Comments about AGT.
Neil K. Guy
Well, thanks to Pierre Tremblay you can now play AGT games on Macs. Up
until now I haven't really participated in the "AGT rules / AGT sucks"
debate, largely because the question was somewhat academic for me. My
primary platform is Macintosh, and neither the AGT compiler nor runtime
have worked on modern Macs for some time.
Anyway, I downloaded CompileAgt and the new version of MultiAventures
and grabbed the source to a dozen odd games from ftp.gmd.de. And, well, I
must admit I'm not exactly bowled over.
I don't want to be too rude, since I know there are some AGT fans around
here, (the co-author of the system not least among them) but the games I
tried out were, frankly, pretty dreadful. Were they all written by young
children who can't spell or construct coherent sentences, for instance?
I'm going to sound like an impossible snob, but I felt I was being
presented reams of examples of how not to write from a remedial English
course.
And then, questions of literary style and intelligent game design aside,
there appear to be legions of outright bugs:
You are in a small room with a low roof. The real distinguishing
feature of this room lies in its exits. There are doorways to the
south and east. To the north there's a large arched door set into the
wall. It's closed.
What Now? open door
Open what door? There isn't any closed door here!
What Now? north
Uh, I hate to be a party-pooper, but there's a door in the way.
Or like this:
What Now? take all
You're already standing up. (And it's a good thing too! You have work to
do!)
I've also noticed the interpreter doing weird stuff like this:
What Now? look
I don't understand 't' as a verb.
Are these bugs in the new Mac port of AGT or are they normal, as it
were? Also, are there any *decent* AGT games out there? I've heard people
speak highly of Shades of Grey, but the source isn't available from
if-archive and the only version I've seen has been a PC executable. I
don't want to write AGT off completely until I've seen what people
consider the best works done in the language.
All this is making me begin to wonder if the admittedly elitist view
that if you make an authoring system *too* easy to use you get somewhat
slipshod games. There are definitely some dreadful TADS and Inform games
out there, but the vast majority are a heck of a lot better than the AGT
products I looked at. Is it that the somewhat steeper learning curve
induces people to take more care in writing their games? Or is it that
TADS and Inform users can benefit from more sophisticated libraries and
parsers? Or is it because people have been reading the excellent manuals
that come with TADS and Inform? I don't know.
- Neil K.
--
the Vancouver CommunityNet * http://www.vcn.bc.ca/
(formerly the Vancouver Regional FreeNet)
Michael Detlefsen
> Well, thanks to Pierre Tremblay you can now play AGT games on Macs. Up
> until now I haven't really participated in the "AGT rules / AGT sucks"
> debate, largely because the question was somewhat academic for me. My
> primary platform is Macintosh, and neither the AGT compiler nor runtime
> have worked on modern Macs for some time.
I'm using AGT on a PowerPC. The compiler and runtime program swork just fine.
> I don't want to be too rude, since I know there are some AGT fans around
> here, (the co-author of the system not least among them) but the games I
> tried out were, frankly, pretty dreadful. Were they all written by young
> children who can't spell or construct coherent sentences, for instance?
> I'm going to sound like an impossible snob, but I felt I was being
> presented reams of examples of how not to write from a remedial English
> course.
I agree with you on the general quality of the writing. Somnetimes I feel
as if English were a second or third language of some game writers.
> And then, questions of literary style and intelligent game design aside,
> there appear to be legions of outright bugs:
>
> You are in a small room with a low roof. The real distinguishing
> feature of this room lies in its exits. There are doorways to the
> south and east. To the north there's a large arched door set into the
> wall. It's closed.
>
> What Now? open door
>
> Open what door? There isn't any closed door here!
>
> What Now? north
>
> Uh, I hate to be a party-pooper, but there's a door in the way.
>
> Or like this:
>
> What Now? take all
>
> You're already standing up. (And it's a good thing too! You have work to
> do!)
>
> I've also noticed the interpreter doing weird stuff like this:
>
> What Now? look
>
> I don't understand 't' as a verb.
>
> Are these bugs in the new Mac port of AGT or are they normal, as it
> were? Also, are there any *decent* AGT games out there? I've heard people
> speak highly of Shades of Grey, but the source isn't available from
> if-archive and the only version I've seen has been a PC executable. I
> don't want to write AGT off completely until I've seen what people
> consider the best works done in the language.
Having worked with AGT (and three of four others), I would say from the
above examples that someone was just a sloppy programmer.
> All this is making me begin to wonder if the admittedly elitist view
> that if you make an authoring system *too* easy to use you get somewhat
> slipshod games. There are definitely some dreadful TADS and Inform games
> out there, but the vast majority are a heck of a lot better than the AGT
> products I looked at. Is it that the somewhat steeper learning curve
> induces people to take more care in writing their games? Or is it that
> TADS and Inform users can benefit from more sophisticated libraries and
> parsers? Or is it because people have been reading the excellent manuals
> that come with TADS and Inform? I don't know.
>
> - Neil K.
>
> --
> the Vancouver CommunityNet * http://www.vcn.bc.ca/
> (formerly the Vancouver Regional FreeNet)
Mike Detlefsen
David Kinder
Neil K. Guy (n...@vcn.bc.ca) wrote:
: Are these bugs in the new Mac port of AGT or are they normal, as it
: were? Also, are there any *decent* AGT games out there? I've heard people
: speak highly of Shades of Grey, but the source isn't available from
: if-archive and the only version I've seen has been a PC executable. I
: don't want to write AGT off completely until I've seen what people
: consider the best works done in the language.
Shades of Gray, Cosmoserve, and Multi-dimensional Thief spring to mind. I'm
no fan of AGT as an authoring system, but some people have written good
games with it. It's just that a lot of people have written bad games with
it. Trying AGT games at random isn't recommended :-) SoG is probably the
best, despite it's flaws.
: All this is making me begin to wonder if the admittedly elitist view
: that if you make an authoring system *too* easy to use you get somewhat
: slipshod games.
I think it's just a question of expectation. Back in the 1980s no-one
expected any amateur text adventures to be any good - so they weren't.
David
Nulldogma
> Is it that the somewhat steeper learning curve
> induces people to take more care in writing their games? Or is it that
> TADS and Inform users can benefit from more sophisticated libraries and
> parsers? Or is it because people have been reading the excellent manuals
> that come with TADS and Inform? I don't know.
Also, most AGT games were written in the late '80s, when the standards for
home-brewed I-F were a lot lower than they are now. Now that there's a
solid base of well-written independent I-F out there (mostly spawned by
TADS and Inform), I'd think most people would be embarrassed to release
anything too amateurish and buggy.
Neil
---------------------------------------------------------
Neil deMause ne...@echonyc.com
http://www.echonyc.com/~wham/neild.html
---------------------------------------------------------
Robert A. DeLisle
David Kinder (kin...@teaching.physics.ox.ac.uk) wrote:
: Shades of Gray, Cosmoserve, and Multi-dimensional Thief spring to mind. I'm
: no fan of AGT as an authoring system, but some people have written good
: games with it. It's just that a lot of people have written bad games with
: it. Trying AGT games at random isn't recommended :-) SoG is probably the
: best, despite it's flaws.
: David
I want to point out that the graphic version of Multi-Dimensional Thief
was written with a generator that the author wrote and not with AGT.
The earlier version was AGT. The file at some sites is buggy. The AGT
file at gmd.de (if it still is there) should not be buggy.
[I sent a 'good' file to replace the buggy one.]
My friend that wrote Klaustrophobia with AGT, and our testing group,
struggled mightily to make it bug-free. It took 18 months of frantic
emails. The AGT authors wrote special passages for her.
Bob Adams
In article <532j6o$4...@milo.vcn.bc.ca>, "Neil K. Guy" <n...@vcn.bc.ca>
writes
> Well, thanks to Pierre Tremblay you can now play AGT games on Macs.
[...]
> I don't want to be too rude, since I know there are some AGT fans around
>here, (the co-author of the system not least among them) but the games I
>tried out were, frankly, pretty dreadful.
What an absolutely daft comparison! To judge the quality of the writing
utility by the quality of some third-party written games is quite
frankly, ludicrous.
Is a badly driven Rolls Royce a poorly built car?
--
Bob Adams
http://www.amster.demon.co.uk
Andrew Plotkin
Bob Adams (ams...@amster.demon.co.uk) wrote:
> > Well, thanks to Pierre Tremblay you can now play AGT games on Macs.
> [...]
> > I don't want to be too rude, since I know there are some AGT fans around
> >here, (the co-author of the system not least among them) but the games I
> >tried out were, frankly, pretty dreadful.
> What an absolutely daft comparison! To judge the quality of the writing
> utility by the quality of some third-party written games is quite
> frankly, ludicrous.
> Is a badly driven Rolls Royce a poorly built car?
But is a badly driven vehicle a badly driven Rolls Royce, or a
horse-wagon careening downhill?
(Is this metaphor a careening wagon? Sorry.)
Put it this way: build a system that even a fool can use, and guess what?
No, I'm not being rude to AGT authors here, since I still haven't played
any AGT games.
I'm being rude to people who use "WorldBuilder", on the Mac.
Phoo.
--Z
--
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the
borogoves..."
Neil K. Guy
Bob Adams (ams...@amster.demon.co.uk) wrote:
: What an absolutely daft comparison! To judge the quality of the writing
: utility by the quality of some third-party written games is quite
: frankly, ludicrous.
Is it? Well, I don't know. Most of the Inform and TADS games I've played
are at the least, competent. All of the AGT games I've played are
dreadful. That seems to me to suggest that there might possibly be some
inherent limitations of the authoring language, and that's what I want to
find out.
Carl Muckenhoupt
n...@vcn.bc.ca (Neil K. Guy) writes:
>Bob Adams (ams...@amster.demon.co.uk) wrote:
>: What an absolutely daft comparison! To judge the quality of the writing
>: utility by the quality of some third-party written games is quite
>: frankly, ludicrous.
> Is it? Well, I don't know. Most of the Inform and TADS games I've played
>are at the least, competent. All of the AGT games I've played are
>dreadful. That seems to me to suggest that there might possibly be some
>inherent limitations of the authoring language, and that's what I want to
>find out.
From what I've seen, AGT's flaws are not chiefly in its limitations.
Rather, it's all in the pitfalls. It's generally possible to do things
the right way, but it's so much easier to do things the wrong way that
that's the more usual method.
I'm thinking specifically of the way AGT extends its vocabulary.
Suppose you have an opening in the wall, large enough to put a stick of
dynamite into. The right thing to do is to make that opening an object.
The easy thing to do is to code a special syntax "PUT DYNAMITE INTO OPENING".
You and I would never do this, of course, but to someone just learning
AGT as their first programming language, it seems eminently reasonable.
Thus, you get phantom nouns that only exist for certain verbs. Is this
the fault of the language?
Well, the language allowed it, didn't it?
--
Carl Muckenhoupt | Text Adventures are not dead!
b...@tiac.net | Read rec.[arts|games].int-fiction to see
http://www.tiac.net/users/baf | what you're missing!
George Caswell
On Sun, 6 Oct 1996, Bob Adams wrote:
> In article <532j6o$4...@milo.vcn.bc.ca>, "Neil K. Guy" <n...@vcn.bc.ca>
> writes
>
> > Well, thanks to Pierre Tremblay you can now play AGT games on Macs.
> [...]
> > I don't want to be too rude, since I know there are some AGT fans around
> >here, (the co-author of the system not least among them) but the games I
> >tried out were, frankly, pretty dreadful.
>
> What an absolutely daft comparison! To judge the quality of the writing
> utility by the quality of some third-party written games is quite
> frankly, ludicrous.
>
Did you read the same message? What I read was that all the AGT games
he'd played were bad.. I don't remember him saying anything about the
system...
> Is a badly driven Rolls Royce a poorly built car?
>
One word- counterexample. The double-quoted asked for one, and now I
am, too.
....T...I...M...B...U...K...T...U... ____________________________________
.________________ _/>_ _______......[George Caswell, CS '99. 4 more info ]
<___ ___________// __/<___ /......[ http://www.wpi.edu/~timbuktu ]
...//.<>._____..<_ >./ ____/.......[ Member LnL+SOMA, sometimes artist, ]
..//./>./ /.__/ /./ <___________.[writer, builder. Sysadmin of adamant]
.//.</.</</</.<_ _/.<_____________/.[____________________________________]
</.............</...................
Phyllis902
n...@vcn.bc.ca (Neil K. Guy) wrote:
> Bob Adams (ams...@amster.demon.co.uk) wrote:
>
> : What an absolutely daft comparison! To judge the quality of the
writing
> : utility by the quality of some third-party written games is quite
> : frankly, ludicrous.
>
> Is it? Well, I don't know. Most of the Inform and TADS games I've played
> are at the least, competent. All of the AGT games I've played are
> dreadful. That seems to me to suggest that there might possibly be some
> inherent limitations of the authoring language, and that's what I want
to
> find out.
Haven't some people in this thread already pointed out that good games
have been written with AGT?
I'm sure AGT has some limitations and bugs. (I'm also sure that TADS and
Inform also have some limitations and bugs.) Since people have written
good games with AGT, clearly AGT isn't so limited that good games cannot
be written with it.
And erky...@netcom.com (Andrew Plotkin) wrote:
>
> Put it this way: build a system that even a fool can use, and guess
what?
Are you certain there's an inverse relationship between being a fool and
being an experienced programmer?
From what I've seen, AGT is one of the few systems that doesn't ask the
game designer to get a degree in computer science before he uses it. This
is not a bad thing.
So long as the develpment system is capable of producing competent games
and the author understands it, or can get his questions about it answered,
then the system can be put to good use.
Let's not begrudge non-programmers the ability to write IF.
- Phyllis
Andrew Plotkin
Carl Muckenhoupt (b...@max.tiac.net) wrote:
> From what I've seen, AGT's flaws are not chiefly in its limitations.
> Rather, it's all in the pitfalls. It's generally possible to do things
> the right way, but it's so much easier to do things the wrong way that
> that's the more usual method.
> I'm thinking specifically of the way AGT extends its vocabulary.
> Suppose you have an opening in the wall, large enough to put a stick of
> dynamite into. The right thing to do is to make that opening an object.
> The easy thing to do is to code a special syntax "PUT DYNAMITE INTO OPENING".
> You and I would never do this, of course, but to someone just learning
> AGT as their first programming language, it seems eminently reasonable.
> Thus, you get phantom nouns that only exist for certain verbs. Is this
> the fault of the language?
> Well, the language allowed it, didn't it?
Hm. In a sense, Inform has the same weakness: it's easier to do
Extend "put" * "dynamite" "into" "opening"
than to make objects and add "Receive" clauses. Why don't people do this?
I think it's because of the big fat manual and the big fat standard
library, which make it somewhat clear how to proceed. (Not *totally*
clear. :-)
And the very large mass of sample code; but after N years, AGT must have
had a lot of sample code. Did it just suck?
Honestly -- not to overinflate certain egos -- I think there's simply a
tradition of quality that got established by the earliest TADS and Inform
games. It's not that it's impossible to do good AGT games, but people
didn't get into the habit.
Nulldogma
> From what I've seen, AGT is one of the few systems that > doesn't ask
the game designer to get a degree in
> computer science before he uses it. This
> is not a bad thing.
Luckily, Mike Roberts accepted my BA in sociology and let me register TADS
anyway.
Kevin P. Soucy
null...@aol.com (Nulldogma) wrote:
>Also, most AGT games were written in the late '80s, when the standards for
>home-brewed I-F were a lot lower than they are now. Now that there's a
>solid base of well-written independent I-F out there (mostly spawned by
>TADS and Inform), I'd think most people would be embarrassed to release
>anything too amateurish and buggy.
Aye. And a point that I thought of while I was at work today is that
not untill recently was the Internet affordably avalable to the
average person. AGT just happened to be the only adventure system
avalable on neighborhood BBS's, but because there was hardly anyone
else in the area who shared that passion for text adventures,
prospective AGT programmers never had anyone to share ideas and
techniques with. So, cruddy games evolved from ideas that at one time
might have actually been good. Just look at some of the most dismal
excuses for AGT games like "Aliens ate My Cardigan"....now think of
how funny that game might have actually been if someone had given the
author a few pointers? That title almost sounds like a Doug Adams book
title.<g> My point is that almost all of those old AGT games have
some seed of an idea deep inside buried under the crap.
"The AGT/AGATE Guy"
Stee...@Mindspring.Com
Bob Adams
In article <Pine.LNX.3.95.961007064138.1733C-
100...@adamant.res.wpi.edu>, George Caswell
<timb...@adamant.res.wpi.edu> writes
>am, too.
>
>....T...I...M...B...U...K...T...U... ____________________________________
>.________________ _/>_ _______......[George Caswell, CS '99. 4 more info ]
><___ ___________// __/<___ /......[ http://www.wpi.edu/~timbuktu ]
>...//.<>._____..<_ >./ ____/.......[ Member LnL+SOMA, sometimes artist, ]
>..//./>./ /.__/ /./ <___________.[writer, builder. Sysadmin of adamant]
>.//.</.</</</.<_ _/.<_____________/.[____________________________________]
></.............</...................
>
I have never understood any of your replies and this one is not an
exception.
Adam J. Thornton
Phyllis902 <phyll...@aol.com> wrote:
>Let's not begrudge non-programmers the ability to write IF.
Actually, let's.
With almost any system and its standard libraries, a non-programmer can get
a game working in short order.
The problem is, it's going to be a boring game.
Neat things only start happening when tricky problems get solved. The
problems exist in any language; however, I much prefer a language that is
flexible and general enough to handle almost any situation (e.g. _Freefall_
for the Zmachine) than a weak and inconsistent one that claims to be for
the non-programmer. *Any* system that claims you can use it to make a
first class game without having to roll up your sleeves and do some coding
is lying to you, plain and simple.
Good IF requires both good prose and good programming. That's part of the
reason so little good IF exists.
Adam
--
ad...@phoenix.princeton.edu | Viva HEGGA! | Save the choad! | 64,928 | Fnord
"Double integral is also the shape of lovers curled asleep":Pynchon | Linux
Thanks for letting me rearrange the chemicals in your head. | Team OS/2
You can have my PGP passphrase when you pry it from my cold, dead brain.
George Caswell
I'll take that as a compliment. Someone said AGT sucked. Someone else
(You??) jumped all over them for condemning the system based on their
assuming that all AGT games sucked. All his experience, and all mine,
with AGT has been with really stupid games, so I would suggest whoever
likes AGT make some effort to provide the counterexample.
Trevor Barrie
phyll...@aol.com (Phyllis902) wrote:
>From what I've seen, AGT is one of the few systems that doesn't ask the
>game designer to get a degree in computer science before he uses it. This
>is not a bad thing.
TADS hardly requires a degree in computer science to use. The equivalent of
a first-year course, maybe, but this level of competence isn't that hard to
teach oneself.
>Let's not begrudge non-programmers the ability to write IF.
Sure. On similar ideological grounds, let's not begrudge the non-engineers
the right to build bridges, and let's not begrudge the non-architects the
right to design houses.:)
I've seen a fair number of purportedly easy-to-use game design kits (for
various types of game), and non of them have shaken my simple belief that,
if you want to make a computer program, you'd better know how to program a
computer.
Bill Hoggett
George Caswell <timb...@adamant.res.wpi.edu> wrote the immortal lines:
>On Tue, 8 Oct 1996, Bob Adams wrote:
>> George Caswell <timb...@adamant.res.wpi.edu> writes
>> > One word- counterexample. The double-quoted asked for one, and now I
>> >am, too.
>> >
>> I have never understood any of your replies and this one is not an
>> exception.
>>
> I'll take that as a compliment. Someone said AGT sucked. Someone else
>(You??) jumped all over them for condemning the system based on their
>assuming that all AGT games sucked. All his experience, and all mine,
>with AGT has been with really stupid games, so I would suggest whoever
>likes AGT make some effort to provide the counterexample.
First, the counterexamples: Shades Of Grey, Klaustrophobia, Multi-
Dimensional Thief, Oklib's Revenge (Must be my "Be Nice To Sue" day <g>).
Sorry Bob, haven't played Helvera so I can't comment.
Second, my opinion: AGT sucks! And here's why I think so: AGT, by it's
very nature, attracts programming newbies. This is fine, except that
newbies, due to inexperience and sometimes ignorance, rely heavily on
defaults. And, like it or not, AGT's defaults are diabolical. Also,
it doesn't help when adventures like the ones written by Ev Cheney
are included as examples. Anyone who's played these will know exactly
what I mean. People will follow the percieved standards, and if given
a bad example, they will follow that too.
Kevin Soucy is talking about creating a SuperAGT compiler (It won't
be called that if Mark Welch has anything to do with it). If the
idea is to keep it as accessible to beginners as the original was,
then the new compiler will have to be MUCH BETTER than TADS or Inform.
This is because it will have to cater for all those situations that the
programmer cannot be bothered or has forgotten to code special code for.
It will have to do this without any noticeable deterioration in the
performance of the finished game. In short, the new compiler will have
to write part of the game itself. Somehow, I don't think that's easy.
Also, from Kevin's posts, I understand this is not his aim anyway.
Just my tuppence...
Bill Hoggett (aka BeeJay) <mas.su...@easynet.co.uk>
My opinions are always those of my employers.
...unless they deny it, of course.
Matthew Amster-Burton
mas.su...@easynet.co.uk (Bill Hoggett) wrote:
>Second, my opinion: AGT sucks! And here's why I think so: AGT, by it's
>very nature, attracts programming newbies. This is fine, except that
>newbies, due to inexperience and sometimes ignorance, rely heavily on
>defaults. And, like it or not, AGT's defaults are diabolical. Also,
>it doesn't help when adventures like the ones written by Ev Cheney
>are included as examples. Anyone who's played these will know exactly
>what I mean. People will follow the percieved standards, and if given
>a bad example, they will follow that too.
Wow, I'm glad someone said this before I did. For god's sake, let's
leave programming to the programmers, okay? Okay, go ahead and create
a database front-end with VB or something, but if you want to put
together something as complex as a text adventure (which one
implementor likened to writing a spreadsheet application), it would
behoove you to join the ranks of actual programmers--really, we're not
an exclusive bunch.
I find that AGT presents an interesting conundrum. I am more
frustrated by a bad TADS game than a bad AGT game, because I expect
that someone who took the time to learn TADS would also take the time
to do the game right. Thankfully, I have almost always found this to
be the case. The hack who picks up a bit of a language and cobbles
together a "Detective" clone does himself and the rest of us a
disservice.
The most interesting parallel I can think of, believe it or not, is
between IF and singing. When I began taking voice lessons, my
attitude about singing was: I would like to be able to sing in style
X or hit note Q ("in the key of Q"), but my voice just doesn't work
that way. I was immediately told and quickly learned to believe that
this belief is useless and untrue: if you only work within the
perceived limits of your system, you will never approach your full
potential.
Using a complete IF system is necessary for creating good IF. If
you're constantly worry about the limitations of AGT, you're being
distracted and constrained from making good art. This is not to say
that TADS and Inform don't have limitations of their own--or that the
human voice does not (ever heard the Tuvan throat singers, though?).
But TADS and Inform encourage a style of game development based on
imagination, not programming traps.
TADS and Inform are better languages for IF than AGT. (I should
probably throw Hugo in there, too, but I've never tried it.) Given
that learning to write IF--in the absence of collaboration--means
learning to program, why not do yourself a favor and pick up one of
the premier languages instead of AGT?
This is not to say that we couldn't use better tools. A Borland or
M$-style IDE geared towards an IF language would really turn me on.
Matthew
----------------------------------------------------------------
The CIA introduced crack to South Central LA to fund the Contras
http://www.sjmercury.com/drugs/
George Caswell
On 10 Oct 1996, Bill Hoggett wrote:
>
> George Caswell <timb...@adamant.res.wpi.edu> wrote the immortal lines:
>
> >On Tue, 8 Oct 1996, Bob Adams wrote:
>
> >> George Caswell <timb...@adamant.res.wpi.edu> writes
> >> > One word- counterexample. The double-quoted asked for one, and now I
> >> >am, too.
> >> >
>
> >> I have never understood any of your replies and this one is not an
> >> exception.
> >>
>
> > I'll take that as a compliment. Someone said AGT sucked. Someone else
> >(You??) jumped all over them for condemning the system based on their
> >assuming that all AGT games sucked. All his experience, and all mine,
> >with AGT has been with really stupid games, so I would suggest whoever
> >likes AGT make some effort to provide the counterexample.
>
> First, the counterexamples: Shades Of Grey, Klaustrophobia, Multi-
> Dimensional Thief, Oklib's Revenge (Must be my "Be Nice To Sue" day <g>).
> Sorry Bob, haven't played Helvera so I can't comment.
>
Thank you for the counterexamples. Nice to know that there's that much
more good effort being made in IF.
> Kevin Soucy is talking about creating a SuperAGT compiler (It won't
> be called that if Mark Welch has anything to do with it). If the
> idea is to keep it as accessible to beginners as the original was,
> then the new compiler will have to be MUCH BETTER than TADS or Inform.
I don't doubt that a new language, written completely from scratch,
-could- surpass TADS or Inform.... I personally think, however, that if
something tries too hard to be too many things, it will be nothing. A
good IF language or a good beginners language... If AGT -can- make it
work both ways, I'll be impressed...
> This is because it will have to cater for all those situations that the
> programmer cannot be bothered or has forgotten to code special code for.
Such a thing may be handy, but certainly not fool-proof. If the
compiler could handle everything the programmer forgot about, the
programmer would be out of a job.
> My opinions are always those of my employers.
>
> ...unless they deny it, of course.
My opinions are my own. They also happen to be yours, yours, yours,
and yours, unless of course, you happen to be completely stupid and wrong.
<g>
Kevin P. Soucy
mas.su...@easynet.co.uk (Bill Hoggett) wrote:
>Kevin Soucy is talking about creating a SuperAGT compiler (It won't
>be called that if Mark Welch has anything to do with it). If the
>idea is to keep it as accessible to beginners as the original was,
>then the new compiler will have to be MUCH BETTER than TADS or Inform.
>This is because it will have to cater for all those situations that the
>programmer cannot be bothered or has forgotten to code special code for.
>It will have to do this without any noticeable deterioration in the
>performance of the finished game. In short, the new compiler will have
>to write part of the game itself. Somehow, I don't think that's easy.
>Also, from Kevin's posts, I understand this is not his aim anyway.
Definately not my aim.<G> Elsewhere in R.A.I-F you can find my
re-emergance post, as well as the explanation on why my system will be
called Text AGATE. My aim is not to create a language, but to expand
on the AGT idea and create a code-free environment from which to write
your games. What do I mean by code-free? Well, it's a complex issue,
so let me try to explain it. There will be three ways to write a game
with Text AGATE: Manual Coding, Menu Assisted, or From Within. Manual
coding will be the same thing as AGT where you write all the code
yourself from a text editor and then let Text AGATE compile it. Menu
Assisted coding and From Within coding are used within the Text AGATE
program itself. When you run Text AGATE it will load up like an
editor with pull-down menus at the top, and your game running on-the
fly in the main area. With Menu-Assisted, you pull down these menus
to design your rooms, nouns, creatures, commands, etc. and Text AGATE
writes the code based on your menu selections. As well as doing this
you can type commands from within the game that will automatically
pull up the required menu or make changes to the existing code, much
like an Internet MU*. Then when you are done creating your game, you
select "Compile Finished Game" from the FILE menu, and Text AGATE
writes the executable game. The significant changes from AGT are the
introduction of RPG-specific settings and meta-tolkens, ANSI graphics,
a BBS-like menu system, the ability to run other programs from within
your game (Like a BBS door), a newly formated command system, and the
ability to play multiplayer games via modem. Couple this with the
ability to import old AGT standard, big, and Master's games, and add
the fact that there are no longer any limits on the amounts of things
you can add, and you've cooked up what Text AGATE is all about.
"The AGT/AGATE Guy"
Stee...@Mindspring.Com
Bob Adams
In article <2285.685...@easynet.co.uk>, Bill Hoggett
<mas.su...@easynet.co.uk> writes
>First, the counterexamples: Shades Of Grey, Klaustrophobia, Multi-
>Dimensional Thief, Oklib's Revenge (Must be my "Be Nice To Sue" day <g>).
>Sorry Bob, haven't played Helvera so I can't comment.
Glad you cannot comment on the last one as it certainly was NOT my
intention to include anything written by me as "high quality". Helvera
for instance, was an old 64k 8-bit adventure that was easily converted
for Amiga, ST & PC users by using AGT. At the time (1990-1) AGT was the
only cross-platform utility that I could find that was capable of doing
the job. Helvera may come under the category of "old but amusing" but
should certainly not be included as a "counterexample".
No, my main point is that _any_ authoring tool is totally reliant on the
user to produce the quality. A bad author will make a rubbish adventure
whatever and however powerful the utility used is.
I did originally write a much longer reply to the points you raised but
after I 'read ahead' and saw some of the other replies and realised that
they lived on a different planet to me - I decided to give up.
Thanks for your reply but I'm now dropping this newsgroup. The noise
level from the empty vessels is just too much.
Adam J. Thornton
In article <53ju7m$c...@camel1.mindspring.com>,
Kevin P. Soucy <stee...@mindspring.com> wrote:
>ANSI graphics, a BBS-like menu system, the ability to run other programs
>from within your game (Like a BBS door)
I don't suppose you could be talked into blowing off ANSI graphics, making
it an ncurses system, and settling for an exec() call so you could write
this in a fashion such that it's portable to any Unix/Unixoid system
easily?
Bill Hoggett
In article <ygtB5IAc...@amster.demon.co.uk> Bob Adams uttered
the immortal last words:
>No, my main point is that _any_ authoring tool is totally reliant on the
>user to produce the quality. A bad author will make a rubbish adventure
>whatever and however powerful the utility used is.
Granted. But most bad authors don't want to spend too much time on
their creation, so TADS and Inform are usually spared their attention.
AGT, on the other hand, allows games to be created quickly, with a
minimum of effort. This invariably results in a bad game and, to a
large extent, is responsible for the huge amount of dross to be found
in the AGT archives. Someone much wiser than me once said something to the
effect of "Nothing worthwhile doing is easy". I subscribe to that view.
Personally, I don't think it is possible for authors to give an objective
view on the tool they have chosen to use. You are biased towards AGT,
I am biased towards Inform, Cardinal proclaims the qualities of Hugo,
others back TADS. There is only one objective way to judge a so called
"authoring tool", and that is by the OVERALL quality of the games created
with it. Compare and make your own judgement.
Furthermore, a point always raised by AGT supporters is:
"Let's not begrudge non-programmers the right to write I-F"
I don't. What I do begrudge non-programmers is the right to PROGRAM
i-f. All the examples I gave elsewhere of good AGT games were the result
of considerable effort on the part of the respective programmers. I
stand to be corrected, of course. But I don't expect to be.
I am a programmer, and writing my own game. When it's finished, I'll be
looking for someone to revamp my text. I may be able to design a game,
but I have no illusions about my storytelling talent (or lack of it).
Similarly, I don't think i-f writers (new or old) should be under any
illusion as to their programming abilities. AGT encourages that illusion.
>I did originally write a much longer reply to the points you raised but
>after I 'read ahead' and saw some of the other replies and realised that
>they lived on a different planet to me - I decided to give up.
That's a shame. It may have helped us to better understand your view, and
maybe opened our minds to something we have so far been blind to. Now I
guess we'll just have to stay ignorant.
>Thanks for your reply but I'm now dropping this newsgroup. The noise
>level from the empty vessels is just too much.
It's your choice. But if so many of the best current i-f authors who
disagree with you are empty vessels, then I feel sorry for you. It must
be hard to be so flawless.
Bill Hoggett (aka BeeJay) <mas.su...@easynet.co.uk>
My opinions are always those of my employers.
Bill Hoggett
On 08-Oct-96, Kevin P. Soucy <stee...@mindspring.com> allowed
himself the luxury to write:
>So, cruddy games evolved from ideas that at one time
>might have actually been good. Just look at some of the most dismal
>excuses for AGT games like "Aliens ate My Cardigan"....now think of
>how funny that game might have actually been if someone had given the
>author a few pointers? That title almost sounds like a Doug Adams book
>title.<g> My point is that almost all of those old AGT games have
>some seed of an idea deep inside buried under the crap.
No doubt. I suspect ALL i-f games have some seed of an idea behind
them. :-)
It's just that while a bad idea will usually result in a bad game, a
good idea does not necesseraly produce a masterpiece. The fact is, there's
much, much more to writing a game than having an idea.
I agree with the point about AGT authors in the late 80's not really
having access to good advice though. Now they have no excuse.
Bill Hoggett (aka BeeJay) <mas.su...@easynet.co.uk>
IF GOD IS LIFE'S SERVICE PROVIDER WHY HAVEN'T I GOT HIS I.P. NUMBER ?
John Francis
This is at least partly due to the fact that there are no decent tools
to assist the non-programmer trying to create IF.
To take your architect example above; What is wrong with being able to
specify "I want a 20x16 living room, aligned N/S, with an 8' slider in
the middle of the north wall ..." as input to an expert system which
has been loaded with a ruleset detailing the local construction codes,
the strengths of the materials being used, etc.
My wife doesn't know how to program a computer, but she can still use
a computer to fill out our tax returns, create a newsletter, etc., etc.
That said, however: currently the only place you can learn how to
create a well-structured computer program is in a general-purpose
programming environment (be it a university course, a book, or just
by experience). While a lot of what is needed to design a good game
is covered in basic computer science (elements of database design,
introduction to algorithms, ...), it is normally presented in the
framework of one (or if you are really lucky, two) programming model(s).
While C++ or Pascal have their uses, they are designed to be rather
more generalised than is required for introductory game design.
In fact this very complexity can be enough to scare off a novice
game designer. Let's face it; even Inform scares several people away,
and that's before you get into the more esoteric capabilities.
If Agate manages to get more people interested in designing computer
games, while not encouraging them to write badly structured games
or constraining them to an overly narrow subset of possibilities in
order to reduce complexity then it will be a good thing. I don't
think it is necessary for it to be able to recreate _every_ game that
has been written in TADS or in Inform (or Hugo or Alan or ....), but
I think it should be able to handle something like the original Zork.
The archives at gmd.de contain some excellent material on how to set
about designing an IF game. A tutorial based on some of this would
be a valuable addition to any game creation package.
Personally I too would like to see an IDE that would allow me to
create (and test) the basic framework of a game such as
o basic room and object descriptions
o simple connectivity
o common actions like get/drop, examine
o rudimentary constraints (isPresent, carrying, capacity, ..)
Once I'd got that sorted out I would like to push a button and have it
generate the game files for my chosen language (TADS, Inform ...)
If I wrote this I'd probably implement it in Win32s, with an Inform
back end code generator.
John Holder
Trevor Barrie (tba...@cycor.ca) wrote:
: if you want to make a computer program, you'd better know how to program a
: computer.
You just made my quotes file with this one!
In a similar vein:
"To know recursion, you must first know recursion"
--
John Holder (jho...@frii.com) http://www.frii.com/~jholder/
UNIX Specialist, Paranet Inc., Denver, Colorado, USA, Earth
Death is just God's way of dropping carrier detect...
Bill Hoggett
On 17-Oct-96 John Francis <jo...@thuridion.com> wrote:
>To take your architect example above; What is wrong with being able to
>specify "I want a 20x16 living room, aligned N/S, with an 8' slider in
>the middle of the north wall ..." as input to an expert system which
>has been loaded with a ruleset detailing the local construction codes,
>the strengths of the materials being used, etc.
Nothing, if you like monotony. What you propose would result in cities
full of houses of different sizes and orientation, yet chillingly
similar. Spooky.
Or, to take another comparison, how about giving a computer a plot,
storyline and characters and expecting it to write your novel by
filling in the narrative and dialogue. What do you get ? Not a classic,
I presume.
>My wife doesn't know how to program a computer, but she can still use
>a computer to fill out our tax returns, create a newsletter, etc., etc.
Maybe I'm just dumb, but I don't see the analogy here. If you think
for one minute that writing i-f (that's designing AND programming) is
anything like creating a newsletter, well we might be seeing a follow
up to "Detective" after all. <please say it ain't so, John>
>That said, however: currently the only place you can learn how to
>create a well-structured computer program is in a general-purpose
>programming environment (be it a university course, a book, or just
>by experience). While a lot of what is needed to design a good game
>is covered in basic computer science (elements of database design,
>introduction to algorithms, ...), it is normally presented in the
>framework of one (or if you are really lucky, two) programming model(s).
>While C++ or Pascal have their uses, they are designed to be rather
>more generalised than is required for introductory game design.
>In fact this very complexity can be enough to scare off a novice
>game designer. Let's face it; even Inform scares several people away,
>and that's before you get into the more esoteric capabilities.
C++ and Pascal are designed to be rather more generalised than is
required for introductory PROGRAM design. Any programmer who is scared
off by this fact will be told in no uncertain terms to seek an
alternative career. The initial complexity of TADS and Inform is
probably the main factor in maintaining the high standard of final
product we are witnessing from them, as they quickly scare off those
not prepared to sweat for their art.
>If Agate manages to get more people interested in designing computer
>games, while not encouraging them to write badly structured games
>or constraining them to an overly narrow subset of possibilities in
>order to reduce complexity then it will be a good thing. I don't
>think it is necessary for it to be able to recreate _every_ game that
>has been written in TADS or in Inform (or Hugo or Alan or ....), but
>I think it should be able to handle something like the original Zork.
That's a tall order there. Remember that AGATE is based on AGT and
that DID encourage and constrain as you describe. The reduction in
complexity is merely an illusion achieved at the expense of the final
result.
>The archives at gmd.de contain some excellent material on how to set
>about designing an IF game. A tutorial based on some of this would
>be a valuable addition to any game creation package.
You mean you want someone to run through how they correctly designed
a game ? I'd like to meet the first volunteer ;)
>Personally I too would like to see an IDE that would allow me to
>create (and test) the basic framework of a game such as
> o basic room and object descriptions
> o simple connectivity
> o common actions like get/drop, examine
> o rudimentary constraints (isPresent, carrying, capacity, ..)
Great...
>Once I'd got that sorted out I would like to push a button and have it
>generate the game files for my chosen language (TADS, Inform ...)
...and not so great. This would only result in BAD TADS and Inform games.
It's not the system used, but the effort expended that makes a good game
in the end.
>If I wrote this I'd probably implement it in Win32s, with an Inform
>back end code generator.
Why not write it in ANSI C, using a modular design, with the system
specific GUI and I/O routines being bolted on at the end. That way
maybe even those who don't use M$Win - heathens ;) - might be taken
on board.
In my view, creating i-f requires three main skills: designing,
writing or storytelling and finally, programming. If you can do
all three passably well you have a chance of creating a good i-f
game. If not, find someone who can fill in the gaps, but don't
expect a computer to do it. They're not very good at creative
thinking, and all the above skills require that.
Bill Hoggett (aka BeeJay) <mas.su...@easynet.co.uk>
The above is purely the result of my computer's AI. I just pressed <RETURN>
Roger Carbol
John Francis wrote:
> This is at least partly due to the fact that there are no decent tools
> to assist the non-programmer trying to create IF.
I'm going to have to go out on a limb and say that this is patently false.
For example, pencils and paper work just fine for the average
non-programmer embarking to write IF.
Of course, if you want to implement the IF on a computer, then odds are
that you might want to have some idea about creating algorithms on
a computer, often called computer programming. But there has been
lots and lots of IF created entirely outside of the sphere of
computers!
> To take your architect example above; What is wrong with being able to
> specify "I want a 20x16 living room, aligned N/S, with an 8' slider in
> the middle of the north wall ..." as input to an expert system which
> has been loaded with a ruleset detailing the local construction codes,
> the strengths of the materials being used, etc.
Many fine examples of the aforementioned expert system exist; they
are biological units called programmers. Collaboration can be a
wonderful thing, whether with an artificial intelligence or a natural one.
Roger Carbol .. r...@col.ca .. I don't know how to program the computer.
John Francis
In reply to my earlier ramblings, Bill Hoggett wrote:
>
> [earlier sections snipped]
> >My wife doesn't know how to program a computer, but she can still use
> >a computer to fill out our tax returns, create a newsletter, etc.
>
> Maybe I'm just dumb, but I don't see the analogy here. If you think
> for one minute that writing i-f (that's designing AND programming) is
> anything like creating a newsletter, well we might be seeing a follow
> up to "Detective" after all. <please say it ain't so, John>
Not from me, you won't! :-)
The point I was trying to make is that a computer can be used by the
non-computer-literate to perform a fairly complex task provided that:
1) The user understands what (s)he wants to do
2) A competent programmer, skilled in the process in question,
has put together an intelligent framework to simplify the task.
To push the newsletter analogy further; all responsibility for content
comes from the user - the page layout program just makes it much easier
to translate the design into reality.
It is possible to create postscript source files using nothing more
than vi (been there; done that), but a good WYSI(almost)WYG document
processor simplifies the task considerably.
>
> [More snippage]
>
> >If Agate manages to get more people interested in designing computer
> >games, while not encouraging them to write badly structured games
> >or constraining them to an overly narrow subset of possibilities in
> >order to reduce complexity then it will be a good thing. I don't
> >think it is necessary for it to be able to recreate _every_ game that
> >has been written in TADS or in Inform (or Hugo or Alan or ....), but
> >I think it should be able to handle something like the original Zork.
>
> That's a tall order there. Remember that AGATE is based on AGT and
> that DID encourage and constrain as you describe. The reduction in
> complexity is merely an illusion achieved at the expense of the final
> result.
I agree with the first part of this, which is why I put the qualifier
clauses in. And even if the reduction in complexity means that the
quality of the resulting game will be limited, I'd be happy to see new
games of the quality of those produced by infocom.
>
> >The archives at gmd.de contain some excellent material on how to set
> >about designing an IF game. A tutorial based on some of this would
> >be a valuable addition to any game creation package.
>
> You mean you want someone to run through how they correctly designed
> a game ? I'd like to meet the first volunteer ;)
>
> >Personally I too would like to see an IDE that would allow me to
> >create (and test) the basic framework of a game such as
> > o basic room and object descriptions
> > o simple connectivity
> > o common actions like get/drop, examine
> > o rudimentary constraints (isPresent, carrying, capacity, ..)
>
> Great...
>
> >Once I'd got that sorted out I would like to push a button and have it
> >generate the game files for my chosen language (TADS, Inform ...)
>
> ...and not so great. This would only result in BAD TADS and Inform games.
> It's not the system used, but the effort expended that makes a good game
> in the end.
Maybe I should have explained that I was straying onto another topic
here,
and this would be the *starting* point for the serious work, not the
end.
This was a description of the kind of tool *I* would like to see as an
aid
to game creation. I'd like to put all my effort into writing the
interesting
parts of the game, not in creating the boring parts. If I decide I want
to
move the cheez(tm) shoppe from Fublio to Anthar I personally would find
it
much easier to use a drag-and-drop IDE than to edit text files.
> >If I wrote this I'd probably implement it in Win32s, with an Inform
> >back end code generator.
>
> Why not write it in ANSI C, using a modular design, with the system
> specific GUI and I/O routines being bolted on at the end. That way
> maybe even those who don't use M$Win - heathens ;) - might be taken
> on board.
Because almost all of the code would be GUI and I/O routines?
The amount of IF-specific code needed would be totally dwarfed
by the code needed to handle the user interface. Unfortunately
GUI code is still the hardest stuff to port between systems.
I'd want to use Win32s so I could use the common dialog controls, etc.
I'm sure that Motif widgets would work just as well, but I haven't used
them for the last three or four years. I don't want to write UI code;
I want to concentrate on the interesting stuff (that sounds familiar ..)
Stop Press: My back end of choice is probably changing to TADS. I wanted
a system with freely available source code for the run-time system (why
write code when you can steal^H^H^H^H^Hadapt it), and TADS now
qualifies.
> In my view, creating i-f requires three main skills: designing,
> writing or storytelling and finally, programming. If you can do
> all three passably well you have a chance of creating a good i-f
> game. If not, find someone who can fill in the gaps, but don't
> expect a computer to do it. They're not very good at creative
> thinking, and all the above skills require that.
I'd never suggest using a computer to replace the skilled practitioner
of game design, or to create the descriptive narrative. But I do feel
that a large percentage of the programming part that currently has to be
done by hand could be entirely automated.
Does the fact that I use Visual C++ and all the tools it provides
(App Wizard, Class Wizard, Component Gallery, ...) mean that I write
worse code? I don't think it does. It just means that I don't have
to rewrite the supporting framework code yet again - I can concentrate
on the interesting part of the problem (I *know* I've heard that before
..)
>
> Bill Hoggett (aka BeeJay) <mas.su...@easynet.co.uk>
>
> The above is purely the result of my computer's AI. I just pressed <RETURN>
Artificial intelligence is all very well, but no substitute for the real
thing.
Jason B Dyer
Bill Hoggett (mas.su...@easynet.co.uk) wrote:
: Nothing, if you like monotony. What you propose would result in cities
: full of houses of different sizes and orientation, yet chillingly
: similar. Spooky.
I haven't seen this happen to entire cities, but there have been
a few housing developments where I have seen this happen.
Jason Dyer
jd...@u.arizona.edu
Trevor Barrie
jho...@frii.com (John Holder) wrote:
>: if you want to make a computer program, you'd better know how to program a
>: computer.
>You just made my quotes file with this one!
My net.presence has been validated at last.:)
>In a similar vein:
>"To know recursion, you must first know recursion"
That's even better. Is it yours?
John Hartnup
Trevor Barrie (tba...@cycor.ca) wrote:
: jho...@frii.com (John Holder) wrote:
: >In a similar vein:
: >"To know recursion, you must first know recursion"
: That's even better. Is it yours?
See the less snappy but more accurate version in "Lists", one
of the competition entries.
JOhn
--
-----------------------------------------------------------
John Hartnup | You can drink your weak lemon drink
sl...@ladle.demon.co.uk| now, or you can save it 'til later.
-----------------------------------------------------------