose I have a gold
coin, a silver coin, a silver candlestick and a gold bracelet. Then
all those objects would have to have the same parse_name. This could
easily become very complicated...

So I think my proposal, to cheat with ChooseObjects, would be
better. Better still would of course be if Graham modified the parser
to make this possible without cheats.

-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From graham@gnelson.demon.co.uk Tue Oct 21 20:08:33 MET DST 1997
Article: 30424 of rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!wuff.mayn.de!Cabal.CESspool!bofh.vszbr.cz!eerie.fr!news.maxwell.syr.edu!news-peer.gsl.net!news.gsl.net!gip.net!news.idt.net!dispose.news.demon.net!demon!news.demon.co.uk!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: [IfComp97] Derivative games (Tempest)
Date: Tue, 21 Oct 1997 10:16:21 +0100 (BST)
Organization: none
Message-ID: <ant210921d07M+4%@gnelson.demon.co.uk>
References: <877057068.1391@dejanews.com> <Pine.SUN.3.91.971017120830.16113B-100000@bigbang.phy.duke.edu> <6286ft$sla$1@calvin.skidmore.edu> <EI7vLr.ntt@undergrad.math.uwaterloo.ca> <19971018025000.WAA10374@ladder02.news.aol.com> <199710192239381367100@d248.b64.cmb.ma.ultra.net> <62hd3c$kr5@bolivia.earthlink.net> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 30
Xref: news.lth.se rec.arts.int-fiction:30424

In article <62hd3c$kr5@bolivia.earthlink.net>, Dave
<URL:mailto:dempls@earthlink.net> wrote:
>    In article <199710192239381367100@d248.b64.cmb.ma.ultra.net>, 
> meta@pobox.com (mathew) wrote:
> 
> >copied designs from Campbell's soup tins; Roy Lichtenstein copied style
> >and even image elements from pulp graphic novels.  The Situationist
> 
> It isn't really all that important, but as a fan of comics I feel the need to 
> point out that Lichtenstein "lifted" entire panels, right down to the word 
> balloons, from romance comics etc.  Comics indexers have been able to identify 
> which specific panels were used, and even which letterer wrote the words.  
> Lichtenstien used this material to create an entirely new work -- not just 
> panels "blown up" to larger size, but works with greatly different intent and 
> context than the originals.

A key point, though, is that Lichtenstein then signed them
"Lichtenstein".  Whereas an adaptation makes no such appropriation.
The by-line of "The Tempest" reads "William Shakespeare", and
whether or not it succeeds (not, by the looks of things) it can't
be accused of cutting from or subverting the context of the "original".

I write "original" in quotation marks because, for a 17th-century
play, what is the original?  The text tends not to be a theatrical
performance.  Original playscripts often don't even identify who's
talking.  Every production is a collaboration with the text.

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From ats@acm.org Wed Oct 22 10:34:33 MET DST 1997
Article: 30431 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!feed1.news.erols.com!howland.erols.net!news.starnet.net!newsreader.wustl.edu!not-for-mail
From: Alan Shutko <ats@acm.org>
Newsgroups: rec.arts.int-fiction
Subject: Re: [IfComp97] Derivative games (Tempest)
Date: 21 Oct 1997 13:43:30 -0500
Organization: Washington University in St. Louis
Lines: 80
Message-ID: <m390vnc5yl.fsf@hubert.wuh.wustl.edu>
References: <62db1e$go$1@news1.atlantic.net> <19971019171901.NAA21807@ladder01.news.aol.com> <877310497.21462@dejanews.com> <62gem8$358sk@fido.asd.sgi.com> <877415196.15936@dejanews.com>
NNTP-Posting-Host: wydo122.wuh.wustl.edu
Mime-Version: 1.0 (generated by SEMI MIME-Edit 0.88 "Tsurugi")
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-Emacs: Emacs 20.2, MULE 3.0 (MOMIJINOGA)
X-Newsreader: Gnus v5.5/Emacs 20.2
Xref: news.lth.se rec.arts.int-fiction:30431

>>>>> "S" =3D=3D Steve Bernard <x96bernard1@wmich.edu> writes:

S> Only recently have the government and the large copyright owning
S> corporations begun to effectively cut out fair use from our
S> copyright law.  Look at how Paramount has been strongarming the
S> maintainers of the various Star Trek websites.

That's not copyright.  That's (usually) trademark defense.  =


Since trademarks _must_ be defended for a company to keep them, they
only have a choice in terms of magnitude, not direction.  Fair use in
terms of trademarks has nothing to do with profitability.  According
to http://www.law.georgetown.edu/lc/internic/trademarks/tm5.html:


     Fair Use involves several factors: (1) the product or service in
     question must be one not readily identifiable without using the
     mark; (2) only so much of the mark or marks may be used as is
     reasonably necessary to identify the product or service; and (3)
     the user must do nothing that suggests endorsement or sponsorship
     by the mark owner.

Since much of the fan web sites use a whole bunch of trademarked Star
Trek (my use of trademark falls under above fair use) items, Paramount
has a legal case to shut them down.

Copyright would come into the picture for sites which have pictures,
scripts, and sounds duplicated from Star Trek shows.  Here, fair use
is different.  From http://www.benedict.com/fund.htm#fund:



  =A7107. Limitations on exclusive rights: Fair Use

  Notwithstanding the provisions of sections 106 and 106A, the fair use
  of a copyrighted work, including such use by reproduction in copies or
  phonorecords or by any other means specified by that section, for
  purposes such as criticism, comment, news reporting, teaching
  (including multiple copies for classroom use), scholarship, or
  research, is not an infringement of copyright. In determining whether
  the use made of a work in any particular case is a fair use the
  factors to be considered shall include -

     1. the purpose and character of the use, including whether such use
	is of a commercial nature or is for nonprofit educational
	purposes;
     2. the nature of the copyrighted work; =

     3. the amount and substantiality of the portion used in relation to
	the copyrighted work as a whole;
     and =

     4.the effect of the use upon the potential market for or value of
       the copyrighted work. =


  The fact that a work is unpublished shall not itself bar a finding of
  fair use if such finding is made upon consideration of all the above
  factors.

Looking at that, it's pretty clear that Paramount has a right to shut
down site with copyrighted materials.  They are non-profit, but not
educational.  They duplicate the whole of copyrighted works, rather
than a small, necessary subset.  And it has a real harm on the
potential market of the original copyrighted works, since people will
often choose to download scripts and pictures rather than purchasing
them.

Simply being non-profit isn't enough to help you on copyright, and it
doesn't matter at all for trademarks.

That said, most of the Trek web sites are only in violation of
trademarks, not copyright.  Whether you agree with it or not,
Paramount has legal ground to stand on.

-- =

Alan Shutko <ats@acm.org> - By consent of the corrupted
Power tools for power losers.


From kjfair@midway.uchicago.edu.REMOVEME Wed Oct 22 10:47:04 MET DST 1997
Article: 30058 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news01.sunet.se!sunic!mn6.swip.net!nntp.uio.no!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!newsfeed.internetmci.com!192.52.106.6!ncar!uchinews!kjfair
From: kjfair@midway.uchicago.edu.REMOVEME (Kenneth Fair)
Subject: Re: [Inform] How to nix implicit taking
X-Nntp-Posting-Host: ntcs-ip52.uchicago.edu
Content-Type: text/plain; charset=ISO-8859-1
Message-ID: <kjfair-ya02408000R1110971922390001@uchinews.uchicago.edu>
Sender: news@midway.uchicago.edu (News Administrator)
Content-Transfer-Encoding: 8bit
Organization: University of Chicago School of Law
X-Newsreader: Yet Another NewsWatcher 2.4.0
References: <kjfair-ya02408000R0910970200590001@uchinews.uchicago.edu> <343F07B9.2340@-.->
Mime-Version: 1.0
Date: Sun, 12 Oct 1997 00:22:38 GMT
Lines: 69
Xref: news.lth.se rec.arts.int-fiction:30058


Recap: I have an object (representing various plants in a room) that I
want to provide some amusing responses for various actions (Take, Eat, etc.)
Unfortunately, actions such as Eat, Disrobe, and Unlock--all actions in
the Grammar with the "held" or "multiheld" token--generate an "implicit
take" for the object, i.e., they send a Take action to the object if the
player is not holding it.  This causes two problems: first, it generates a
"(first taking the <object>)" message; and second, it moves directly to
a Take action, thereby returning my amusing Take response rather than
my amusing Eat response.

I think I've generated a solution for this problem, but I'd like to see
if someone who is more knowledgeable about the parser can shoot a hole in it.

The solution revolves around the variable "action_to_be."  The parser stores
the action actually called for by the player here, rather than in the "action"
variable, which can contain other actions generated by the parser.

So to prevent the Take rule from stealing the Eat action, one can do this:

Object -> foliage "foliage"
  with  name "planter" "flowers" "flower" "planters"
          "plant" "plants" "foliage" "tree" "trees",
        description  "You admire the foliage and take a whiff of 
            the flowers.  Mmm, they smell lovely.  All of the foliage
            is certainly well kept.";
        react_before [;
          if (self == noun) {
            switch (action_to_be) {
              ##Eat: "You absently-mindedly snap off a leaf and pop it in 
                  your mouth, a move you regret instantly.  Blecch!  You
                  spit the leaf out and toss it back.";
                  }
               }
        ],
        before [;
          Smell: "You take a deep whiff of the plants.  The airy floral
              scent transports you for an instant back to happier days
              at home.";
          Take: "What, are you going to become a gardener now?";
       ],
   has scenery edible;   

If there is an implicit take, action_to_be will have the actual action,
in this case Eat, while action has the value Take.  Since the parser
has generated a <Take foliage> action, noun is set to foliage.  Thus
(self == noun) and the react_before rule traps only those actions 
affecting the foliage object.  Note that Smell works fine as before
since Smell does not generate an implicit take.


The first problem of the "(first taking the <object>)" message is easily
solved by a LibraryMessages object:

Object  LibraryMessages
  with  before [;
           Miscellany: if (lm_n == 26) {
               if (lm_o == foliage) rtrue; ! This keeps the (first taking)
                                           ! message from appearing
             }
        ];

See Section 21 of the Designer's Manual.

-- 
KEN FAIR - U. Chicago Law  | <http://student-www.uchicago.edu/users/kjfair>
Of Counsel, U. of Ediacara | Power Mac! | CABAL(tm) | I'm w/in McQ - R U?
 "My vulva is wide opened, a cave of darkness where the people . . . hang out 
  at night . . . .  There is music playing in my Womb."  - Doctress Neutopia


From fake-mail@anti-spam.address Thu Oct 23 09:33:01 MET DST 1997
Article: 30514 of rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!news.IAEhv.nl!news.cs.utwente.nl!news1.spinne.com!news-feed.inet.tele.dk!newsfeed.direct.ca!newsgate.direct.ca!van-52-0329.direct.ca!user
From: fake-mail@anti-spam.address (Neil K.)
Newsgroups: rec.arts.int-fiction
Subject: Re: AMNESIA by Thomas Disch
Date: Wed, 22 Oct 1997 19:57:35 -0700
Organization: Capn Grocibles Discount Groceries
Lines: 30
Message-ID: <fake-mail-2210971957350001@van-52-0329.direct.ca>
References: <62lul1$c30@drn.zippo.com>
NNTP-Posting-Host: van-52-0329.direct.ca
X-Newsreader: MT-NewsWatcher 2.3.1
x-no-archive: yes
Xref: news.lth.se rec.arts.int-fiction:30514

daryl@cogentex.com (Daryl McCullough) wrote:

> I found an interview with the science-fiction writer Thomas Disch (I hope he
> doesn't mind being called that, but the books that I've read by him are
> what I would call science fiction) about an interactive fiction that he
> wrote called "AMNESIA" back in 1993 (or earlier).
> 
> Has anyone played it, and does anyone know how to get a copy of it, and
> can anyone summarize it?

 It was available for the PC, I bought a copy from someone off the net,
but it's unlikely you'll find pirated versions available because it's
heavily copy-protected. It dates back to 1986.

 In that interview, as I recall, Disch moans and groans about how *awful*
and *stupid* and *unappreciative* the game-buying public were not to buy
zillions of copies of his brilliant masterwork, but the general consensus
on this newsgroup has been, again as I recall, that it was a heavily
flawed piece of junk. Not least amongst its many problems being the fact
that you kept starving to death every 5 turns.

 One day I'll dig up a second-hand 5.25" drive from somewhere so I can try
it again. Very cheesy packaging, not up to Infocom standards, though it
was clearly an attempt to compete with them.

 - Neil K.

-- 
 t e l a  computer consulting + design   *   Vancouver, BC, Canada
      web: http://www.tela.bc.ca/tela/   *   email: tela @ tela.bc.ca


From jcompton@typhoon.xnet.com Thu Oct 23 09:33:17 MET DST 1997
Article: 30518 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!www.nntp.primenet.com!globalcenter0!news.primenet.com!nntp.primenet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news-peer.sprintlink.net!news-pull.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!198.147.221.37!news.xnet.com!typhoon.xnet.com!not-for-mail
From: jcompton@typhoon.xnet.com (Jason Compton)
Newsgroups: rec.arts.int-fiction
Subject: Re: AMNESIA by Thomas Disch
Date: 23 Oct 1997 04:23:35 GMT
Organization: XNet - The Midwest's Leading Network Service Provider - (630) 983-6064
Lines: 30
Message-ID: <62mjg7$oar$2@flood.xnet.com>
References: <62lul1$c30@drn.zippo.com> <fake-mail-2210971957350001@van-52-0329.direct.ca>
NNTP-Posting-Host: typhoon.xnet.com
X-Trace: flood.xnet.com 877580615 24923 (None) 198.147.221.70
X-Complaints-To: news-admin@news.xnet.com
X-Newsreader: TIN [UNIX 1.3 950824BETA PL0]
Xref: news.lth.se rec.arts.int-fiction:30518

Neil K. (fake-mail@anti-spam.address) wrote:
: 
:  In that interview, as I recall, Disch moans and groans about how *awful*
: and *stupid* and *unappreciative* the game-buying public were not to buy
: zillions of copies of his brilliant masterwork, but the general consensus
: on this newsgroup has been, again as I recall, that it was a heavily
: flawed piece of junk. Not least amongst its many problems being the fact
: that you kept starving to death every 5 turns.

Yeah.  And the multiple copyright dates on Amnesia lead me to believe that
Disch kept trying and trying and trying to publish the thing in novel form
and kept failing and failing and failing, so his mediocre story got turned
into a mediocre games.

:  One day I'll dig up a second-hand 5.25" drive from somewhere so I can try
: it again. Very cheesy packaging, not up to Infocom standards, though it
: was clearly an attempt to compete with them.

Electronic Arts manuals were really inconsistent in quality.  Sometimes
they were very well done, other times they looked like they were printed
out on dot-matrix printers.

I have the original 64 version and even the damn copy protection wheel,
but very little motivation to pull the game out and try playing it again.

-- 
Jason Compton                                jcompton@xnet.com
Editor-in-Chief, Amiga Report Magazine       Anchor, Amiga Legacy
http://www.cucug.org/ar/                     http://www.xnet.com/~jcompton/
There are only dreams...                     ...like any other.


From dbs@sol33.cs.wisc.edu Thu Oct 23 22:47:47 MET DST 1997
Article: 30552 of rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!wuff.mayn.de!cosy.sbg.ac.at!newsfeed.Austria.EU.net!newscore.univie.ac.at!newsfeed.nacamar.de!uninett.no!news.maxwell.syr.edu!ais.net!iagnet.net!news.altair.com!uwvax!sol33.cs.wisc.edu!dbs
From: dbs@sol33.cs.wisc.edu (Daniel Shiovitz)
Newsgroups: rec.arts.int-fiction
Subject: Re: [IfComp97] Derivative games (Tempest)
Date: 23 Oct 1997 17:33:29 GMT
Organization: U of Wisconsin CS Dept
Lines: 27
Message-ID: <62o1p9$pi7@spool.cs.wisc.edu>
References: <Pine.SGI.3.95L.971023141415.27234B-100000@ebor.york.ac.uk>
NNTP-Posting-Host: sol33.cs.wisc.edu
Xref: news.lth.se rec.arts.int-fiction:30552

In article <Pine.SGI.3.95L.971023141415.27234B-100000@ebor.york.ac.uk>,
Den of Iniquity  <dmss100@york.ac.uk> wrote:
[..]
>Lists - well, people were too busy bowing down and prostrating themselves
>before an idol carved in your image to really sit up and take notice. 
>Although Lists fairly clearly lies in the realm of 'non-games', it's
>practically impossible to define a game for the purposes of excluding a
>'non-game' from the competition, so attempting to draw up such rules would
>be a pointless exercise. 

Dibs on the title "Escape from the Temple of the Zarf God" for next
year's comp.

(The cultists chant "Owah owah owah, Zarf, hear our humble plea. We
 are outside the metal gate and can't get through. We have a red fruit
 and a brick. Please, enlighten us with a portion of your thought
 processes, great and all-knowing yet incomprehensible to mortal man."
 Suddenly, a lightning bolt blasts down from the heavens, incinerating
 them instantly. "NO HINTS!" booms a gigantic voice.)

[..]
>Den (can't stand offal myself)
-- 
(Dan Shiovitz) (dbs@cs.wisc.edu) (look, I have a new e-mail address)
(http://www.cs.wisc.edu/~dbs) (and a new web page also)
(the content, of course, is the same)



From boutel1g@wcc.govt.nz Fri Oct 24 09:23:46 MET DST 1997
Article: 30556 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!newsfeed.internetmci.com!199.0.154.56!ais.net!logbridge.uoregon.edu!news.uoregon.edu!news.hn.netlink.co.nz!comp.vuw.ac.nz!wcc.govt.nz!wcc.govt.nz!nntp
Newsgroups: rec.arts.int-fiction
Subject: Re: [IfComp97] Derivative games (Tempest)
Message-ID: <01bcdff2$4a4448c0$67330a0a@WC034319.wcc.govt.nz>
From: "Giles Boutel" <boutel1g@wcc.govt.nz>
Date: 24 Oct 97 09:29:03 NZST
References: <Pine.SGI.3.95L.971023141415.27234B-100000@ebor.york.ac.uk> <62o1p9$pi7@spool.cs.wisc.edu>
Organization: Wellington City Council
Nntp-Posting-Host: 10.10.51.103
X-Newsreader: Microsoft Internet News 4.70.1161
Lines: 22
Xref: news.lth.se rec.arts.int-fiction:30556



Daniel Shiovitz <dbs@sol33.cs.wisc.edu> wrote in article
<62o1p9$pi7@spool.cs.wisc.edu>...


> Dibs on the title "Escape from the Temple of the Zarf God" for next
> year's comp.
> 
> (The cultists chant "Owah owah owah, Zarf, hear our humble plea. We
>  are outside the metal gate and can't get through. We have a red fruit
>  and a brick. Please, enlighten us with a portion of your thought
>  processes, great and all-knowing yet incomprehensible to mortal man."
>  Suddenly, a lightning bolt blasts down from the heavens, incinerating
>  them instantly. "NO HINTS!" booms a gigantic voice.)

I can see it now - a whole SoFarian religion based not around the question
of 'how did we begin?' but 'what the hell is the ending all about?"

-Giles

(does this count as hubris? I loved the ending, really I did)


From spatula@javanet.com.andbacon Fri Oct 24 15:34:18 MET DST 1997
Article: 30480 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news-ge.switch.ch!newscore.univie.ac.at!europa.clark.net!208.134.241.18!newsfeed.internetmci.com!208.134.60.8!news.javanet.com!spatula
From: spatula@javanet.com.andbacon (tv's Spatch)
Newsgroups: rec.arts.int-fiction
Subject: Re: ifMUD - what is this exactly?
Date: 22 Oct 1997 15:34:32 GMT
Organization: The Group W Bench
Lines: 21
Message-ID: <slrn64s785.gsc.spatula@kona.javanet.com>
References: <01bcded8$0772d680$684812ac@MyPC.futurenet.co.uk>
NNTP-Posting-Host: kona.javanet.com
X-Newsreader: slrn (0.9.4.3 UNIX)
Xref: news.lth.se rec.arts.int-fiction:30480

Wormer, he's a dead man!  Marmalard: dead!  Tony Ellis: 
> I've seen several references to the ifMUD - can some kind person tell me
> what this is and how I can enter it?

The ifMUD is a magical place where elvenfolk sit in hollow trees and make
cookies all day long.  No, wait.  The ifMUD is a magical place where
wizards ramble about, frotzing anything that moves.  Well, perhaps not.
The ifMUD is a magical place where strange powers and Perl collide to
provide a meeting place for IF authors, players, and random passers-by.
It's accessible via the Web at http://fovea.retina.net/ifmud for a Java
client, or you can telnet to fovea.retina.net 4000 and enjoy the fun right
there.  Or something.


- spatch, semi-kinda-maybe spokesperson and guy with pointy hat -

-- 
spatula@javanet.com.andbacon is but one of many Spatula tentacles. Rar.
         More fun than hax0ring E's random sig script -- PUTPBAD!
Let http://www.javanet.com/~spatula/booth.html show you how it's done.
       "And I say that holding my nose.  Deliberately." - H. Jon Benjamin


From ldaly@cs.bu.edu Fri Oct 24 15:34:19 MET DST 1997
Article: 30517 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!feed1.news.erols.com!cpk-news-hub1.bbnplanet.com!cam-news-feed2.bbnplanet.com!news.bbnplanet.com!news.bu.edu!ldaly
From: ldaly@cs.bu.edu (Art Gecko)
Newsgroups: rec.arts.int-fiction
Subject: Re: ifMUD - what is this exactly?
Date: 23 Oct 1997 03:51:04 GMT
Organization: The House of Erotic Massage
Lines: 18
Message-ID: <62mhj8$qti$1@news1.bu.edu>
References: <01bcded8$0772d680$684812ac@MyPC.futurenet.co.uk> <62l97c$qdd$1@joe.rice.edu> <slrn64sd4s.gsc.spatula@kona.javanet.com>
NNTP-Posting-Host: csa.bu.edu
X-Newsreader: TIN [version 1.2 PL2]
Xref: news.lth.se rec.arts.int-fiction:30517

tv's Spatch (spatula@javanet.com.andbacon) wrote:
: Wormer, he's a dead man!  Marmalard: dead!  Lucian Paul Smith: 

: > The best way to get an account on the ifMUD is via the web:
: > 
: > http://fovea.retina.net:4001/
: > 

: I knew I had forgotten one of the gateways.  Zounds!

There are a great many gateways, my son.  All will be revealed to you.

--Liza

--
Visit the ifMUD and shout "Timing!":
http://fovea.retina.net:4001/



From carl@earthweb.com Sat Oct 25 18:49:30 MET DST 1997
Article: 30619 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!newsfeed.ecrc.net!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!newsfeed.internetmci.com!199.201.128.9!news.silcom.com!news.genuity.net!not-for-mail
From: Carl Muckenhoupt <carl@earthweb.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: AMNESIA by Thomas Disch
Date: Thu, 23 Oct 1997 15:05:27 -0400
Organization: EarthWeb
Lines: 54
Message-ID: <344F9FF7.2631@earthweb.com>
References: <62lul1$c30@drn.zippo.com> <fake-mail-2210971957350001@van-52-0329.direct.ca>
Reply-To: carl@earthweb.com
NNTP-Posting-Host: ben.earthweb.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.0 (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:30619

Neil K. wrote:
> 
> daryl@cogentex.com (Daryl McCullough) wrote:
> 
> > Has anyone played it, and does anyone know how to get a copy of it, and
> > can anyone summarize it?
> 
>  It was available for the PC, I bought a copy from someone off the net,
> but it's unlikely you'll find pirated versions available because it's
> heavily copy-protected. It dates back to 1986.
> 
>  In that interview, as I recall, Disch moans and groans about how *awful*
> and *stupid* and *unappreciative* the game-buying public were not to buy
> zillions of copies of his brilliant masterwork, but the general consensus
> on this newsgroup has been, again as I recall, that it was a heavily
> flawed piece of junk. Not least amongst its many problems being the fact
> that you kept starving to death every 5 turns.

I'll second that.  Back when I played the game, I was quite fond of
Disch's sci-fi works.  And indeed, the story isn't bad.  You wake up in
a hotel room in New York with no memory of who you are or how you got
there.  You have two parallel goals: survive, and piece together your
past.  To do both, you need to find people who used to know you. 
However, it turns out that your amnesia is a chronic condition, so most
of the people who know you don't know much more about you than you do. 
Ultimately, you learn most of your story - including crucial details
about the reason for your condition - from a diskette containing your
old lab notes.

Unfortunately, the gameplay is lousy.  You spend far more time doing
boring, repetetive things, like wandering through Manhattan, or
acquiring and consuming food, than you spend advancing the plot.  When
you do get a chance to advance the plot, the chances are you'll blow it
because of something you didn't know to be important.  There's an essay
about Planetfall that you can find from Stephen Van Egmond's site,
complaining about how it deliberately produces tedium and confusion. 
The same complaint can be made about Amnesia ten times over.

Furthermore, Disch did himself a disservice by refusing to admit this. 
I recall an interview - I don't know if it's the same one you cite - in
which he claimed that the game was highly innovative, and unappreciated
because it didn't conform to the players' expectations.  This is the
artist's equivalent of "I was just following orders": an excuse that is
universally applicable, but never satisfactory.  I have seen other game
authors use the same excuse, and in some cases I can see some basis for
it.  _Return to Zork_, for example, tried very hard to give players a
completely new experience, and part of the reason that it failed as
badly as it did was that it did not adequately communicate to the player
how it was meant to be played.  This is still the fault of the game
itself rather than the players - after all, did we have trouble catching
on that "Lemmings" can't be played like "Super Mario Brothers"?  But at
least we can identify in hindsight what the authors of RTZ were reaching
for.  In Amnesia, I'm still unsure what it was that Disch thought was so
innovative.  I hope he wasn't referring to the tedious parts.


From graham@gnelson.demon.co.uk Sun Oct 26 13:45:27 MET 1997
Article: 30646 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!4.1.16.34.MISMATCH!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news.maxwell.syr.edu!netnews.com!news2.euro.net!bullseye.news.demon.net!demon!news.demon.co.uk!demon!gnelson.demon.co.uk!graham
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: [IFcomp97] Questions and Observations (no spoilers)
Date: Sun, 26 Oct 1997 02:36:18 +0100 (BST)
Organization: none
Message-ID: <ant260118bc8M+4%@gnelson.demon.co.uk>
References: <Pine.SGI.3.95L.971022155839.12841D-100000@ebor.york.ac.uk> <62t3ir$s5e@drn.zippo.com> <erkyrathEIMD1K.G69@netcom.com> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 30
Xref: news.lth.se rec.arts.int-fiction:30646

In article <erkyrathEIMD1K.G69@netcom.com>, Andrew Plotkin
<URL:mailto:erkyrath@netcom.com> wrote:
> Daryl McCullough (daryl@cogentex.com) wrote:
> > Yes. The rule is that each winner picks one prize, but I wonder whether
> > it may have been better to lump several prizes together. For
> > the purpose of promoting the contest and encouraging excellence, does
> > it work better to have a small number of winners who get big prizes,
> > or a larger number of winners who get small prizes?
> 
> I would say the latter.

Agreed.

> Historically, we've *had* the latter. (I think the biggest prize has
> always been US$100.) Call me cynical, but I really doubt we would have 
> seen a better contest if the prize had been $1000. 

I actually felt slightly guilty cashing the cheque, too (though I
applaud Martin Braun's generosity, especially as he insisted on
paying the extra cost of writing the cheque in sterling).

I mean, I got over it and bought myself a luxury-model wok... with
which I cooked this evening's dinner, incidentally, so you could say
I'm still living off the proceeds from "Sherbet".  But the range and
oddballness of prizes is much more us, somehow.  It could almost be
the list of parts needed in "Leather Goddesses of Phobos".

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From mol@bartlet.df.lth.se Sun Oct 26 13:53:14 MET 1997
Article: 30648 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: 26 Oct 1997 13:51:58 +0100
Organization: Societas Datori Universitatis Lundensis
Lines: 366
Message-ID: <62vede$psn$1@bartlet.df.lth.se>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:30648


What follows is more of an essay than an ordinary Usenet post, mostly
a bit theoretical, but leading up to a concrete proposal for an
improvement of the Inform Library. Hope you don't find it too
long-winded - it may provide some change from the Contest discussions.


If you've been following my earlier posts in the "Skipping
disambiguation" thread, I'm looking at ways of implementing a more
convenient (for the programmer) conversation system in Inform.

This problem is of course not limited to Inform, so I'm not putting an
"[INFORM]" marker in the subject, but my practical problem right now
is how to do this in Inform. The general conclusions should be valid
for all languages that implement a "Ask troll about axe"-style
conversation.

Before we start, let me add that I'm aware of menu-driven
conversation, and its drawbacks and advantages. For reasons of my own,
I don't want such an interface. Please don't tell me that I _should_
want it :-).

=== A First Attempt at Conversation ===

A first, naive, attempt at conversation is to implement the "ask" verb
as any other verb. In Inform, the grammar line would be

verb 'ask' * creature 'about' noun -> Ask;

which, to aid the Informally challenged, means that the parser accepts
commands of the type "ask <creature> about <noun>", where <creature>
is any actor that's in scope, and <noun> is the name of any object in
scope.

This has, of course, the obvious drawback that you can only
talk about things you can see, which may add some unconscious humour
to the game, but is hardly convenient.

The next step is to change the scope definition for the "noun". In
Inform, it's easy to write a grammar line that accepts any noun in the
game, not just those in scope. This allows you to talk about things
not in sight, but it leads to another undesirable effect (supposing
that there are two axes in the game):

| > Ask troll about axe.
|
| Which axe do you mean, the troll's axe or the magic axe of Frobozz?

First, this is the parser speaking, not the troll. This may - or may
not - break mimesis.

** Question #1 to the audience: would you find it disturbing to have the
** parser interrupt your dialogue with the troll in this way?

If you think that it is disturbing, it is possible, but far from
trivial to make the Inform parser put its words in the mouth of the
troll:

| > Ask troll about axe.
|
| The troll says: "Which axe do you mean, my axe or the magic axe
| of Frobozz?"

I'm not sure this is an improvement, though, for two reasons: firstly,
the response must be personalized, so it sounds like the troll's
voice, not like Graham Nelson's (note that it says "my axe", not "the
troll's axe"). And this may be hard; adventure trolls are usually not
as grammatical as my example, so you might prefer that it mangle the
response in some way.

=== On the Problems Caused by Disambiguation ===

More seriously, we now have an omniscient troll that knows about every
object in the game. This can spoil the game (never mind the troll,
what if the _player_ isn't supposed to *know* that there is a magic
axe?). It can also lead to more unconscious comedy:

| > Ask troll about underwear.
|
| The troll says: "Which underwear do you mean, the princess' lace
| undies or your Mickey Mouse Y-fronts?"

(Again, one is lead to wonder: how does the troll know all this?)


One could of course limit the definition of scope to certain
well-defined "conversation topics" that the player is supposed to know
about, possibly including all objects that the player has already
seen. In this case, disambiguation wouldn't be running wild like
above. We'd still get questions from the parser (or from the troll):

| > Ask troll about swat.
|
| The troll asks you which "swat" you mean, the Akond of Swat or
| the special police force.

** Question #2: Would you prefer to get such disambiguation questions
** from the parser or from the troll?

I think such a system would work. However, it would require a lot of
foresight by the game author. For example, in the case of the axes
above, if the player isn't supposed to know about the Magic Axe, it
can't be defined as a conversation topic, or the disambiguator will
reveal its existence. And if the player hasn't seen the troll's axe
yet (because the troll is hiding it behind his back until he can get
close enough to decapitate the player), it shouldn't be a topic
either.

But should this mean that the player should be unable to converse
about axes in general? Probably not; so you'd have to add "abstract"
conversation topics like "axes-in-general". This could easily become
quite complicated; in the general case, we'd need to build up some
knowledge representation of what is known to the player, to the troll,
and what the troll wants to reveal to the player. Of course, realistic
NPC interaction *is* a Turing-complete problem, so we shouldn't expect
it to be easy...

=== The Case of Encyclopedias ===

There are a couple of related cases where disambiguation is more
clearly the right thing to do. Consider, for example, an encyclopedia:

| > Consult encyclopedia about London.
|
| Which London do you mean: London, England; London, Ontario; or Jack
| London?

In this case, I think it's very natural that you have a set of
"encyclopedia subjects" and that the parser disambiguates between
them. After all, this is how real-world interaction with an
encyclopedia works: once you start browsing it, you know which
subjects it contains and you have to choose which "London" to read
about. (I'm of course referring to conventional encyclopedias, not
quasi-intelligent quasi-NPC's like the Hitchhiker's Guide in the
eponymous game).

Gareth Rees has implemented a system for encyclopedias that work like
this; it's available from the IF-archive as frobozzica.inf (under
inform5.5/examples). It also handles some other problems, such as not
revealing the existence of topcis when disambiguating the _indirect_
object of "look up" (i.e. if "zork" is mentioned in a book that you
haven't seen yet, you shouldn't get disambiguation messages like "What
do you want to look up the zork in" that tell you that *some* book
mentions Zork, even though the ones you have seen don't).


=== Telling vs. Asking: the Difference in Scope ===

I also have a feeling that disambiguation may be more appropriate for
the telling than for asking. To me, it feels more natural that the
parser requires me to be specific in the case of

| > Tell troll about coin.
|
| Which coin do you mean, the gold coin or the silver coin?

than in

| > Ask troll about coin.
|
| Which coin do you mean, the gold coin, the silver coin, or the
| platinum coin in the king's pocket?

Because if you think about it, the action of telling is fundamentally
different from that of asking when it comes to scope: the scope of a
question is the entire universe of discourse, while the scope of
telling is usually limited to what you know.

In programming terms, it seems much easier to model the notion of
"subjects you know about and may want to tell people about", than
"subjects that you might want to ask people about".

** Question #3: What are your views on this? Am I just
** over-theorizing, or should there really be a difference in the
** notion of scope between "ask" and "tell"?


=== What TADS Does === 

The standard TADS parser omits the disambiguation altogether, and
simply chooses the first object in the game that matches the input. So
in our axe example, when the player types "ask troll about axe", the
troll really sees the question as either "ask troll about troll's axe"
or "ask troll about magic axe", depending on which of the objects is
first found by the parser.

This avoids all the problems with the methods I've described
previously. However, it creates two new problems.

The first problem is that the programmer can't know which of the
objects will actually be sent to the troll's "ask" code. Or perhaps he
or she can deduce it from the order words are defined, but code that
assumes anything about the order of word definitions int he internal
parser tables will be very fragile and hard to maintain - a small,
seemingly insignificant change to the code can lead to the order
between the axes being reversed.

So the only sensible way out is for the programmer to perform an
explicit "reverse disambiguation" (ambiguation?): the code will contain
lots of tests like

if (topic = troll_axe or topic = magic_axe) 

which is OK if there are only two objects that have the noun "axe",
but what if the word in question is "door" and there are 69 doors in
the game? (OK, the programmer would probably have to derive all the
doors from a common base class and check the class of the topic, but
that's a bit of a hack in this context).

More seriously, the troll's actions will have to take this into
account. Even if topic = magic_axe, the player may still have *meant*
the troll's axe, so the troll can't start talking about the magic
axe.

=== What if the Troll Misunderstands You? ===

Or perhaps he can? Perhaps NPC's, just as real people, should
interpret questions as they want, not the way the questioner wants? If
you "ask troll about axe", and the troll is thinking about the magic
axe, rather than about the axe he's absent-mindedly toting around,
shouldn't the troll reply about what's occupying his mind?

| > Ask troll about axe.
|
| The troll gets an eager glint in his eyes. "Yes, the Magic Axe of
| Frobozz! I've been searching for it for years. Do you know where it
| is?"
|
| The troll suddenly seems to notice the axe in his hand. "Or did you
| mean _this_ old thing? It's good for chopping up adventurers, but
| it's not magic."

The problem with this is, unfortunately, that it won't work: If the
magic axe happens to be first in the parser's tables, "Ask troll about
magic axe" would lead to the very same message being sent to the
troll, and in that case the reply would be totally inappropriate.

So this would lead to the conclusion that the "TADS" method of
conversation isn't the ultimate solution, either. (To be fair to TADS,
there are ways of getting around this, looking at what the player
really typed, but then we're leaving the attractive simplicity of the
"no disambiguation" method far behind).

=== Deferring Disambiguation ===

A solution to this problem would be the following:

When parsing a command like "ask troll about axe", don't perform any
disambiguation, but don't cheat by just choosing the first object that
matches "axe". Instead, send the NPC a *list* of all the objects that
match the input.

In the TADS case, the troll's doAskAbout method would not receive the
single object magic_axe (or troll_axe, depending on the state of the
parser tables), but rather the list [ magic_axe, troll_axe ]. (In
Inform, which doesn't have any list datatype, one would have to use an
array instead, or a different interface altogether, but it would be
possible to solve).

The disambiguation would then be performed by the troll, rather than
by the parser. The troll's code could then do several different things:

* It could do as in current TADS and just check the first member
  of the list.

* It could see if the magic axe was mentioned somewhere in the list,
  and in that case (being precoocupied with magic axes for the moment)
  go off on its rant, or if it isn't just say something about its own
  axe.

* It could confess to being confused and ask the player.

| > Ask troll about coins.
|
| The troll looks puzzled. "Are you talking about gold or silver
| coins? Makes a lot of difference, you know."

Of course, in this last case it would be nice to be able to use the
parser's normal disambiguation routines, so that it would suffice to
say "gold" rather than rephrasing the entire question.

** Question #4: If the troll asks a dismabiguation question like this
** last example, would it be very inconvenient if you had to rephrase
** the entire question?


=== Deferring Parsing: How Inform Does It ===

The Inform Library chooses the most radical solution of all: don't
even try to parse NPC conversation, just pass the entire input on to
the NPC code. (This is not entirely true: the parser *does* parse the
first word in a question, but that's not very helpful if it is, say,
a common adjective like "big").

This solution has the obvious advantage of being maximally flexible.

It is also a major pain for the programmer, who in effect has to
write his or her own parser. And this is a non-trivial task: suppose
the Magic Axe of Frobozz has the "name" property

name 'magic' 'axe' 'of' 'frobozz'.

Then *any* combination of those words can be used to refer to it. And
objects can have parse_name routines that perform any strange
matching, and there are global entyr points that can be used to change
the way nouns are parsed.


=== Some Concrete Proposals re Inform ===

This situation would be very much simplified if the Inform parser were
user-callable. I think it would be sufficient to have a simple
predicate

MatchesInput(object)

that returned true if the object's name matched the input (or, rather,
the "consult topic" set up by the parser before sending the Ask
message to the actor), and false if it didn't.

That way, the troll object could perform its own parsing simply by
calling MatchesInput on all objects it had information about.

An alternative would be to have a function

NextMatchingObject()

that iterated over the object tree and returned the object number of
the next object that matched the input, and 0 if no more objects
matched. This would be a substitute for sending a list of matching
words.


Note: As Andrew Plotkin has pointed out (he proposed the equivalent of
the MatchesInput function in a post in the "Skipping Disambiguation"
thread), a function such as MatchesInput would not be very difficult
to write. However, I feel that such a function should really be an
official part of the Library. If nothing else, that would ensure that
when the Library changes its way of parsing words, the MatchesInput
function would change accordingly. The same holds for
NextMatchingObject, if one chooses to implement that function.

=== Conclusions ===

Above, I've looked at the pros and cons of a number of ways of
implementing NPC conversation, in particular the "ask" verb, paying
special attention to issues of scope and disambiguation. Some of the
proposed methods are clearly insufficent, while others are sufficient
but require much effort by the programmer.

I was earlier (in the "Skipping Disambiguation" thread) considering
implementing the "TADS method" of conversation in Inform. However, I
must draw the conclusion that that method is not entirely
satisfactory, and that it is better to defer disambiguation and let
the NPC object handle that.

The Inform method of leaving parsing to the NPC object is by far the
most flexible. However, in order to reduce the workload on the
programmer it would be desirable to make the internals of the parser
user-callable. Two possible interfaces for such a callable parser are
proposed.

-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From mol@bartlet.df.lth.se Mon Oct 27 10:21:15 MET 1997
Article: 30648 of rec.arts.int-fiction
Path: news.lth.se!not-for-mail
From: mol@bartlet.df.lth.se (Magnus Olsson)
Newsgroups: rec.arts.int-fiction
Subject: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: 26 Oct 1997 13:51:58 +0100
Organization: Societas Datori Universitatis Lundensis
Lines: 366
Message-ID: <62vede$psn$1@bartlet.df.lth.se>
NNTP-Posting-Host: bartlet.df.lth.se
NNTP-Posting-User: mol
Xref: news.lth.se rec.arts.int-fiction:30648


What follows is more of an essay than an ordinary Usenet post, mostly
a bit theoretical, but leading up to a concrete proposal for an
improvement of the Inform Library. Hope you don't find it too
long-winded - it may provide some change from the Contest discussions.


If you've been following my earlier posts in the "Skipping
disambiguation" thread, I'm looking at ways of implementing a more
convenient (for the programmer) conversation system in Inform.

This problem is of course not limited to Inform, so I'm not putting an
"[INFORM]" marker in the subject, but my practical problem right now
is how to do this in Inform. The general conclusions should be valid
for all languages that implement a "Ask troll about axe"-style
conversation.

Before we start, let me add that I'm aware of menu-driven
conversation, and its drawbacks and advantages. For reasons of my own,
I don't want such an interface. Please don't tell me that I _should_
want it :-).

=== A First Attempt at Conversation ===

A first, naive, attempt at conversation is to implement the "ask" verb
as any other verb. In Inform, the grammar line would be

verb 'ask' * creature 'about' noun -> Ask;

which, to aid the Informally challenged, means that the parser accepts
commands of the type "ask <creature> about <noun>", where <creature>
is any actor that's in scope, and <noun> is the name of any object in
scope.

This has, of course, the obvious drawback that you can only
talk about things you can see, which may add some unconscious humour
to the game, but is hardly convenient.

The next step is to change the scope definition for the "noun". In
Inform, it's easy to write a grammar line that accepts any noun in the
game, not just those in scope. This allows you to talk about things
not in sight, but it leads to another undesirable effect (supposing
that there are two axes in the game):

| > Ask troll about axe.
|
| Which axe do you mean, the troll's axe or the magic axe of Frobozz?

First, this is the parser speaking, not the troll. This may - or may
not - break mimesis.

** Question #1 to the audience: would you find it disturbing to have the
** parser interrupt your dialogue with the troll in this way?

If you think that it is disturbing, it is possible, but far from
trivial to make the Inform parser put its words in the mouth of the
troll:

| > Ask troll about axe.
|
| The troll says: "Which axe do you mean, my axe or the magic axe
| of Frobozz?"

I'm not sure this is an improvement, though, for two reasons: firstly,
the response must be personalized, so it sounds like the troll's
voice, not like Graham Nelson's (note that it says "my axe", not "the
troll's axe"). And this may be hard; adventure trolls are usually not
as grammatical as my example, so you might prefer that it mangle the
response in some way.

=== On the Problems Caused by Disambiguation ===

More seriously, we now have an omniscient troll that knows about every
object in the game. This can spoil the game (never mind the troll,
what if the _player_ isn't supposed to *know* that there is a magic
axe?). It can also lead to more unconscious comedy:

| > Ask troll about underwear.
|
| The troll says: "Which underwear do you mean, the princess' lace
| undies or your Mickey Mouse Y-fronts?"

(Again, one is lead to wonder: how does the troll know all this?)


One could of course limit the definition of scope to certain
well-defined "conversation topics" that the player is supposed to know
about, possibly including all objects that the player has already
seen. In this case, disambiguation wouldn't be running wild like
above. We'd still get questions from the parser (or from the troll):

| > Ask troll about swat.
|
| The troll asks you which "swat" you mean, the Akond of Swat or
| the special police force.

** Question #2: Would you prefer to get such disambiguation questions
** from the parser or from the troll?

I think such a system would work. However, it would require a lot of
foresight by the game author. For example, in the case of the axes
above, if the player isn't supposed to know about the Magic Axe, it
can't be defined as a conversation topic, or the disambiguator will
reveal its existence. And if the player hasn't seen the troll's axe
yet (because the troll is hiding it behind his back until he can get
close enough to decapitate the player), it shouldn't be a topic
either.

But should this mean that the player should be unable to converse
about axes in general? Probably not; so you'd have to add "abstract"
conversation topics like "axes-in-general". This could easily become
quite complicated; in the general case, we'd need to build up some
knowledge representation of what is known to the player, to the troll,
and what the troll wants to reveal to the player. Of course, realistic
NPC interaction *is* a Turing-complete problem, so we shouldn't expect
it to be easy...

=== The Case of Encyclopedias ===

There are a couple of related cases where disambiguation is more
clearly the right thing to do. Consider, for example, an encyclopedia:

| > Consult encyclopedia about London.
|
| Which London do you mean: London, England; London, Ontario; or Jack
| London?

In this case, I think it's very natural that you have a set of
"encyclopedia subjects" and that the parser disambiguates between
them. After all, this is how real-world interaction with an
encyclopedia works: once you start browsing it, you know which
subjects it contains and you have to choose which "London" to read
about. (I'm of course referring to conventional encyclopedias, not
quasi-intelligent quasi-NPC's like the Hitchhiker's Guide in the
eponymous game).

Gareth Rees has implemented a system for encyclopedias that work like
this; it's available from the IF-archive as frobozzica.inf (under
inform5.5/examples). It also handles some other problems, such as not
revealing the existence of topcis when disambiguating the _indirect_
object of "look up" (i.e. if "zork" is mentioned in a book that you
haven't seen yet, you shouldn't get disambiguation messages like "What
do you want to look up the zork in" that tell you that *some* book
mentions Zork, even though the ones you have seen don't).


=== Telling vs. Asking: the Difference in Scope ===

I also have a feeling that disambiguation may be more appropriate for
the telling than for asking. To me, it feels more natural that the
parser requires me to be specific in the case of

| > Tell troll about coin.
|
| Which coin do you mean, the gold coin or the silver coin?

than in

| > Ask troll about coin.
|
| Which coin do you mean, the gold coin, the silver coin, or the
| platinum coin in the king's pocket?

Because if you think about it, the action of telling is fundamentally
different from that of asking when it comes to scope: the scope of a
question is the entire universe of discourse, while the scope of
telling is usually limited to what you know.

In programming terms, it seems much easier to model the notion of
"subjects you know about and may want to tell people about", than
"subjects that you might want to ask people about".

** Question #3: What are your views on this? Am I just
** over-theorizing, or should there really be a difference in the
** notion of scope between "ask" and "tell"?


=== What TADS Does === 

The standard TADS parser omits the disambiguation altogether, and
simply chooses the first object in the game that matches the input. So
in our axe example, when the player types "ask troll about axe", the
troll really sees the question as either "ask troll about troll's axe"
or "ask troll about magic axe", depending on which of the objects is
first found by the parser.

This avoids all the problems with the methods I've described
previously. However, it creates two new problems.

The first problem is that the programmer can't know which of the
objects will actually be sent to the troll's "ask" code. Or perhaps he
or she can deduce it from the order words are defined, but code that
assumes anything about the order of word definitions int he internal
parser tables will be very fragile and hard to maintain - a small,
seemingly insignificant change to the code can lead to the order
between the axes being reversed.

So the only sensible way out is for the programmer to perform an
explicit "reverse disambiguation" (ambiguation?): the code will contain
lots of tests like

if (topic = troll_axe or topic = magic_axe) 

which is OK if there are only two objects that have the noun "axe",
but what if the word in question is "door" and there are 69 doors in
the game? (OK, the programmer would probably have to derive all the
doors from a common base class and check the class of the topic, but
that's a bit of a hack in this context).

More seriously, the troll's actions will have to take this into
account. Even if topic = magic_axe, the player may still have *meant*
the troll's axe, so the troll can't start talking about the magic
axe.

=== What if the Troll Misunderstands You? ===

Or perhaps he can? Perhaps NPC's, just as real people, should
interpret questions as they want, not the way the questioner wants? If
you "ask troll about axe", and the troll is thinking about the magic
axe, rather than about the axe he's absent-mindedly toting around,
shouldn't the troll reply about what's occupying his mind?

| > Ask troll about axe.
|
| The troll gets an eager glint in his eyes. "Yes, the Magic Axe of
| Frobozz! I've been searching for it for years. Do you know where it
| is?"
|
| The troll suddenly seems to notice the axe in his hand. "Or did you
| mean _this_ old thing? It's good for chopping up adventurers, but
| it's not magic."

The problem with this is, unfortunately, that it won't work: If the
magic axe happens to be first in the parser's tables, "Ask troll about
magic axe" would lead to the very same message being sent to the
troll, and in that case the reply would be totally inappropriate.

So this would lead to the conclusion that the "TADS" method of
conversation isn't the ultimate solution, either. (To be fair to TADS,
there are ways of getting around this, looking at what the player
really typed, but then we're leaving the attractive simplicity of the
"no disambiguation" method far behind).

=== Deferring Disambiguation ===

A solution to this problem would be the following:

When parsing a command like "ask troll about axe", don't perform any
disambiguation, but don't cheat by just choosing the first object that
matches "axe". Instead, send the NPC a *list* of all the objects that
match the input.

In the TADS case, the troll's doAskAbout method would not receive the
single object magic_axe (or troll_axe, depending on the state of the
parser tables), but rather the list [ magic_axe, troll_axe ]. (In
Inform, which doesn't have any list datatype, one would have to use an
array instead, or a different interface altogether, but it would be
possible to solve).

The disambiguation would then be performed by the troll, rather than
by the parser. The troll's code could then do several different things:

* It could do as in current TADS and just check the first member
  of the list.

* It could see if the magic axe was mentioned somewhere in the list,
  and in that case (being precoocupied with magic axes for the moment)
  go off on its rant, or if it isn't just say something about its own
  axe.

* It could confess to being confused and ask the player.

| > Ask troll about coins.
|
| The troll looks puzzled. "Are you talking about gold or silver
| coins? Makes a lot of difference, you know."

Of course, in this last case it would be nice to be able to use the
parser's normal disambiguation routines, so that it would suffice to
say "gold" rather than rephrasing the entire question.

** Question #4: If the troll asks a dismabiguation question like this
** last example, would it be very inconvenient if you had to rephrase
** the entire question?


=== Deferring Parsing: How Inform Does It ===

The Inform Library chooses the most radical solution of all: don't
even try to parse NPC conversation, just pass the entire input on to
the NPC code. (This is not entirely true: the parser *does* parse the
first word in a question, but that's not very helpful if it is, say,
a common adjective like "big").

This solution has the obvious advantage of being maximally flexible.

It is also a major pain for the programmer, who in effect has to
write his or her own parser. And this is a non-trivial task: suppose
the Magic Axe of Frobozz has the "name" property

name 'magic' 'axe' 'of' 'frobozz'.

Then *any* combination of those words can be used to refer to it. And
objects can have parse_name routines that perform any strange
matching, and there are global entyr points that can be used to change
the way nouns are parsed.


=== Some Concrete Proposals re Inform ===

This situation would be very much simplified if the Inform parser were
user-callable. I think it would be sufficient to have a simple
predicate

MatchesInput(object)

that returned true if the object's name matched the input (or, rather,
the "consult topic" set up by the parser before sending the Ask
message to the actor), and false if it didn't.

That way, the troll object could perform its own parsing simply by
calling MatchesInput on all objects it had information about.

An alternative would be to have a function

NextMatchingObject()

that iterated over the object tree and returned the object number of
the next object that matched the input, and 0 if no more objects
matched. This would be a substitute for sending a list of matching
words.


Note: As Andrew Plotkin has pointed out (he proposed the equivalent of
the MatchesInput function in a post in the "Skipping Disambiguation"
thread), a function such as MatchesInput would not be very difficult
to write. However, I feel that such a function should really be an
official part of the Library. If nothing else, that would ensure that
when the Library changes its way of parsing words, the MatchesInput
function would change accordingly. The same holds for
NextMatchingObject, if one chooses to implement that function.

=== Conclusions ===

Above, I've looked at the pros and cons of a number of ways of
implementing NPC conversation, in particular the "ask" verb, paying
special attention to issues of scope and disambiguation. Some of the
proposed methods are clearly insufficent, while others are sufficient
but require much effort by the programmer.

I was earlier (in the "Skipping Disambiguation" thread) considering
implementing the "TADS method" of conversation in Inform. However, I
must draw the conclusion that that method is not entirely
satisfactory, and that it is better to defer disambiguation and let
the NPC object handle that.

The Inform method of leaving parsing to the NPC object is by far the
most flexible. However, in order to reduce the workload on the
programmer it would be desirable to make the internals of the parser
user-callable. Two possible interfaces for such a callable parser are
proposed.

-- 
Magnus Olsson (mol@df.lth.se, zebulon@pobox.com)
------    http://www.pobox.com/~zebulon   ------
     Not officially connected to LU or LTH.


From dmss100@york.ac.uk Mon Oct 27 20:59:57 MET 1997
Article: 30697 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!news.maxwell.syr.edu!howland.erols.net!recycled.news.erols.com!nntp.news.xara.net!xara.net!peer.news.nildram.co.uk!Aladdin!aladdin.net!ns2.aladdin.net!RMplc!rmplc.co.uk!yama.mcc.ac.uk!news.york.ac.uk!ebor.york.ac.uk!dmss100
From: Den of Iniquity <dmss100@york.ac.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: Mon, 27 Oct 1997 18:49:18 +0000
Organization: The University of York, UK
Lines: 139
Sender: dmss100@york.ac.uk
Message-ID: <Pine.SGI.3.95L.971027174633.14709A-100000@ebor.york.ac.uk>
References: <62vede$psn$1@bartlet.df.lth.se>
NNTP-Posting-Host: ebor.york.ac.uk
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
In-Reply-To: <62vede$psn$1@bartlet.df.lth.se>
Xref: news.lth.se rec.arts.int-fiction:30697

Hi, excellent post...

On 26 Oct 1997, Magnus Olsson wrote:
>What follows is more of an essay than an ordinary Usenet post

[and so on. Read the original post.]

>** Question #1 to the audience: would you find it disturbing to have the
>** parser interrupt your dialogue with the troll in this way?

No, I think we're already used to it. Avoiding it as much as possible is a
good thing but in the long run, it's hardly unusual.


>| > Ask troll about swat.
>|
>| The troll asks you which "swat" you mean, the Akond of Swat or
>| the special police force.

Who, or why, or which, or _what_ is the Akond of Swat?

>** Question #2: Would you prefer to get such disambiguation questions
>** from the parser or from the troll?

Ideally the troll, but putting questions into the voice of the troll could
be difficult. As in the above case where the question comes from the
troll, but without actually quoting him, well, that would work, but if
it's a stereotypically dull troll, it does seem a little incongruous to
use grammatically correct questions. Especially if the troll had to refer
to objects with complex or polysyllabic names: "The troll asks you which
book you mean, the big red book, the spell book or the encyclopaedia of
mediaeval architecture?"

Of course, the aforementioned omniscient troll could be quite fun as a
much maligned, very misunderstood character. :)

-> ASK TROLL ABOUT AXES 
- 
-The troll thinks for a moment, humming and hahing to himself, then, after
-scrabbling around on the floor for a chalky stone, he proceeds to scratch
-two straight crossing lines on the south wall. "Well," he says, 
-"conventionally, in two-dimensional space, we refer to the x-axis," he
-points out the horizontal line, "and the y-axis" he indicates the 
-vertical line. He pauses for a moment while he produces a delicate
-pince-nez from some concealed pocket and perches it precariously on his
-snout-like nose. "These are arranged perpendicularly such that any point
-in the two-dimensional plane can be identified by its distance from a
-given origin, here..."

>But should this mean that the player should be unable to converse
>about axes in general? Probably not; so you'd have to add "abstract"
>conversation topics like "axes-in-general". This could easily become
>quite complicated [...]

On a slightly different line to the one you're heading down, topics of
conversation should be separate from the objects to which they refer.
Perhaps every NPC should have an invisible bucket of concepts which they
can talk about. If one NPC tells another (N)PC about (say) 'crown_info_1',
then a duplicate of that piece of information is added to the new (N)PC's
bucket of knowledge, which they can then copy to another (N)PC's bucket,
and so on. This would be difficult to implement smoothly, perhaps a
programmer would prefer pointers to some global knowledge-net rather than
duplicates of the information itself. The end result is a world in which
the NPC's can trade gossip.

In the Delphine Software (Graphical) Game 'Cruise For A Corpse' (was it
given a different title in other countries?) a large part of the
advancement of plot involved obtaining information from one character and
then wandering across the ship to tell another character in order to glean
some more information which must then be told to yet another person - all
by menu-driven conversation, I must add. A little tedious, I have to say,
though thankfully interspersed with standard 'hunt the pixel' scenes and
'use everything on everything else' puzzles. This is one example of such a
system, though simplistic in that the NPC's didn't seem to do much talking
to each other and the number of options was limited. I think it's fair to
say that in it's limited usage it took the idea too far.

It does solve the problem of the troll knowing about the princess' panties
or the platinum coin, and so on.


>=== The Case of Encyclopedias ===
>
>There are a couple of related cases where disambiguation is more
>clearly the right thing to do. Consider, for example, an encyclopedia:
>
>| > Consult encyclopedia about London.
>|
>| Which London do you mean: London, England; London, Ontario; or Jack
>| London?

One minor point - does anyone else think that a specially tailored answer
like the following would be a better response? 

  "Under 'London' the index lists the following: [...] Which would you 
  like to look up?"

I suppose this comes down to the 'which would you prefer, the voice of the
parser or that of the troll' question, with the book replacing the troll.

[Back to my 'idea bucket' point - the book has its own special 'ideas
container' from which any character capable of reading can draw, if you'd
want to implement that.]


>** Question #3: What are your views on this? Am I just
>** over-theorizing, or should there really be a difference in the
>** notion of scope between "ask" and "tell"?

Actually, I don't think there's much of a difference. If you can ask about
something you should be able to tell about it. If you can ask about axes
'in general' then you should be able to tell about axes in general too. An
'ask' is just a 'tell' turned upside-down. Disambiguation should work
equally for both, I feel. Of course with tell, you've got less flexibility
about where the disambiguation is coming from. It would be a bit bizarre
if, before telling the troll about the coin, he asked which one you meant,
the gold or the silver one.


>** Question #4: If the troll asks a dismabiguation question like this
>** last example, would it be very inconvenient if you had to rephrase
>** the entire question?

Yes. It's not really that hard to rephrase the question with our modern
interpreters with their command histories and cursor-compatible
insertion-editing and yet I'd be miffed if I had to use it instead of just
typing 'gold'.

 =========================================================================

Slightly separate from Magnus' points then, would people benefit from a
'transferable knowledge' library (or am I ignorant of some similar
existing feature in some system; I'll admit I'm not that familiar with any
particular system, preferring theory to practice)? Has anyone worked on
such a system? 

--
Den



From dgatewood@bigfoot.com Mon Oct 27 21:00:38 MET 1997
Article: 30701 of rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!wuff.mayn.de!news.fh-hannover.de!newsfeed.nacamar.de!newsfeed.ecrc.net!newshub2.home.com!newshub1.home.com!news.home.com!Supernews60!supernews.com!Supernews69!not-for-mail
From: Dave Gatewood <dgatewood@bigfoot.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: Mon, 27 Oct 1997 14:51:19 -0500
Organization: All USENET -- http://www.Supernews.com
Lines: 52
Message-ID: <3454F0B7.54@bigfoot.com>
References: <62vede$psn$1@bartlet.df.lth.se> <Pine.SGI.3.95L.971027174633.14709A-100000@ebor.york.ac.uk>
Reply-To: dgatewood@bigfoot.com
NNTP-Posting-Host: 28687@207.42.201.106
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01 (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:30701

Den of Iniquity wrote:

> On a slightly different line to the one you're heading down, topics of
> conversation should be separate from the objects to which they refer.

However, it would be nice if the programmer could "associate" one or
more topics of conversation with an object, so that once the magic axe
had entered the player's scope, the topic "magic axe" would
automatically be deposited into the player's "bucket of concepts."  This
would save the programmer from doing a "concept transfer" for each and
every object in the game that the player encounters.  (Inevitably, when
done well, there would still be a lot of "concept movement" for the
programmer to do manually - but just the bare minimum outlined above
would be sufficient to mimic the current system.)

> Perhaps every NPC should have an invisible bucket of concepts which they
> can talk about.

This is exactly what I would like.  This would give a consistent system
for checking to see if a concept is "in scope" both for the PC and the
NPC.  It would then be a simple(*) matter of creating cleverly flexible
default responses for when the concept is _not_ in scope.  I posted
examples of this in a reply to Magnus's original post, but the basic
idea is this:

If the concept is not in the PC's "bucket", the _parser_ shouldn't allow
either "ask" or "tell"; whereas if the concept is not in the NPC's
"bucket", the _NPC_ shouldn't allow "ask" (by replying stupidly: "I
dinna know nothin' aboot that").  (Note that an NPC doesn't need to
"have" a concept to accept a "tell" action.)

(*) Of course, setting up a realistic "buckets" system in the first
place is not so simple; I'm just saying that _if_ such a system existed,
Magnus's questions about scope would then be easily solved.

> >| > Consult encyclopedia about London.
> 
>   "Under 'London' the index lists the following: [...] Which would you
>   like to look up?"

Once you suggest the existence of an index, you have to allow the player
to read the index, thereby giving the player a list of all of the
"consultable" topics in the encyclopedia.  Often this is precisely what
you want to avoid (take, for example, The History of the Meldrewes).  I
guess you might say that the flip side of the "Omniscient Troll problem"
is the "Conveniently Abridged Encyclopedia problem."

If we instead have the parser handle all questions of disambiguation
(based on the concepts in the PC's bucket), we avoid both of these
problems.

Dave


From dns361@merle.acns.nwu.edu Mon Oct 27 21:01:20 MET 1997
Article: 30698 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!4.1.16.34.MISMATCH!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!newsfeed.internetmci.com!128.174.5.49!vixen.cso.uiuc.edu!newsfeed.acns.nwu.edu!news.acns.nwu.edu!merle.acns.nwu.edu!dns361
From: Second April <dns361@merle.acns.nwu.edu>
Newsgroups: rec.arts.int-fiction,rec.games.int-fiction
Subject: Re: The End of the Age of Magick in the GUE
Date: Mon, 27 Oct 1997 12:38:26 -0600
Organization: Northwestern University, Evanston, IL, US
Lines: 63
Message-ID: <Pine.HPP.3.93.971027123018.11957A-100000@merle.acns.nwu.edu>
References: <62ueah$72q$1@soenews.ucsd.edu>
NNTP-Posting-Host: merle.acns.nwu.edu
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
In-Reply-To: <62ueah$72q$1@soenews.ucsd.edu>
Xref: news.lth.se rec.arts.int-fiction:30698 rec.games.int-fiction:27158

> In any case, something I wondered during Sorcerer and Beyond
> Zork was the signifance and possible metaphor of the End of the Age of
> Magick.  In Sorcerer, when you use magic you create a shadow and
> every time you cast a spell your shadow gets more powerful.

Do you mean Spellbreaker?

  It's
> a lot like a software company that comes up with a hot concept:
> a competitor always springs up to copy your idea.  In Beyond Zork, they
> talk about the End of the Age of Magick and the beginning of the Age of
> Science.  It's interesting to note that BZ was released around the
> same time that Infocom was transitioning from pure text
> adventures.  It is also well known that Infocommies 
> were proponents of games that were driven by story line (Age of Magick)
> rather than having technical superiority be their main selling point
> (Age of Science).

Hmmm...well, as Magnus pointed out, it's a general theme, and I'd say it
draws more specifically on Ursula LeGuin's Earthsea trilogy, which
features a similar sequence. What precisely LeGuin was driving at is open
to debate, but one could certainly argue that she has end-of-innocence
themes going on.

> Okay, I might be reaching here and this hypothesis is probably quite a
> bit far-feteched, but what I'm really wondering is this: Why would
> Dave Lebling destroy Magick in his storyline (Spellbreaker)?  That kind
> of event puts a damper on having a sequel to the Enchanter
> series.  It really affects the entire Zork series as anything else
> would have to either take place before or during Spellbreaker (e.g.
> BZ, Z0) or would have to be a story without Magick.  I realize that he
> was probably inspired by a similar plot device in a fantasy book,
> but I can't help but wonder what was on his (their) mind when
> he (they) made such a far-reaching change in the mechanics of the GUE.
> Was there an ulterior motive for ending the Age of Magick?

Well, remember that Beyond Zork provided an out, so to speak, in
safeguarding the Coconut for an age "beyond Magick, beyond Science...",
about which there isn't much to say besides conjecture. Whatever the
Coconut was actually supposed to do/represent, anyway, it pointed toward
some sort of continuation of the fantasy-Magick idea. For a future game?
Dunno, but the ending of Beyond Zork does point toward a sequel, and it's
worth speculating on what that sequel would've been like. (My guess,
admittedly influenced by the Enchanter trilogy: some evil type tries to
use the Coconut to recapture magic for his own ends.)

Interesting question, though. Someone who's played RTZ and ZN and ZGI
wanna say how/whether this is addressed, when the games are set in time,
etc.?

Duncan Stevens
d-stevens@nwu.edu
312-654-0280

The room is as you left it; your last touch--
A thoughtless pressure, knowing not itself
As saintly--hallows now each simple thing,
Hallows and glorifies, and glows between
The dust's gray fingers, like a shielded light.

--from "Interim," by Edna St. Vincent Millay




From dgatewood@bigfoot.com Tue Oct 28 11:14:22 MET 1997
Article: 30701 of rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!wuff.mayn.de!news.fh-hannover.de!newsfeed.nacamar.de!newsfeed.ecrc.net!newshub2.home.com!newshub1.home.com!news.home.com!Supernews60!supernews.com!Supernews69!not-for-mail
From: Dave Gatewood <dgatewood@bigfoot.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: Mon, 27 Oct 1997 14:51:19 -0500
Organization: All USENET -- http://www.Supernews.com
Lines: 52
Message-ID: <3454F0B7.54@bigfoot.com>
References: <62vede$psn$1@bartlet.df.lth.se> <Pine.SGI.3.95L.971027174633.14709A-100000@ebor.york.ac.uk>
Reply-To: dgatewood@bigfoot.com
NNTP-Posting-Host: 28687@207.42.201.106
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01 (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:30701

Den of Iniquity wrote:

> On a slightly different line to the one you're heading down, topics of
> conversation should be separate from the objects to which they refer.

However, it would be nice if the programmer could "associate" one or
more topics of conversation with an object, so that once the magic axe
had entered the player's scope, the topic "magic axe" would
automatically be deposited into the player's "bucket of concepts."  This
would save the programmer from doing a "concept transfer" for each and
every object in the game that the player encounters.  (Inevitably, when
done well, there would still be a lot of "concept movement" for the
programmer to do manually - but just the bare minimum outlined above
would be sufficient to mimic the current system.)

> Perhaps every NPC should have an invisible bucket of concepts which they
> can talk about.

This is exactly what I would like.  This would give a consistent system
for checking to see if a concept is "in scope" both for the PC and the
NPC.  It would then be a simple(*) matter of creating cleverly flexible
default responses for when the concept is _not_ in scope.  I posted
examples of this in a reply to Magnus's original post, but the basic
idea is this:

If the concept is not in the PC's "bucket", the _parser_ shouldn't allow
either "ask" or "tell"; whereas if the concept is not in the NPC's
"bucket", the _NPC_ shouldn't allow "ask" (by replying stupidly: "I
dinna know nothin' aboot that").  (Note that an NPC doesn't need to
"have" a concept to accept a "tell" action.)

(*) Of course, setting up a realistic "buckets" system in the first
place is not so simple; I'm just saying that _if_ such a system existed,
Magnus's questions about scope would then be easily solved.

> >| > Consult encyclopedia about London.
> 
>   "Under 'London' the index lists the following: [...] Which would you
>   like to look up?"

Once you suggest the existence of an index, you have to allow the player
to read the index, thereby giving the player a list of all of the
"consultable" topics in the encyclopedia.  Often this is precisely what
you want to avoid (take, for example, The History of the Meldrewes).  I
guess you might say that the flip side of the "Omniscient Troll problem"
is the "Conveniently Abridged Encyclopedia problem."

If we instead have the parser handle all questions of disambiguation
(based on the concepts in the PC's bucket), we avoid both of these
problems.

Dave


From lpsmith@rice.edu Tue Oct 28 11:14:36 MET 1997
Article: 30711 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!4.1.16.34.MISMATCH!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!howland.erols.net!newsxfer3.itd.umich.edu!gumby!newspump.wustl.edu!rice!ural.owlnet.rice.edu!lpsmith
From: lpsmith@rice.edu (Lucian Paul Smith)
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: 27 Oct 1997 22:31:36 GMT
Organization: Rice University, Houston, Texas
Lines: 69
Message-ID: <6334o8$c3m$1@joe.rice.edu>
References: <62vede$psn$1@bartlet.df.lth.se> <Pine.SGI.3.95L.971027174633.14709A-100000@ebor.york.ac.uk> <3454F0B7.54@bigfoot.com>
NNTP-Posting-Host: ural.owlnet.rice.edu
X-Newsreader: TIN [version 1.2 PL2]
Xref: news.lth.se rec.arts.int-fiction:30711

Dave Gatewood (dgatewood@bigfoot.com) wrote:
: Den of Iniquity wrote:

: > Perhaps every NPC should have an invisible bucket of concepts which they
: > can talk about.

: This is exactly what I would like.  This would give a consistent system
: for checking to see if a concept is "in scope" both for the PC and the
: NPC.  It would then be a simple(*) matter of creating cleverly flexible
: default responses for when the concept is _not_ in scope.  I posted
: examples of this in a reply to Magnus's original post, but the basic
: idea is this:

: If the concept is not in the PC's "bucket", the _parser_ shouldn't allow
: either "ask" or "tell"; whereas if the concept is not in the NPC's
: "bucket", the _NPC_ shouldn't allow "ask" (by replying stupidly: "I
: dinna know nothin' aboot that").  (Note that an NPC doesn't need to
: "have" a concept to accept a "tell" action.)

: (*) Of course, setting up a realistic "buckets" system in the first
: place is not so simple; I'm just saying that _if_ such a system existed,
: Magnus's questions about scope would then be easily solved.

Yes!  This is exactly what I would like to see.  And, barring that, this
is exactly what I would like to code up, and actually plan on doing, after
I recover from the competition.

My idea is to do this with flags.  Like the 'moved' and 'visited' flag,
you could have a 'concept' flag which would be updated along with those
other two.  (This would involve searching the library to find where this
occurs, and adding code for the new flag, too.)  You can't just use
'moved' and 'visited' as they currently stand, since then you couldn't
have static or scenery objects in this new concept scope.

At any rate, if you used this system with NPC's, you'd either have to have
a flag for each of them, or implement some other sort of system.  I liked
the idea someone posted earlier in the thread of having a scope routine
within the NPC.  The posted method wouldn't be ideal, but could be
slightly modified to something like:

Extend first "ask"
   * animate 'about' scope=AskScope   ->Ask;

[AskScope;
  if (noun provides SpecialAskScope)
  {
     noun.SpecialAskScope;
     rtrue;
  }
  else objectloop (x has concept) PlaceInScope(x);
  rtrue;
];

You could then use a clever combination of topic objects and the
ChooseObjects routine to deal with any other disambiguation problems.  For
example, in the hypothetical 'axe' example, if you wanted a general
response to 'ASK TROLL ABOUT AXES', you could create an object 'axes',
place it out of the way some where, but add it to concept scope, then tell
ChooseObjects to pick it over any of the other axes if the verb was Ask,
Tell, or Say, and to not pick it if the verb was something else.

This thread will have different applications depending on what you're
talking about--writing a game, or writing a library.  If you're writing a
game, all you need to account for is your particular constellation of
NPC's.  If you're writing a library, you need to take into consideration
all possible constellations, and so provide 'hooks' for the programmer to
override default behavior.

-Lucian


From erkyrath@netcom.com Tue Oct 28 11:14:47 MET 1997
Article: 30719 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!news.edu.sollentuna.se!news.oru.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Message-ID: <erkyrathEIqq7q.E0v@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <62vede$psn$1@bartlet.df.lth.se> <Pine.SGI.3.95L.971027174633.14709A-100000@ebor.york.ac.uk> <3454F0B7.54@bigfoot.com> <6334o8$c3m$1@joe.rice.edu>
Date: Tue, 28 Oct 1997 02:43:02 GMT
Lines: 61
Sender: erkyrath@netcom9.netcom.com
Xref: news.lth.se rec.arts.int-fiction:30719

Lucian Paul Smith (lpsmith@rice.edu) wrote:
> Dave Gatewood (dgatewood@bigfoot.com) wrote:
> : Den of Iniquity wrote:

> : > Perhaps every NPC should have an invisible bucket of concepts which they
> : > can talk about.

> : If the concept is not in the PC's "bucket", the _parser_ shouldn't allow
> : either "ask" or "tell"; whereas if the concept is not in the NPC's
> : "bucket", the _NPC_ shouldn't allow "ask" (by replying stupidly: "I
> : dinna know nothin' aboot that").  (Note that an NPC doesn't need to
> : "have" a concept to accept a "tell" action.)

> : (*) Of course, setting up a realistic "buckets" system in the first
> : place is not so simple; I'm just saying that _if_ such a system existed,
> : Magnus's questions about scope would then be easily solved.

> Yes!  This is exactly what I would like to see.  And, barring that, this
> is exactly what I would like to code up, and actually plan on doing, after
> I recover from the competition.

> My idea is to do this with flags. 

I think it would be safer and ultimately more flexible to have the 
concept system entirely separate. Don't try to represent a concept and a 
physical object with the same Inform object. Have a separate scope token 
for concepts that never touches the "standard" scope system at all.

This means that many significant objects will have parallel objects in 
the scope system. But this isn't really any extra code; it just divides 
things up a little more than you'd like. The new properties you're 
already defining wind up in a different object.

Naturally, every physical object can have a "concept_is" property, which 
points to a concept object. When the physical object is in scope, the 
library checks "obj.concept_is" and turns on its "seen" attribute. Two 
different axes (for example) might well point ot the same axe concept 
object. A dozen gold coins would definitely all point at the "money" 
concept. Contrariwise, a lot of objects would have no related concept at 
all, signifying that their concept is "nothing interesting."

The scope token would need one unique property: if the input words don't
match any object in concept-scope, it would match a "nothing interesting"
scope object, as opposed to saying "I don't know about any such thing." 
(Or maybe not.) In any case, I think this is as easy as always putting 
"nothing interesting" in scope and using ChooseObjects to give it a 
low weight.

This system makes a lot of the problems that have been discussed
evaporate. It doesn't solve the naive version of the problem, which is to
be able to ask the troll about every rock, door, and tree you've
encountered in the game. But I don't think anyone really wanted that in
the first place. :)

--Z


-- 

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


From dbs@coyote.cs.wisc.edu Tue Oct 28 11:15:19 MET 1997
Article: 30721 of rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news01.sunet.se!sunic!newscore.univie.ac.at!newsfeed.nacamar.de!news-feed.inet.tele.dk!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!ais.net!vixen.cso.uiuc.edu!uwm.edu!uwvax!coyote.cs.wisc.edu!dbs
From: dbs@coyote.cs.wisc.edu (Daniel Shiovitz)
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: 28 Oct 1997 04:46:57 GMT
Organization: U of Wisconsin CS Dept
Lines: 46
Message-ID: <633qo2$1so@spool.cs.wisc.edu>
References: <62vede$psn$1@bartlet.df.lth.se> <3454F0B7.54@bigfoot.com> <6334o8$c3m$1@joe.rice.edu> <erkyrathEIqq7q.E0v@netcom.com>
NNTP-Posting-Host: coyote.cs.wisc.edu
Xref: news.lth.se rec.arts.int-fiction:30721

In article <erkyrathEIqq7q.E0v@netcom.com>,
Andrew Plotkin <erkyrath@netcom.com> wrote:
[..]
>I think it would be safer and ultimately more flexible to have the 
>concept system entirely separate. Don't try to represent a concept and a 
>physical object with the same Inform object. Have a separate scope token 
>for concepts that never touches the "standard" scope system at all.
>
>This means that many significant objects will have parallel objects in 
>the scope system. But this isn't really any extra code; it just divides 
>things up a little more than you'd like. The new properties you're 
>already defining wind up in a different object.
>
>Naturally, every physical object can have a "concept_is" property, which 
>points to a concept object. When the physical object is in scope, the 
>library checks "obj.concept_is" and turns on its "seen" attribute. Two 
>different axes (for example) might well point ot the same axe concept 
>object. A dozen gold coins would definitely all point at the "money" 
>concept. Contrariwise, a lot of objects would have no related concept at 
>all, signifying that their concept is "nothing interesting."

I'm not sure how useful it would be, but it would be kinda cool if
classes could also have associated concept objects, perhaps for
plurals:

>ASK TROLL ABOUT FIRE CROWN
"Ooh, the mystic fire crown of frobozz! That hot!"

>ASK TROLL ABOUT ICE CROWN
"Ooh, the mystic ice crown of frobozz! That cold!"

>ASK TROLL ABOUT MYSTIC CROWNS
"Og heard of many mystic crowns. They neat-o."

I suppose you could also do this with a secondary plural_concept_is
property, and then you wouldn't have people asking about classes. Hmm.

[..]
>--Z
-- 
(Dan Shiovitz) (dbs@cs.wisc.edu) (look, I have a new e-mail address)
(http://www.cs.wisc.edu/~dbs) (and a new web page also)
(the content, of course, is the same)





From femaledeer@aol.com Tue Oct 28 11:15:22 MET 1997
Article: 30717 of rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news01.sunet.se!sunic!newscore.univie.ac.at!newsfeed.nacamar.de!news-peer.gsl.net!news.gsl.net!gip.net!portc01.blue.aol.com!audrey02.news.aol.com!not-for-mail
From: femaledeer@aol.com (FemaleDeer)
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: 28 Oct 1997 01:04:03 GMT
Lines: 26
Message-ID: <19971028010400.UAA28638@ladder02.news.aol.com>
NNTP-Posting-Host: ladder02.news.aol.com
X-Admin: news@aol.com
Organization: AOL http://www.aol.com
References: <3454F0B7.54@bigfoot.com>
Xref: news.lth.se rec.arts.int-fiction:30717

>From: Dave Gatewood <dgatewood@bigfoot.com>

>However, it would be nice if the programmer could >"associate" one or more
 topics of conversation with an >object, so that once the magic axe
>had entered the player's scope, the topic "magic axe" >would automatically be
 deposited into the player's "bucket >of concepts."  This would save the
 programmer from doing >a "concept transfer" for each and
>every object in the game that the player encounters.  >(Inevitably, when done
 well, there would still be a lot of >"concept movement" for the programmer to
 do manually - >but just the bare minimum outlined above
>would be sufficient to mimic the current system.)

Hmm, I may not know what I am talking about, but isn't this concept bucket idea
 already in the Christminister source code, specifically the topics of
 conversation combined with using "DecodeTopic"? I use a variation of that for
 NPC conversation. I can also see associating an attribute of some kind with
 the possible topics of conversation, such as  "moved" meaning a player has
 found it, or "known" or something, meaning the NPC knows about it. But like I
 said, I probably don't know what I am talking about, because I am not sure if
 that is what this thread is discussing.

FD :-)
------------------------------------------------------------------------------
FemaleDeer@aol.com       "Good breeding consists in 
concealing how much we think of ourselves and how 
little we think of the other person."             Mark Twain


From dmss100@york.ac.uk Tue Oct 28 14:07:18 MET 1997
Article: 30727 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!4.1.16.34.MISMATCH!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!dispose.news.demon.net!demon!uknet!yama.mcc.ac.uk!news.york.ac.uk!ebor.york.ac.uk!dmss100
From: Dennis Smith <dmss100@york.ac.uk>
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: Tue, 28 Oct 1997 12:06:21 +0000
Organization: The University of York, UK
Sender: dmss100@york.ac.uk
Message-ID: <Pine.SGI.3.95L.971028115710.28339A-100000@ebor.york.ac.uk>
References: <62vede$psn$1@bartlet.df.lth.se> <Pine.SGI.3.95L.971027174633.14709A-100000@ebor.york.ac.uk> <3454F0B7.54@bigfoot.com>
NNTP-Posting-Host: ebor.york.ac.uk
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
In-Reply-To: <3454F0B7.54@bigfoot.com>
Lines: 33
Xref: news.lth.se rec.arts.int-fiction:30727

On Mon, 27 Oct 1997, Dave Gatewood wrote:

> whereas if the concept is not in the NPC's "bucket", the _NPC_ shouldn't
>allow "ask" (by replying stupidly: "I dinna know nothin' aboot that"). 

The 'mad Scotsman' form of expressing ignorance? :)

>>   "Under 'London' the index lists the following: [...] Which would you
>>   like to look up?"
>
>Once you suggest the existence of an index, you have to allow the player
>to read the index, thereby giving the player a list of all of the
>"consultable" topics in the encyclopedia.

Ah, good point. Then: "The book has the following entries on London: [...]
Which would you like to look up?"

I wonder if you could exploit disambiguation to achieve the effect of
giving the player a 'fortuitous' clue. If when looking up 'London,
England', one is presented with the opportunity to learn about Jack
London, seemingly totally irrelevant, who wouldn't look it up anyway, out
of curiosity? Then the entry for Jack London could contain a clue to
another problem.

Does anyone besides me think that this would be cool? :)

Of course it wouldn't work in all cases, such as the reverse, in which you
might initially specify 'Jack' and thereby eliminate the need for
disambiguation.

--
Den



From graham@gnelson.demon.co.uk Tue Oct 28 14:35:50 MET 1997
Article: 30730 of rec.arts.int-fiction
From: Graham Nelson <graham@gnelson.demon.co.uk>
Newsgroups: rec.arts.int-fiction,rec.games.int-fiction
Subject: Re: The End of the Age of Magick in the GUE
Date: Mon, 27 Oct 1997 22:56:40 +0000 (GMT)
Organization: none
Message-ID: <ant2722400b0M+4%@gnelson.demon.co.uk>
References: <62ueah$72q$1@soenews.ucsd.edu> 
NNTP-Posting-Host: gnelson.demon.co.uk
X-NNTP-Posting-Host: gnelson.demon.co.uk [194.222.103.187]
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
X-Newsreader: ANT RISCOS Marcel [ver 1.26]
Lines: 21
Path: news.lth.se!solace!xinit!adm.icenet.no!news.IAEhv.nl!news.oru.edu!nntp.mid-ga.com!news1.mid-ga.com!news.pagesat.net!news.he.net!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news.idt.net!dispose.news.demon.net!demon!news.demon.co.uk!demon!gnelson.demon.co.uk!graham
Xref: news.lth.se rec.arts.int-fiction:30730 rec.games.int-fiction:27181

In article <62ueah$72q$1@soenews.ucsd.edu>, Dennis Lou
<URL:mailto:dlou@worf.ucsd.edu> wrote:
> 
> Okay, I might be reaching here and this hypothesis is probably quite a
> bit far-feteched, but what I'm really wondering is this: Why would
> Dave Lebling destroy Magick in his storyline (Spellbreaker)?  That kind
> of event puts a damper on having a sequel to the Enchanter
> series.  It really affects the entire Zork series...

Because it is a brilliant and appropriate ending to the trilogy.
Really, because it's artistically right, and hang continuity.  The
end of "Spellbreaker" works itself out like a Bach fugue, with only
one possible ending.  For a game which is basically a collection
of short scenes, "Spellbreaker" has a superb architecture to it,
and the ending makes the whole thing right.

Did I mention that I'm a fan?

-- 
Graham Nelson | graham@gnelson.demon.co.uk | Oxford, United Kingdom



From lpsmith@rice.edu Tue Oct 28 22:19:38 MET 1997
Article: 30741 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!4.1.16.34.MISMATCH!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!howland.erols.net!newsxfer3.itd.umich.edu!gumby!newspump.wustl.edu!rice!ural.owlnet.rice.edu!lpsmith
From: lpsmith@rice.edu (Lucian Paul Smith)
Newsgroups: rec.arts.int-fiction
Subject: Re: Encyclopedias and Omniscient Trolls: About scope, disambiguation and NPC Conversation
Date: 28 Oct 1997 18:20:14 GMT
Organization: Rice University, Houston, Texas
Lines: 40
Message-ID: <635acu$ivu$1@joe.rice.edu>
References: <62vede$psn$1@bartlet.df.lth.se> <Pine.SGI.3.95L.971027174633.14709A-100000@ebor.york.ac.uk> <3454F0B7.54@bigfoot.com> <Pine.SGI.3.95L.971028115710.28339A-100000@ebor.york.ac.uk>
NNTP-Posting-Host: ural.owlnet.rice.edu
X-Newsreader: TIN [version 1.2 PL2]
Xref: news.lth.se rec.arts.int-fiction:30741

Dennis Smith (dmss100@york.ac.uk) wrote:
: On Mon, 27 Oct 1997, Dave Gatewood wrote:

: I wonder if you could exploit disambiguation to achieve the effect of
: giving the player a 'fortuitous' clue. If when looking up 'London,
: England', one is presented with the opportunity to learn about Jack
: London, seemingly totally irrelevant, who wouldn't look it up anyway, out
: of curiosity? Then the entry for Jack London could contain a clue to
: another problem.

: Does anyone besides me think that this would be cool? :)

Zork Zero did this, actually, in a slightly different way.

***Spoilers below!*******


















For one puzzle, it was obvious that you had to look up 'jester' in the
Encyclopedia Frobozzica.  However, when you did, the picture of the jester
was ripped out, and you could see a partial text for something about
'jerrimore' on the next page.  If you looked up 'jerrimore', you got a
clue for a completely different puzzle in the game.

-Lucian


From dcornelson@placet.com Wed Oct 29 13:00:35 MET 1997
Article: 30759 of rec.arts.int-fiction
From: dcornelson@placet.com (David A. Cornelson)
Subject: Re: The End of the Age of Magick in the GUE
Date: Tue, 28 Oct 1997 22:35:35 -0600
Message-ID: <878099235.14211@dejanews.com>
Newsgroups: rec.arts.int-fiction
Organization: Deja News Posting Service
Path: news.lth.se!solace!xinit!adm.icenet.no!news.IAEhv.nl!news.cs.utwente.nl!newsfeed.eris.dera.gov.uk!delos.dra.hmg.gb!dispose.news.demon.net!demon!news.idt.net!news-peer.gsl.net!news.gsl.net!gip.net!news-peer.sprintlink.net!news-pull.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!204.238.120.130!jump.net!grunt.dejanews.com!not-for-mail
References: <62ueah$72q$1@soenews.ucsd.edu> <ant2722400b0M+4%@gnelson.demon.co.uk>
X-Article-Creation-Date: Wed Oct 29 04:27:17 1997 GMT
X-Authenticated-Sender: dcornelson@placet.com (David A. Cornelson)
X-Http-User-Agent: Mozilla/4.0 (compatible; MSIE 4.0; Windows 95)
X-Originating-IP-Addr: 199.174.190.245 (ad61-245.compuserve.com)
Lines: 52
Xref: news.lth.se rec.arts.int-fiction:30759

In article <62ueah$72q$1@soenews.ucsd.edu>, Dennis Lou
<URL:mailto:dlou@worf.ucsd.edu> wrote:
> >
> > Okay, I might be reaching here and this hypothesis is probably quite a
> > bit far-feteched, but what I'm really wondering is this: Why would
> > Dave Lebling destroy Magick in his storyline (Spellbreaker)?  That kind
> > of event puts a damper on having a sequel to the Enchanter
> > series.  It really affects the entire Zork series...
>
> Because it is a brilliant and appropriate ending to the trilogy.
> Really, because it's artistically right, and hang continuity.  The
> end of "Spellbreaker" works itself out like a Bach fugue, with only
> one possible ending.  For a game which is basically a collection
> of short scenes, "Spellbreaker" has a superb architecture to it,
> and the ending makes the whole thing right.
>
> Did I mention that I'm a fan?
>

Also remember that Infocom's demise was a very quick thing. They tried to
build the database program (I forget the name of it) that would help them
diversify, but that too failed. When Spellbreaker was written, they
probably had every intention of doing a science trilogy (or two) which
would then lead back to the rediscovery of magick or the balance being
restored.

It's actually one of the more disappointing aspects of the Infocom
trademarks. One of us (Graham) would likely be an excellent author of
such a continuation.

Imagine the anticipation...first comes "Inventor", where a guild
apprentice is learning about the wheel and the inclined plane and he
discovers some evil in the GUE that confronts with all of his newfound
knowledge.

A year passes and the second episode, eagerly awaited, appears....and
it's called "Builder", and then the finally, "Engineer", at the end of
which we find a great machine that runs everything in the GUE and at the
center of the machine is the Coconut of Quendor!

I say we petition Activision, plead for our rights as true Infocom
adventurers. They call us "Zorkers", a name I detest. They simply do not
understand. Text cannot be compared to pictures (egadz!), prose produces
the greatest animations and graphics in the world...in our minds!

Well.....umm.......yes.......I guess I'm a fan too.

David A. Cornelson, Chicago (bang bang Al Capone)
dcornelson@placet.com

-------------------==== Posted via Deja News ====-----------------------
      http://www.dejanews.com/     Search, Read, Post to Usenet


From jholder@io.frii.com Wed Oct 29 17:34:35 MET 1997
Article: 30765 of rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news00.sunet.se!sunic!mn6.swip.net!nntp.uio.no!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!news-peer.sprintlink.net!news.sprintlink.net!Sprint!newsfeed.internetmci.com!208.206.176.15!dimensional.com!frii.com!io.frii.com!jholder
From: J. Holder <jholder@io.frii.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: The End of the Age of Magick in the GUE
Date: 29 Oct 1997 16:26:40 GMT
Organization: Front Range Internet, Inc.
Lines: 43
Message-ID: <637o41$e5k$1@europa.frii.com>
References: <62ueah$72q$1@soenews.ucsd.edu> <ant2722400b0M+4%@gnelson.demon.co.uk> <878099235.14211@dejanews.com>
NNTP-Posting-Host: io.frii.com
Xref: news.lth.se rec.arts.int-fiction:30765

Thus spake David A. Cornelson <dcornelson@placet.com>:
: I say we petition Activision, plead for our rights as true Infocom
: adventurers. They call us "Zorkers", a name I detest. They simply do not
: understand. Text cannot be compared to pictures (egadz!), prose produces
: the greatest animations and graphics in the world...in our minds!

No, it was Infocom who invented the term "Zorkers".
(the original Infocom.)

To quote Zork I:

"The prayer is inscribed in an ancient script, rarely used today. It
seems to be a philippic against small insects, absent-mindedness, and the
picking up and dropping of small objects. The final verse consigns trespassers 
to the land of the dead. All evidence indicates that the beliefs of the ancient
Zorkers were obscure."

and

"The engravings were incised in the living rock of the cave wall by an
unknown hand. They depict, in symbolic form, the beliefs of the ancient
Zorkers. Skillfullly interwoven.... [etc]"

To quote Zork II:

"The room contains the earthly remains of the mighty Flatheads, twelve
somewhat flat heads mounted securly on poles.  While the room might be expected
to contain funerary urns or other evidence of the ritual practices of the
ancient Zorkers, it is empty of all such objects.  There is...[etc]"


As should be obvious from the context, Zorkers are a people
who inhabit the Zork universe, such as DimWit Flathead.  We, on the
other hand, are really adventurers (in the Zorks, at least), as can be 
seen in the turnabout scene in Enchanter, where we are actually a 
"modern Zorker", and we manipulate the Adventurer.

So, you may detest the term, but in at least three of the Infocom
adventures, we are cast as Zorkers. (the Enchanter trilogy)
-- 
John Holder (jholder@frii.com)         http://www.frii.com/~jholder/
Sr. Programmer Analyst, J.D.Edwards World Source Company, Denver, CO
http://www.jdedwards.com/


From jcmason@undergrad.math.uwaterloo.ca Wed Oct 29 19:37:01 MET 1997
Article: 30770 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news00.sunet.se!sunic!mn6.swip.net!nntp.uio.no!uninett.no!news.algonet.se!4.1.16.34.MISMATCH!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!howland.erols.net!torn!kwon!watserv3.uwaterloo.ca!undergrad.math.uwaterloo.ca!jcmason
From: jcmason@undergrad.math.uwaterloo.ca (Joe Mason)
Subject: Re: Infocom Sighting!!!!!
Sender: news@undergrad.math.uwaterloo.ca (news spool owner)
Message-ID: <EItowC.2v3@undergrad.math.uwaterloo.ca>
Date: Wed, 29 Oct 1997 17:07:24 GMT
References: <345561c4.2299083@nntp.ix.netcom.com> <Pine.GSO.3.96.971028084350.20131A-100000@ucsu.Colorado.EDU> <34561b4b.1093671@news.alt.net> <635g7r$hr6@solaria.cc.gatech.edu>
Nntp-Posting-Host: picard.math.uwaterloo.ca
Organization: University of Waterloo
Lines: 29
Xref: news.lth.se rec.arts.int-fiction:30770

In article <635g7r$hr6@solaria.cc.gatech.edu>,
Peter Ilberg <ilberg@cc.gatech.edu> wrote:
>
>: I believe it was intentional.  Who but Douglas Adams would name a
>: robot Floyd?
>
>Steve Meretzky?

Hey, that reminds me!  I've had this magazine for years, never even made the
connection until now:

>From the MAD Star Wars pardody, _The Force and I_ (The MAD Star Wars Musical),
during the Jawa swap meet scene:

(sung to the tune of _My Favourite Things_):

Slightly used robots with brains electronic
Self operating with bodies bionic
Full of ambition and now unemployed:
That's what we've got in a second-hand droid

...

Here's a hum-dinger from Alpha Centaurus
Give him a kick and he'll sing the next chorus
Name's RK-4 but he answers to Floyd
That's what we've got in a second-hand droid!

Joe


From erkyrath@netcom.com Thu Oct 30 21:56:42 MET 1997
Article: 30809 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!mn6.swip.net!nntp.uio.no!www.nntp.primenet.com!globalcenter0!news.primenet.com!nntp.primenet.com!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: [Inform] Making a trenchcoat
Message-ID: <erkyrathEIvrwD.M5H@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <34585431.5673@ucla.edu> <63a85o$9iv$1@joe.rice.edu>
Date: Thu, 30 Oct 1997 20:07:24 GMT
Lines: 46
Sender: erkyrath@netcom5.netcom.com
Xref: news.lth.se rec.arts.int-fiction:30809

Lucian Paul Smith (lpsmith@rice.edu) wrote:
> Jon Petersen (enoto@ucla.edu) wrote:

> : I'm trying to implement a trenchcoat.  Being a coat, it has left
> : and right pockets (which are just containers attached to the coat 
> : object).  Being a trenchcoat, you can put your shotgun, or laptop, or 
> : whatever, inside the coat itself as well, and button up (close) the coat 
> : to hide whatever's inside it.

> If you absolutely must have the trench coat be a container and also have
> pockets, use add_to_scope (section 28).

Yet another option: have *three* containers, the left pocket, the right 
pocket, and the inside-of-coat. The coat itself contains those three 
objects, but the coat does not have "container"; it has "transparent" 
instead. 

Then you have to route a few actions from the coat to the inside-of-coat. 
This isn't too bad:

Object coat
with
before [;
  Open:
    <<Open insidecoat>>;
  Close:
    <<Close insidecoat>>;
  Receive:
    <<Insert noun insidecoat>>;
  Search:
    <<Search insidecoat>>;
],
has transparent;

And you might want to add some detail to the description, as well. 

Getting the inventory listing right is probably a pain, but it's going to
be a pain no matter what you do. I'd change the inventory routine to list
them manually (calling WriteListFrom.)

--Z

-- 

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


From ravipind@fast.net Tue Nov  4 13:10:05 MET 1997
Article: 30886 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!feed1.news.erols.com!howland.erols.net!newsxfer.itd.umich.edu!sunqbc.risq.qc.ca!nntp.flash.net!newsgate.ptd.net!fastnet!news.fast.net!not-for-mail
From: ravipind@fast.net
Newsgroups: rec.arts.int-fiction,rec.games.int-fiction
Subject: Re: The End of the Age of Magick in the GUE
Date: 3 Nov 1997 23:45:31 GMT
Organization: The Frobozz Magic Usenet Company
Lines: 40
Distribution: world
Message-ID: <63lnmr$k6j$1@news1.fast.net>
NNTP-Posting-Host: x2-01-027.phl.fast.net
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Newsreader: <QVT/News ver. 4.1>
Xref: news.lth.se rec.arts.int-fiction:30886 rec.games.int-fiction:27329

Interesting question, though. Someone who's played RTZ and ZN and ZGI
>wanna say how/whether this is addressed, when the games are set in time,
>etc.?

Well, ZGI comes with a timeline.  I don't know how "official" you'd consider 
that though...

The 'eras' as defined by this chronology are:

Pre-Time BE - 0 GUE 		Before Entharion Civilization
0 GUE - 659 GUE		Entharion Dynasty
659 GUE - 883 GUE		Flathead Dynasty
883 GUE - 1047 GUE		The Fallen Empire
1047 GUE - "Present"		The Inquisition

The games seem to fall as:

Zork 0 		883 GUE (the fall of the Flathead Dynasty)
Zork I - III 		948 GUE 
Zork: Nemesis		949 GUE
Enchanter		952 GUE
Sorcerer		957 GUE
Spellbreaker		966 GUE 
Beyond Zork		966 GUE
Zork: GI		"Present" (ie c. 1070 GUE)

I'm not sure where RTZ would fall, but it's beyond the 'present' in Z:GI, 
right?  I vaguely remember it being a time-period in the time-travel gallery 
in Z:N.

Did I forget anything?

I have to say that if magic DOES come back at the end of Z:GI (I haven't 
gotten that far) I'll be pretty disappointed.  I don't think 100 years really 
counts as an 'Age of Science.'

d
who's actually trying to come up with a timeline for the novel "Infinite Jest" 
and keeping a character list as he's re-reading it.  Anyone who's read it 
knows what I'm talking about when I say it's an insane project, but fun. :)


From mkkuhner@phylo.genetics.washington.edu Wed Nov  5 13:55:11 MET 1997
Article: 30922 of rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news00.sunet.se!sunic!mn6.swip.net!nntp.uio.no!news-feed.ifi.uio.no!uninett.no!news.maxwell.syr.edu!news-peer.sprintlink.net!news.sprintlink.net!Sprint!worldnet.att.net!news.u.washington.edu!phylo.genetics.washington.edu!mkkuhner
From: mkkuhner@phylo.genetics.washington.edu (Mary K. Kuhner)
Newsgroups: rec.arts.int-fiction
Subject: Twelve Days of Comp97
Date: 5 Nov 1997 05:00:45 GMT
Organization: University of Washington, Seattle
Lines: 22
Distribution: world
Message-ID: <63ouht$lt7$1@nntp5.u.washington.edu>
NNTP-Posting-Host: phylo.genetics.washington.edu
X-Trace: nntp5.u.washington.edu 878706045 22439 (None) 140.142.64.4
X-Complaints-To: help@cac.washington.edu
Xref: news.lth.se rec.arts.int-fiction:30922

I hope this will not be considered "discussion of the games"....

**

On the twelfth day of the contest Comp97 brought to me:

Twelve cosy houses,
Eleven obstructed doors,
Ten boring bathrooms,
Nine furry animals,
Eight wicked villains,
Seven home computers,
Six self-references,
Five tasty feasts!
Four romantic bits,
Three chairs to climb, 
Two marine biologists,
And a scrapple factory!

(Inform and JACL games only, one instance counted per game.)

Mary Kuhner mkkuhner@genetics.washington.edu


From blore@ibm.net Thu Nov  6 09:05:47 MET 1997
Article: 30943 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
From: blore@ibm.net (Steven Howard)
Date: Wed, 05 Nov 97 19:15:41 -0500
Subject: [Comp 97] Side Bets
X-Newsreader: MR/2 Internet Cruiser Edition for OS/2 v1.40a b42 
NNTP-Posting-Host: 129.37.55.125
Message-ID: <34613981.0@news2.ibm.net>
Organization: IBM.NET
Lines: 25
Path: news.lth.se!feed1.news.luth.se!luth.se!news-ge.switch.ch!newscore.univie.ac.at!europa.clark.net!208.134.241.18!newsfeed.internetmci.com!165.87.194.248!newsm.ibm.net!ibm.net!news2.ibm.net!129.37.55.125
Xref: news.lth.se rec.arts.int-fiction:30943

For no reason at all, I'm now accepting *BY E-MAIL ONLY* side bets on the
1997 Competition.  They're not really "bets", since no money is changing
hands.  All you can "win" is the right to say "I told you so" for a year.

Anyone can play: authors, beta-testers, fictional characters, I don't 
care.  All you have to do is E-mail me with your predictions on the
following topics.  After the results of the actual competition are 
posted, I'll announce the prediction results.  

1.  What will be the top three games (in order)?
2.  What will be the bottom three games (in order)?
3.  Which game that you hated will place in the top five?
4.  Which game that you loved will place in the bottom five?
5.  (Extra credit) Identify as many anonymous authors as you can.
    (Note: anonymous authors probably shouldn't get the extra
     credit for guessing their own identities, but they will.)

Remember: Do not post follow-ups to this post.  E-mail me.

========
Steven Howard
blore@ibm.net

"Are you a COBOL programmer?" "No, but people tell me I look like one."



From mkkuhner@phylo.genetics.washington.edu Fri Nov  7 19:15:33 MET 1997
Article: 30963 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news.algonet.se!newsfeed.direct.ca!news-peer.sprintlink.net!news-pull.sprintlink.net!news-in-east.sprintlink.net!news.sprintlink.net!Sprint!148.59.19.27!nntp.msen.com!ftl.msen.com!conch.msen.com!not-for-mail
From: mkkuhner@phylo.genetics.washington.edu (Mary K. Kuhner)
Newsgroups: news.groups.reviews,rec.arts.int-fiction,rec.games.int-fiction
Subject: REVIEW: rec.arts.int-fiction & rec.games.int-fiction
Supersedes: <63le2b$v6s$1@nntp6.u.washington.edu>
Followup-To: rec.arts.int-fiction,rec.games.int-fiction
Date: 6 Nov 1997 23:51:05 -0500
Organization: University of Washington, Seattle
Lines: 105
Approved: news-groups-reviews@uunet.uu.net (Michael Bauser)
Message-ID: <63u6np$7c0@conch.msen.com>
NNTP-Posting-Host: conch.msen.com
Mime-Version: 1.0
PICS-Label: (PICS-1.1 "http://www.weburbia.com/safe/" l r (s 0))
Xref: news.lth.se news.groups.reviews:201 rec.arts.int-fiction:30963 rec.games.int-fiction:27439

Newsgroup: rec.arts.int-fiction
Moderator: none
Description: Discussions about interactive fiction.
FAQ-Location: ftp://ftp.gmd.de/if-archive/rec.arts.int-fiction/FAQ

Newsgroup: rec.games.int-fiction
Moderator: none
Description: All aspects of interactive fiction games.
FAQ-Location: ftp://ftp.gmd.de/if-archive/rec.games.int-fiction/FAQ

Reviewer: Mary K. Kuhner <mkkuhner@phylo.genetics.washington.edu>
Review-Date: 3 Nov 1997 21:00:59 GMT

This pair of newsgroups covers interactive fiction (IF), which in practice
mainly means text adventures:  computer games in which the player types
in words and phrases to advance the action, such as _Zork_.  It also
encompasses hypertext fiction and similar media, but these are less
often discussed.

An example of the genre: 

---

You are in a small, bare prison cell.  A bare bulb dangles from the
ceiling, but it is dead:  the only light comes in through a barred
window.  A low shelf serves as a bed.

>take bulb

You can't reach it from here.

>stand on bed

You climb up on the narrow shelf.  From here you can see a glimpse
of the prison yard through the window.

>take bulb

You carefully unscrew the dead bulb, leaving an empty socket.

>inventory

You have a gray prison uniform and a dead lightbulb.

[...]

---

The two newsgroups have specialized functions, and crossposting between
them is usually not appropriate.

[Moderator's Note: I'm crossposting this review to both groups.  Please
don't hurt me.]

Rec.arts.int-fiction is for discussion of creating and appreciating 
interactive fiction.  Topics include the nuances of game
programming, plotting, prose-writing for interactive fiction,
characterization and so forth.  Game reviews can be posted here,
or sent to any of the several excellent on-line review newsletters.
Please be careful to avoid or clearly label spoilers.

There is lively discussion of both commercially available games and
fan-written ones:  the newsgroup supports a substantial game-
writing community and many of the games produced are of very high
quality, comparable (in this reviewer's opinion) to Infocom at its
best.  Text adventure programming requires skill and creativity,
but the basics are relatively easy to learn:  it's a very accessible
art form for the amateur.

Rec.games.int-fiction is for topics directly related to playing
interactive fiction games:  hints, availability of games, help in
getting them to run on particular machines, and so forth.  Again, please
be careful to clearly signpost spoilers, even for old games:
there's always someone out there playing _Zork_ for the very first
time.

General consensus is that bugs in games, unless they are amusing
or illustrative, should be emailed to the author rather than posted
to either newsgroup.  Fixes ("You can get game X to run on an Amiga
if you do Y") are entirely welcome, particularly on r.g.int-fiction.

These two newsgroups help support the yearly Interactive Fiction Writing
Contest:  participants write short (two-hour) IF games and 
submit them in October to be judged by all comers.  The 1997
competition is currently underway, with 35 games submitted.  Judging
will continue until the end of December 1997, so there's still time
to become a judge.  The competition games, archives for the newsgroups, 
and huge amounts of freely available IF games and materials (including 
several programming languages for game writing) are available at the 
FTP site ftp.gmd.de in directory /pub/if-archives.   This site is
the world's most amazing archive for all matters related to IF, and
well worth a visit if you are interested in the genre.

RAIF and RGIF are not for discussion of fiction in general, only
interactive fiction:  the readers will become annoyed if you post
advertisements for your novel or magazine, unless it is interactive
(and that should mean more than "available on a Web page").

Mary Kuhner mkkuhner@genetics.washington.edu

-- 
For more information about news.groups.reviews,
see <URL: http://web.superb.net/islander/ngr >.
This review will be permanently archived at <URL:
http://web.superb.net/islander/ngr/Reviews/Rec/rec.arts.int-fiction.1.html >


From kjfair@midway.uchicago.edu.REMOVEME Tue Nov 11 10:34:04 MET 1997
Article: 31032 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!news.edu.sollentuna.se!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!howland.erols.net!vixen.cso.uiuc.edu!uchinews!kjfair
From: kjfair@midway.uchicago.edu.REMOVEME (Kenneth Fair)
Subject: Re: IF combat example
X-Nntp-Posting-Host: ntcs-ip85.uchicago.edu
Message-ID: <kjfair-1111970143260001@ntcs-ip85.uchicago.edu>
Sender: news@midway.uchicago.edu (News Administrator)
Organization: University of Chicago School of Law
X-Newsreader: MT-NewsWatcher 2.3.5
References: <346794DF.1B5C@stc.net>
Date: Tue, 11 Nov 1997 07:43:26 GMT
Lines: 61
Xref: news.lth.se rec.arts.int-fiction:31032

In article <346794DF.1B5C@stc.net>, bard@stc.net wrote:

>Here's how, in my opinion, one could use battles in an IF game.
>
>A goblin leaps from the bushes from the north.
>HP=10 >attack n
>The goblin dodges your clumsy blow.
>The goblin knocks you hard on the head for 3 damage.
>HP=7  >cast fire n
>You utter a few words and cast a fire spell.
>The spell roasts the goblin for 10 damage, killing it.
>Gain whatever exp points (if any) and get whatever treasure (if any).

The nice thing, though, is that you don't have to tell the player the
numerical information.  I think it makes the game more realistic to have
an exchange like this:

A goblin leaps from the bushes to the north!

>ATTACK GOBLIN
You swing at the goblin, but miss.

The goblin whacks you over the head with his staff.  The room spins, and
you shake your head to get your bearings.

You feel a bit woozy.

>AGAIN
You execute a nice pirouette inside the goblin's guard and clobber him in
the stomach with your staff.  He says "Whoof!" and staggers backwards.
He appears to be in some pain.

The goblin growls low, feints to the left, and twists, bringing the staff
in low to whack you in the right leg.

You feel woozy, and you're limping.

>POINT WAND AT GOBLIN
You point the wand at the goblin.

The goblin swings at you, but misses.

You feel woozy, and you're limping.

>CAST FLAMBE
You utter the word "Flambe!" and a jet of fire shoots from the end of
the wand, enveloping the goblin.  He screams in agony and keels over,
quite dead.

>SEARCH CORPSE
You find a golden ring and some coins in the goblin's charred leather
pouch.  There's also a piece of paper, but it's too burnt to read.


Now isn't that a more interesting description to read?

-- 
KEN FAIR - U. Chicago Law  | <http://student-www.uchicago.edu/users/kjfair>
Of Counsel, U. of Ediacara | Power Mac! | CABAL(tm) | I'm w/in McQ - R U?
      "Any smoothly functioning technology will be 
       indistinguishable from a rigged demo."  Isaac Asimov


From erkyrath@netcom.com Wed Nov 12 12:36:51 MET 1997
Article: 31045 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!solace!xinit!adm.icenet.no!news.edu.sollentuna.se!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: IF combat example
Message-ID: <erkyrathEJIMv3.sF@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <346794DF.1B5C@stc.net> <kjfair-1111970143260001@ntcs-ip85.uchicago.edu> <649tab$1g4$1@news1.atlantic.net> <64aqjs$ed2$1@news.one.net> <OclO04x78GA.182@upnetnews03>
Date: Wed, 12 Nov 1997 04:23:27 GMT
Lines: 42
Sender: erkyrath@netcom5.netcom.com
Xref: news.lth.se rec.arts.int-fiction:31045

Brock Kevin Nambo (newsmaster@earthling.net) wrote:

>  This is a very odd thing that I know at least
> some of the Zork games had and possibly many others.. "It appears to be half
> full." "It seems as if something is blocking it."  Yada Yada.  It may be
> that the stereotypical adventure player doesn't believe everything he/she
> sees, but does the game have to tell us that?  Do modern IF authors have the
> urge to write such uncertainties?

Sure, I do it all the time.

Perception is not exact. If I put in my game "It appears to be half 
full," I'm leaving open the possibility that you may get a volumetric 
flask later in the game and find that it's 49.78% full. Or that the upper 
half is full of an invisible liquid.

If you reach an opaque door, try to open it, and get "It feels as if 
something is blocking it," then that's how it feels. You can't see what 
it is (otherwise I'd say "There's a fallen orc blocking the door.") Maybe 
it's a false door and *can't* open. All sorts of possibilities.

I don't expect you to distrust all statements of that sort. I mean, I 
won't make a puzzle where you must -- with no hints at all -- realize 
that your perceptions are misleading you. But maybe there *will* be hints 
later. Or maybe you'll just discover later that you're wrong, no puzzle 
involved.

I don't want to say "The door is blocked" when the door really isn't 
blocked. That would be sort of unfair.

In the combat example, it's even more relevant, because it's hard for the 
protagonist to estimate damage. I don't want to print "The goblin is 
nearly dead." and then "The goblin was faking! It jumps up and eats your 
head." Better to emphasize the uncertainty by saying "The goblin looks 
nearly dead."

--Z

-- 

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


From sothoth@xxx.usa.net Wed Nov 12 18:39:24 MET 1997
Article: 31047 of rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!news-ge.switch.ch!newscore.univie.ac.at!03-newsfeed.univie.ac.at!alijku02!not-for-mail
From: "Gunther Schmidl" <sothoth@xxx.usa.net>
Newsgroups: rec.arts.int-fiction
Subject: Re: On Combat Systems (was: Re: Authoring Tool: Combat?)
Date: Tue, 11 Nov 1997 14:32:54 +0100
Organization: Johannes Kepler University Linz
Lines: 127
Message-ID: <649muq$1ble$1@alijku04.edvz.uni-linz.ac.at>
References: <6459vi$7l9$1@news1.tor.acc.ca>
NNTP-Posting-Host: async04.edvz.uni-linz.ac.at
X-Newsreader: Microsoft Outlook Express 4.71.1712.3
X-MimeOLE: Produced By Microsoft MimeOLE V4.71.1712.3
Xref: news.lth.se rec.arts.int-fiction:31047

0. NOTE
-------
This is a rather lengthy discourse about combat in Interactive Fiction
systems. It contains possible spoilers for some I-F games (Zork 1, Beyond
Zork, Journey, Hollywood Hijinx, Border Zone, Arthur, Sylenius Mysterium),
so be careful.

1. COMBAT IN EXISTING GAMES
---------------------------
As far as I know, there are five Infocom games that use actual hand-to-hand
(or sword-to-sword, or chainsaw-to-bat, or whatever) combat: Zork 1, Beyond
Zork, Hollywood Hijinx, Arthur and Journey.

1.1 ZORK 1
----------
The system used in Zork 1 contains mainly of one command: ATTACK x WITH y.
Depending on the experience and status of the player, a hit is more or less
likely to succeed. For example, the thief is very hard to dispatch unless
you have a lot of experience and are rested enough from previous fights (i.
e. enough turns have passed since you've last been hit), although it's
possible to carry around gold sarcophagi and tons of other stuff.
The designers themselves were not very happy with this system, and so it was
never again used in this form.

1.2 BEYOND ZORK
---------------
One of the "style experiments" of Infocom, Beyond Zork has a rather
"RPG-ish" feel about it, with the player having different attributes:
Endurance, Strength, Dexterity, Compassion, Intelligence, Luck. These are
used in combat to determine the outcome of each hit, done again via the
simple "ATTACK MONSTER" command, conveniently placed on the F7 key this
time.
However, there is at least one different possibility of getting rid of each
enemy besides fighting. Some involve diverse Rods, others take advantage of
some monsters' special disabilities (like the sensitive eyes of the
Dornbeast; it can be irritated by cutting an onion), and some (like the dust
bunnies or Christmas Tree monsters) can't be fought at all, and must be
tricked.

1.3 JOURNEY
-----------
Someone mentioned this game, but as far as I remember, there is but one
avoidable fight in this game, and it can be "walked around" if you have a
scout. The combat system lets you win a fight in one way only, and it
usually involves clever thinking and usage of spells than brute force if one
wants to avoid injury or death of party members.

1.4 HOLLYWOOD HIJINX
--------------------
I name this game only because the end-game has a hilarious combat scene, in
which one must determine which of the movie props that lie around is not
really a prop, but a real weapon to get rid of your foe, but it again
consists only of "ATTACK x WITH y", and is really just for fun.

1.5 ARTHUR
----------
Arthur lets the player fight out a tournament in one place, but it is very
clear what you have to do to win, and it involves no commands, but simply
gives you choices a, b, c, of which you must select the correct ones.

2. REAL-TIME COMBAT
-------------------

2.1 BORDER ZONE
---------------
Interestingly enough, Border Zone, Infocom's only real-time game, has lots
of time-dependent puzzles, but none of them involves fighting.

2.2 WHAT I'M DOING
------------------
In my current game-in-progress (If I ever get it done :), I plan to include
a real-time combat system. To get an impression of this, play the arcade
section of the competition game "Sylenius Mysterium" for the principle of
the real-time mechanism.
The combat system would respond to single key-presses, e. g. L for attack
left, R for attack right, B for block or whatever, and react to the player's
input as far as possible, but not making a combat impossible to win and not
too linear (several strategies can win you a battle). This, however, is only
a plan, and I'm not too sure if I ever get so far. It would look something
like this (the lines coming in real-time)

XYZ swings his sword at you!
The sword is now dangerously close to your throat
>B        [block]
You manage to parry the blow just in time. Your opponent stumbles back!
>S        [stab]
You stab the sword at him, and manage to pierce his side! As you pull out
the sword, blood starts rinning from the wound.
Seemingly unaffected, XYZ quickly throws a dagger at you!
The dagger comes closer!
The dagger is now dangerously close!
>D        [duck]
Too late! The dagger hits you.
[...]

3. COMMENTS
-----------
My personal opinion is that real-time combat would be the most interesting
form of combat in I-F, taking out the static "HIT x WITH y" part. However,
it may prove difficult for slowly-reacting players (but then, one could
always implement difficulty settings) and could be hard to program (though I
don't think so).
Another idea would be to just let a player win a combat when he's properly
equipped to do so, and warn him otherwise, though it might be hard to find
reasons for that, and rather reminds me of AGT (shudder).
The most uninteresting form is still the one seen in Zork 1 and Beyond Zork,
and random elements in games have always proven to be more of a nuisance.
If, however, the outcome depends on the player's dexterity (and I mean the
*player's* dexterity, not the dexterity of the on-screen character like in
Beyond Zork), it might just prove to be very interesting and challenging.
HOWEVER, good measure should be used; to take Sylenius Mysterium again, I
couldn't even get through the first level, and four (or five?) levels may
prove to be too much for even the most patient of players. Same for
extra-long, extra-hard combats.

--

+------------------------+----------------------------------------------+
+ Gunther Schmidl        + "I couldn't help it. I can resist everything +
+ Ferd.-Markl-Str. 39/16 +  except temptation"           -- Oscar Wilde +
+ A-4040 LINZ            +----------------------------------------------+
+ Tel: 0732 25 28 57     + http://gschmidl.home.ml.org - new & improved +
+------------------------+---+------------------------------------------+
+ sothoth (at) usa (dot) net + please remove the "xxx." before replying +
+----------------------------+------------------------------------------+




From erkyrath@netcom.com Thu Nov 13 09:30:15 MET 1997
Article: 31052 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!feed1.news.erols.com!howland.erols.net!news-peer.sprintlink.net!news.sprintlink.net!Sprint!worldnet.att.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Ethical game-design question...
Message-ID: <erkyrathEJJunA.G4C@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <19971112195501.OAA07324@ladder02.news.aol.com>
Date: Wed, 12 Nov 1997 20:09:10 GMT
Lines: 36
Sender: erkyrath@netcom12.netcom.com
Xref: news.lth.se rec.arts.int-fiction:31052

Zey Guy (zeyguy@aol.com) wrote:
> O.K., here's the sitch.

> I have been working on completing my game, Bedlam. There was one puzzle that I
>  spent a lot of time coding and perfecting. Then much to my dismay, I play Zork
>  Grand Inquisitor to find that they used the exact same puzzle themselves. I
>  would hate to chuck all that time and effort down the drain, but nor do I want
>  to be accused of copying.

I don't think you will be; you can honestly claim you came up with the 
idea independently, and there's no reason to push you. (Unless Bedlam 
will be a commercial graphical game.) Stick a comment in your game's 
"about" text if you want to make sure everyone knows what happened.

Legally: puzzles are much too abstract a territory to involve copyright 
law at all. 

(You could conceivably *patent* a game puzzle -- I know there are patents
on board games, based on the "method of manipulating the pieces", ie the
movement rules. But I'm certain that Activision hasn't patented ZGI.)

> Has this happened to anybody else?

Certainly. I didn't play Unnkulia 2 until long after I'd written "Change 
in the Weather." Turns out one puzzle is nearly identical. Oopsie. 

But nobody complained. Dozens of games have mazes, and does anyone 
complain? (Well, yes. But they're not complaining that the idea is 
copied. :-)

--Z

-- 

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


From david_lebling@avid.com Mon Nov 24 10:45:43 MET 1997
Article: 31214 of rec.arts.int-fiction
Path: news.lth.se!solace!news.ecn.uoknor.edu!feed1.news.erols.com!cpk-news-hub1.bbnplanet.com!cam-news-feed5.bbnplanet.com!news.bbnplanet.com!avid.com!news@avid.com
From: David Lebling <david_lebling@avid.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: IF history quiz [possibly off-topic]
Date: Wed, 19 Nov 1997 11:32:46 -0500
Organization: Avid Technology
Lines: 17
Message-ID: <347314AE.426C@avid.com>
References: <01bcf2b2$95165320$LocalHost@bertha>
NNTP-Posting-Host: cilantro.avid.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01Gold (WinNT; I)
Xref: news.lth.se rec.arts.int-fiction:31214

HairBrain wrote:
> Did people make a lot (if any) adventure games in between the release of
> Adventure and the release of Zork?

The only one we ever saw that was roughly contemporaneous with Zork (and
I'm talking the original mainframe Zork, not the much later Infocom
release) was called Haunt, and it came from CMU, if I recall correctly.
There was never a commercial version of it that I've heard of.

If you're referring to the commercial releases of Adventure and Zork,
there were lots of games in between, but that's because Zork had to wait
for the technology to catch up with its excessive size -- it was, I
think, the first game that _required_ a disk (a floppy disk, not a hard
disk).

      Dave Lebling
      (Orig. Imp.)


From winalski@lspace.zko.dec.com Mon Nov 24 10:46:00 MET 1997
Article: 31231 of rec.arts.int-fiction
Path: news.lth.se!newsfeed.sunet.se!news99.sunet.se!news01.sunet.se!130.240.42.8.MISMATCH!feed1.news.luth.se!luth.se!feed1.news.erols.com!cpk-news-hub1.bbnplanet.com!su-news-hub1.bbnplanet.com!news.bbnplanet.com!news1.digital.com!pa.dec.com!usenet
From: Paul Winalski <winalski@lspace.zko.dec.com>
Newsgroups: rec.arts.int-fiction
Subject: Re: IF history quiz [possibly off-topic]
Date: Wed, 19 Nov 1997 18:53:53 -0500
Organization: Digital Equipment Corporation
Lines: 18
Message-ID: <34737C11.A9595E9F@lspace.zko.dec.com>
References: <01bcf2b2$95165320$LocalHost@bertha> <347314AE.426C@avid.com>
NNTP-Posting-Host: lspace.zko.dec.com
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 4.01 [en] (WinNT; I)
X-Priority: 3 (Normal)
Xref: news.lth.se rec.arts.int-fiction:31231

David Lebling wrote:
> 
> The only one we ever saw that was roughly contemporaneous with Zork (and
> I'm talking the original mainframe Zork, not the much later Infocom
> release) was called Haunt, and it came from CMU, if I recall correctly.
> There was never a commercial version of it that I've heard of.

Yes, Haunt was done at CMU.  It was written as a non-procedural,
rule-based system in the OPS family of rule-based programming
languages being done at CMU at the time.  The version that I
saw was in OPS-5 (and a rewrite of an earlier one in OPS-4, if
I recall correctly).  At the time of its release, it was the
largest rule-based system in existence.

Getting Haunt to work was one of the more fun parts of our
OPS-5 compiler project many years ago.

--PSW


From dlamb@cgo.wave.ca Mon Nov 24 10:46:08 MET 1997
Article: 31234 of rec.arts.int-fiction
Message-ID: <34739EA1.977B5936@cgo.wave.ca>
Date: Wed, 19 Nov 1997 21:21:21 -0500
From: David Alex Lamb <dlamb@cgo.wave.ca>
X-Mailer: Mozilla 4.03 [en] (WinNT; U)
MIME-Version: 1.0
Newsgroups: rec.arts.int-fiction
Subject: Haunt (was Re: IF history quiz)
References: <01bcf2b2$95165320$LocalHost@bertha> <347314AE.426C@avid.com>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
NNTP-Posting-Host: cgowave-17-253.cgocable.net
Lines: 12
Path: news.lth.se!solace!mn6.swip.net!nntp.uio.no!www.nntp.primenet.com!globalcenter1!news.primenet.com!nntp.primenet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!news1.bellglobal.com!news.uunet.ca!news.cgocable.net!cgowave-17-253.cgocable.net
Xref: news.lth.se rec.arts.int-fiction:31234

David Lebling wrote:
> 
> HairBrain wrote:
> > Did people make a lot (if any) adventure games in between the release of
> > Adventure and the release of Zork?
> 
> The only one we ever saw that was roughly contemporaneous with Zork  > was called Haunt, and it came from CMU, if I recall correctly.

I believe Haunt was post-Zork, but perhaps not much post.  It was
written by John Laird (an AI grad student at CMU in the late 70's and
very early 80's) as an exercise in writing code in Production Systems (I
think he used ops5, but am not sure).


From badger@aquarius.scs.uiuc.edu Mon Nov 24 10:47:36 MET 1997
Article: 31315 of rec.arts.int-fiction
Path: news.lth.se!solace!mn6.swip.net!nntp.uio.no!news.uoregon.edu!logbridge.uoregon.edu!ais.net!vixen.cso.uiuc.edu!aquarius!badger
From: badger@aquarius.scs.uiuc.edu (Jonathan Badger)
Newsgroups: rec.arts.int-fiction
Subject: Re: IF history quiz [possibly off-topic]
Date: 24 Nov 1997 02:39:07 GMT
Organization: University of Illinois at Urbana-Champaign
Lines: 28
Message-ID: <65apcb$d5q$1@vixen.cso.uiuc.edu>
References: <01bcf2b2$95165320$LocalHost@bertha> <347314AE.426C@avid.com> <655lj2$b71$1@thorn.cc.usm.edu> <816.265T1039T5643264@mistral.co.uk> <657iu9$b8e$1@flood.xnet.com> <Pine.SGI.3.95L.971123134912.12127I-100000@ebor.york.ac.uk>
NNTP-Posting-Host: aquarius.scs.uiuc.edu
Xref: news.lth.se rec.arts.int-fiction:31315

Den of Iniquity <dmss100@york.ac.uk> writes:

>On 22 Nov 1997, Jason Compton wrote:

>>Mercifully, the tape phenomenon on the 64 died a quick death in the US by
>>1985, but I know it lived a long life in the UK and Europe.

Of course the C64 had the unique property of having a disc drive
nearly as painfully slow as a tape drive, so maybe the folks in the
Old World weren't as crazy as they seem...

>In UK (and Europe?) disk drives were not widely sold, were about four
>times as expensive as the tape drives and had poor reputations. [...]

Oh, tape drives were much cheaper in the US than disc drives too, but
we were willing to pay the extra price here. Even as early as 1982,
when I got my first computer (a Franklin Ace, a clone of the Apple
][+) disc drives were considered a necessity, not a luxury (at least
among Apple ][ users). Part of the reason was Americans generally
bought computers because they wanted them for a useful reason (word
processing or spreadsheets), and then only later discovered computers
were fun. 

In constrast I get the feeling that Brits simply bought computers then
not because they had a defined task in mind but because they just
thought computers were neat -- why else would anyone in their right
mind purchase Sinclair ZX-81s as the Brits did in droves?  Not that I
disagree that computers are inherently neat, of course.


From kjfair@midway.uchicago.edu Tue Nov 25 13:53:53 MET 1997
Article: 31342 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!fu-berlin.de!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!newsfeed.internetmci.com!128.174.5.49!vixen.cso.uiuc.edu!uchinews!kjfair
From: kjfair@midway.uchicago.edu (Kenneth Fair)
Subject: Re: [Inform] "I don't know the word 'vocabulary'."
X-Nntp-Posting-Host: ntcs-ip81.uchicago.edu
Message-ID: <kjfair-2411971650350001@ntcs-ip81.uchicago.edu>
Sender: news@midway.uchicago.edu (News Administrator)
Organization: University of Chicago School of Law
X-Newsreader: MT-NewsWatcher 2.3.5
References: <3477ED1E.7A02@ucla.edu> <ant231120868M+4%@gnelson.demon.co.uk> <3478FD9B.6461@hatch.net>
Date: Mon, 24 Nov 1997 22:50:33 GMT
Lines: 93
Xref: news.lth.se rec.arts.int-fiction:31342

In article <3478FD9B.6461@hatch.net>, jeff@hatch.net wrote:

>Yeah, Inform's method is certainly better.  One of my first experiences
>with Infocom's parser was when I saw the garlic in Zork I and
>immediately "cheated" to find out that a vampire existed somewhere. 
>Very bad.  But one of my first experiences with Inform's parser was when
>I tried to "pull on ropes" in Curses, and the error message I got fooled
>me into thinking that Inform didn't understand the word "ropes."  A dumb
>mistake on my part, certainly.

I agree wholeheartedly with Graham's choice of making the default ambiguous.
In fact, I think he's done a fantastic job with the library of providing 
sensible default replies.  They are very difficult to do in the general
case, when you can't exactly be sure if the input is singular or plural,
or male or female, or what have you.  Try going through the library
messages in English.h and testing them against various inputs.  You'll
find that they work in virtually every situation you can think of.

>What I'd prefer, though, is for the parser to simply pretend not to know
>the vocabulary for specific objects until you see them for the first
>time.  That solves the Infocom problem, but still gives quite specific
>messages about which words it doesn't understand.  (I haven't
>implemented this feature yet on my system; right now I'm using the
>Infocom method.)

Isn't this more confusing?  Then if I say, "pick up the burin" before I 
have seen the burin, it will reply, "I don't know the word 'burin.'"
But after you see it, it will know the word?  This seems even less
consistent.

In any case, this can be implemented in Inform by adding a global "seen" 
attribute to all objects that have been in scope at some point during the
game.  Then, you check to see if the object has the seen attribute before
performing a CANTSEE_PE parser error fix.

Let me provide the code.  I will use a ParserError routine instead of 
the LibraryMessages object I used before.  The two methods should be
interchangeable, as far as I can tell, but this provides a bit better
safekeeping with respect to future compatibility.

First, define the attribute:

   Attribute seen;

We need a BeforeParsing routine to give seen to all objects in scope:

   [ BeforeParsing x;
      objectloop (x) if (TestScope(x)) give x seen;
   ];


Then define the ParserError entry point routine:

[ IsAWordIn w obj prop   k l m;
   k = obj.&prop; l = (obj.#prop)/2;
   for (m = 0 : m < l : m++)
      if (w==k-->m) rtrue;
   rfalse;
];
         
[ ParserError pe i obj;
      if (pe == CANTSEE_PE) {
         wn = 1;
         while (i = NextWordStopped()) {
            if (i == -1) "You can't see that here.";
!----------
            objectloop (obj)
               if (IsAWordIn(i, obj, name) && obj hasnt seen) jump HaventSeenIt;
!----------
         }
.HaventSeenIt;
         print "^I don't know the word ~";
         for (i = 0 : i < WordLength(wn-1) : i++) {
            print (char) WordAddress(wn-1)->i;
         }
         ".~";
      }
];

The primary problem with this routine is that it does not work for objects
that have a parse_name routine.  The two lines between the dividers test
to see if the word under consideration can be found in the name property
of all objects in the game.  Of course, objects that provide a parse_name
routine will not have their names in the name property, so this section
would have to be modified to take that into account.

Overall, I'd say the current Inform reply is probably the best bet.

-- 
KEN FAIR - U. Chicago Law  | <http://student-www.uchicago.edu/users/kjfair>
Of Counsel, U. of Ediacara | Power Mac! | CABAL(tm) | I'm w/in McQ - R U?
      "Any smoothly functioning technology will be 
       indistinguishable from a rigged demo."  Isaac Asimov


From dgatewood@bigfoot.com Thu Nov 27 09:27:30 MET 1997
Article: 31421 of rec.arts.int-fiction
Path: news.lth.se!solace!xinit!newsfeed5.telia.com!masternews.telia.net!news-out.internetmci.com!newsfeed.internetmci.com!199.60.229.5!newsfeed.direct.ca!Supernews60!supernews.com!Supernews69!not-for-mail
From: Dave Gatewood <dgatewood@bigfoot.com>
Newsgroups: rec.arts.int-fiction,rec.games.int-fiction
Subject: Re: Hard puzzles in IF?
Date: Wed, 26 Nov 1997 19:07:58 -0500
Organization: All USENET -- http://www.Supernews.com
Lines: 10
Message-ID: <347CB9DE.15A9@bigfoot.com>
References: <347A38CD.1D44E957@telegram.infi.net> <Pine.A41.3.95L.971125155003.51998E-100000@login4.isis.unc.edu> <65fn3l$342$1@lyra.csx.cam.ac.uk> <Pine.SGI.3.95L.971126115100.13933A-100000@ebor.york.ac.uk>
Reply-To: dgatewood@bigfoot.com
NNTP-Posting-Host: 6900@207.42.201.70
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 3.01 (Win95; I)
Xref: news.lth.se rec.arts.int-fiction:31421 rec.games.int-fiction:27963

Den of Iniquity wrote:
> 
> Y'know, I think that kind of puzzle would be just too hard. Why can't
> people think up some easier puzzles, like maybe having a rug on the floor
> that you have to lift up to find a TRAPDOOR leading to some hidden
> dungeons! That would be cool.

Ack!  Would you please include spoiler warnings?  I'm gonna try one of
those so-called "text adventures" I keep hearing about, just as soon as
I catch that gosh-durn Wumpus.  (If only this printer weren't so slow.)


From erkyrath@netcom.com Thu Nov 27 15:01:32 MET 1997
Article: 31416 of rec.arts.int-fiction
Newsgroups: rec.arts.int-fiction
Path: news.lth.se!feed1.news.luth.se!luth.se!sunqbc.risq.qc.ca!newsfeed.direct.ca!ais.net!ix.netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Inform Designer's Manual .pdf version
Message-ID: <erkyrathEK9n0M.40C@netcom.com>
Organization: Netcom On-Line Services
X-Newsreader: TIN [version 1.2 PL2]
References: <65ctvh$q3o@kruuna.Helsinki.FI> <01bcf9dd$532edcc0$LocalHost@default> <erkyrathEK8KML.Iw8@netcom.com> <65gb6t$avl$1@nntp5.u.washington.edu> <fake-mail-2611970008370001@van-as-14a04.direct.ca>
Date: Wed, 26 Nov 1997 18:21:58 GMT
Lines: 72
Sender: erkyrath@netcom7.netcom.com
Xref: news.lth.se rec.arts.int-fiction:31416

Neil K. (fake-mail@anti-spam.address) wrote:

> > To get slightly (only slightly) back on topic, does anyone understand
> > why formatting that looks good on screen (like the Usenet standard
> > no-ident, skip-between-paragraphs) looks bad in print and vice versa?

>  I think quite a lot of factors come into play.

[Much stuff about physical display configuation: resolution, font, 
contrast, luminance, portability.]

Yup, I agree with all of that.

>  I think one of the biggest limitations that monitors place on prose is
> that it tends to chop text up into little bite-sized chunks. Read a book -
> you're going to read long paragraphs without a problem. Long pages. But a
> typical IF game has little screen-sized bits. Anything longer seems
> annoyingly excessive. Take Dave Baggett's Legend, for example. In (what I
> believe was) an attempt to lend a more literary quality to his work, he
> had fairly lengthy room descriptions and cut scenes from time to time.
> These multi-paragraph chunks wouldn't have seemed particularly long in a
> book, but I feel that they helped undermine Legend's popularity somewhat,
> as people were a bit overwhelmed with text. (particularly the
> introduction, which is several screens long on a typical monitor)

However, I don't agree with this. I'm pretty sure that the discomfort IF 
players feel with long text blocks is entirely a matter of custom; we're 
used to short pieces of text between input, not long ones. (Or perhaps 
it's more like "one event per input, not many events", although that's 
not exactly right either.)

Yes, staring at a lot of CRT text is uncomfortable, but when you play a
computer game you stare at the CRT for *hours*, regardless. I don't think
typing frequency makes that much difference. 

I think that if we had perfect dynamic paper, which felt and looked just 
like printed paper but was computer controlled, the long text scenes in 
_Legend_ would be just as uncomfortable. 

> Dave
> tried to address this by introducing the innovation of user-configurable
> paragraph breaks (optional blank lines between paragraphs) but I don't
> know if that could go far enough, ultimately.

I keep meaning to add better paragraph support to MaxZip/MaxTADS. User 
control of inter-paragraph spacing (which *should* be controlled by the 
interpreter anyway, not the game program) and optional paragraph 
indentation. (Or outdentation, if desired.) Combined with the 
proportional fonts I already support, it could make a big difference.

>  I'm facing this lengthy-text problem with my game in progress myself.
> There are a handful of key scenes with multi-screen pieces of text. On a
> big GUI monitor with a nice proportional font, they don't look too bad,
> though they're still a bit long. But on an 80x25 text-only display they
> seem really rather lengthy.

Hm. I always played _Legend_ in a large window with a proportional font; 
I never even tried it in 80x25. I guess that would be even worse. 

>  I wonder how IF would have developed if we had output devices that
> produced print-quality dynamic output. Would works of IF contain more
> prose-length (hopefully not prosaic!) chunks or would we still have the
> short, clipped, paragraphs typical to contemporary IF?

As I said, I think the latter.

--Z

-- 

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


