“You have probably never heard of Professor Moriarty?” said he.

“You have probably never heard of Professor Moriarty?” said he.

“Never.”

“Aye, there’s the genius and the wonder of the thing!” he cried. “The man pervades London, and no one has heard of him. That’s what puts him on a pinnacle in the records of crime.”

Of course, it’s not true that no one has heard of Brian Moriarty — this sort of slipped under the radar as far as I can tell, but Moriarty recently appeared at a M*U*S*H event and gave a little /rant on the state of text games.

Andrew Plotkin, Jota (I’m assuming Admiral Jota of Lost Pig), and Emily Short (again an assumption but I think the Emily in the transcript is the same) were there too — it’s an interesting read all around.

The log is typical of mush transcripts so I’ve liberally edited the thing into a more readable presentation — however keep in mind you lose out by not taking in the transcript in its unadulterated form.

Also note that this is an IF-heavy conversation taking place on a mush, with about equal representation from the IF and mush sides — so sometimes people are talking past each other with respect to what text games mean to them.

By the way, there’s a surprise appearance halfway through the log that I’m not including in my transcript — so check out the link above.

My edited transcript after the cut:


zarf takes a seat in the Right Front Row.
Moriarty looks around the lecture hall.
Jota takes a seat in the Right Front Row.
Emily takes a seat in the Left Front Row.

Moriarty has arrived.
Moriarty peers out over the hall to study the attendees.
Moriarty shuffles his notes importantly.
Moriarty pokes curiously at the holographic projector. “Don’t have these where I teach …”

AnneLions says, “It’s just like an overhead projector! Only… holographic.”

Zebranky has arrived.
Zebranky says, “Welcome, everyone, to the eighth(?) annual(?) Innovations and NEw Directions in Text-Based Gaming Conference! Or ITBG.”

Nämmyung applauds!
Orange_Guest applaud
Chaz claps his hands!
tramp applauds.

Zebranky says, “Our first speaker this afternoon is Professor Brian Moriarty, of such renowned titles as… well, you can read the announcement I made yourself. 🙂 And with that I’ll give him the mic. Please give him a warm welcome!”

tramp applauds.
jaybird woot!@
zarf claps
Nämmyung applauds!
AnneLions applauds and stuff.
ToadyOne cheers!
Purple_Guest applauds

Moriarty steps up to the microphone, which (he notes with some
disappointment) is NOT holographic.
tramp chuckles.
AnneLions snickers.

Moriarty says, “Thanks for the introduction, Zebranky. To begin with, I invite you all to take a look at the emote I typed a moment ago. No graphics game in the world I am aware of is capable of expressing disappointment. But here, I was able to do it easily. In 16 bytes, without a schedule or a budget. I suppose I’m preaching to the choir when I say this, but despite 20+ years of development, graphics games STILL have not even begun to approach the expressive power, subtlety and suppleness of text.”

AnneLions says, “Hear, hear.”
jaybird says, “Yeh”

Moriarty says, “Yet, it’s also obvious that these unique qualities of interactive text are not particularly valued. I was recently appointed to the position of Professor of Practice in Game Design at Worcester Polytech. This has two happy results. First, I am now truly and actually PROFESSOR Moriarty.

“Second, I am now in a position to teach students about the wonderful properties of text-based gaming.”

Orange_Guest laughs
AnneLions says, “Hooray!”
tramp applauds.
jaybird says, “Yay for text games!”

Moriarty says, “Yes, strangely, and perhaps sadly, I find myself reluctant to do so. The reason for my reluctance is simple. If I show an 18-year-old aspiring game designer a modern text game, they will simply wrinkle their nose at it. This not so much because they have to READ or TYPE to use it (although some might not be particularly excited about that). It’s simply because the actual screen appearance of modern text games is so GOD DAMNED UGLY!

“The interface I am using to address you today looks essentially identical to the interface I was using to play GEMSTONE back in 1987! I simply don’t understand why the visual appearance of text games has not evolved in the past 25 years.”

“Aside from a few IF systems capable of presenting UGLY screens that look like HTML circa 1995, text games seem to be stuck at around 1986, the year I completed my game Trinity. There is exactly ONE interpreter, Gargoyle, that even BEGINS to address the issues involved in making interactive digital typography attractive. And even Gargoyle is basically designed to mimic the DEC VT-100 terminals used by the Infocom implementors to write Zork!

“Text games enthusiasts seem curiously unconcerned about this issue. I believe that, for this excellent medium to attract new reader/players and writers, the community needs to take a long, hard look at basic presentation issues, and find ways to make these things more appealing to look at.

“Another long, hard look is needed at the natural language interface (which, to save time, I’ll call the “parser”). Despite 20+ years of hardware evolution, the parsers found in contemporary text games are essentially no different in expressive power than the one I used to write Beyond Zork. This is simply inexcusable. An average desktop machine is capable of much, MUCH more than the glorified toys we were writing for back then. But contemporary games don’t exploit this massive power at all.

“This is partly the fault of what I call Z-Code Nostalgia. Z-Code Nostalgia is a wasting disease which originated in England. It has hampered the development of IF by locking writers down to a technology standard which was cutting-edge in 1990, and laughable in 2009.

“Did I write 1990? Excuse me. I meant 1980.

“I know that better virtual machines have been implemented since then (GLULX, for instance). But these standards aren’t yet fully supported by the most popular development systems yet. And even if they WERE supported, the standard libraries of the popular dev systems aspire to no more than Beyond Zork-level parsing ability.

“I remember going to a trade show to exhibit Beyond Zork back in fall of 1987. The story was always the same. Some curious retail executive would sit down in front of a computer loaded with BZ. I would explain that you typed ‘real English sentences’ to interact. They would nod, turn to the keyboard and type WHERE AM I? End of demo.”

Zebranky grins.

Moriarty says, “Now, I know some modern games attempt to handle these embarrassments by sensing common ‘mistakes’ like this and nudging the player towards simple imperative sentences. But THAT is exactly the problem! Typing WHERE AM I is NOT a mistake!!! Why can’t my 3 GHz Core 2 Quad, with hundreds of times as much processing power as the DECSystem-20 mainframe I used to create Trinity, understand and ACT on a simple sentence like WHERE AM I? It’s a scandal, pure and simple.”

“The standards for natural language parsing in IF are laughably, pathetically low. I find it a supreme irony that Inform 7, the most popular IF dev system, is capable of being programmed (in a limited way) using natural language … to create games that are utterly incapable of doing so!”

Moriarty turns his flame off.

zarf brushes ash off his smoking jacket

Moriarty says, “These are the two main challenges for the IF community, as I see it. First, to design a user interface (or better, a whole suite of interfaces) that take advantage of the extremely high resolution and color depth of modern displays to create text games that are BEAUTIFUL … that is, attractive to look at and easy to interact with. The utilitarian thing is NOT working.

“And second, to get up to speed with the enormous advances that have been made in natural language processing, and harness the raw speed and massive storage of modern PCs to ACTUALLY fulfill the promise of typing in real English sentences”.”

Moriarty says, “I should leave time for some questions. Anyone?”

Taladan raises a hand with more of a comment than a question.
Taladan says, “What you’re talking about is a paradigm shift in the way text based games are approached. I think this is awesome, however, we’re talking about a group of people (currently) who have projects (like TinyFugue) thatare developed by one or a select few people that tend to show no interest in ‘keeping up with the jonses’ as far as Graphical games are concerned. Unicode is one example, the language parser is another…how do you see your vision for the future of TBG’s meshing together with the established models of text based gaming out there? Or are we talking about a need for a generational shift in thought instead of personal shifts?”

Moriarty says, “I don’t think this community should be trying to ‘keep up with the Joneses’. You can’t … and you don’t have to. As simple a technology as Adobe Flash is more than capable of creating beautiful animated text displays with full kerning and antialiasing, on every significant platform (except iPhone).

“What’s needed, I think, is an ORGANIZED effort to unite all the people making different clients. Failing that, a single visionary could do the trick. This is NOT that difficult, technically. It’s more of an AESTHETIC challenge now.”

Nämmyung says, “Regarding the parser, what would some example of syntactic complexity that games should be going for?”

Moriarty says, “What you use is important. The development tools must be able to express the new aesthetics. And that is really the issue. The current dev systems are only able to express a limited range of nostalgic aesthetic choices.”

Taladan nods, “As an addendum to my question, if I may: I see what you mean, but too, you do have a lot of folks who are invested in the clients/scripts/whatnot that they use…that personal investment in useage lends to them perpetuating the use of that client because, well, that’s what they know, no?

Moriarty says, “Certainly. Plenty of chickens and eggs to go around here. :)”

Moriarty says, “Anyway, it’s past my hour. Thank you all for listening. Hope I didn’t offend anyone. Despite my remarks, there is some truly heroic work going on in the IF field. I simply fear that much of it will be for naught if beauty and power aren’t added to the equation soon.”

Moriarty says, “Thank you all very much.”

AnneLions says, “I’d like to note that I’ve been hearing the mantra ‘text-based gaming must change’ for the past 10 years (how long I’ve been on MU*). In that time, I’ve yet to see any really decent suggestions/ideas on exactly how to go about that. I’ve seen Pueblo, which was a nice idea but seemed to have been implemented poorly. I’ve seen FANSI, which is pretty, but doesn’t work across all platforms (yet). Unicode would be nice, yes.. But what else is there to change about text-based gaming? If you start to get away from the current (telnet) technology, you start to lose the ‘text based’ aspect. You’ll also lose a subset of players; the blind. I’ve met a significant number of blind people who use text-based games as essentially the only games they _can_ play. This counts MU* more than IF, which is a different ballgame altogether. On a related note, a book of today is would be recognizable and is essentially the same as a book printed 400 years ago. Parser aside, text is text, so I’m not sure what else COULD be changed. In other words, after hearing this sort of thing for a decade, I want to see a mockup that /works/ and doesn’t alienate the current playerbase.”

Moriarty says, “AnneLions, you are exactly correct. What is needed is an EXAMPLE that makes it obvious what COULD be done. One really compelling example will do more than any number of years of griping.”

jaybird says, “Yeah… I was just going to say, as a blind person, text-based games are really valuable to me. Visual flash is nice, as long as it doesn’t get in the way of blind people being able to play.”

Jota says, “Are you presently working on such an example?”

Moriarty says, “Jaybord, there’s more than enough power available to fully accomodate the needs of both sighted and non-sighted people. The truth is, NEITHER group is being particularly well served by current technology or aesthetics. Jota, the answer is: Try and stop me. :)”

Emily says, “There are some fairly strong reasons not to give the player full natural language (and not to pretend you are doing so: ‘just type English sentences’ is effectively a lie). I think Facade is a decent example of why not to do that. Leaving the lid off what you can express to the computer means it’s much harder for the player to understand his range of effective actions in the game world.

“It’s easy enough to make WHERE AM I produce a room description(and in fact various games do do that). It’s much harder to cope with the expectations raised that way”

Nämmyung says, “Well, the underlying point seems to be that games need to be more ambitious / capable in handling a wider range of player interactions, rather than just dropping in a fancier parser while keeping the same restricted set of choices in what they can do.”

Moriarty says, “Obviously, if we make the parser understand more, we will need to find a way to accomodate the potential content explosion. But raw power can help us here, too. Look at chatbots, for instance. Some marvelous work being done there … none of which has made its way into IF, as far as I can see.”

Emily says, “People have experimented with chatbot assimilation to IF”

jaybird says, “Also, I don’t really see the Z-machine going anywhere soon, limited though it is. It’s probably the most widely available IF VM.”

AnneLions notes that chatbots can be /very/ tedious to program.

Emily says, “The problem is that they don’t really deal with the problem of agency” If the player can type anything and everything he wants and the results are handled mysteriously, it’s much harder to understand what effect he’s having”

Moriarty says, “I’m not suggesting that chatbots are the solution, of course. But they ARE an example of using raw processing power to solve a problem that would have been impossible to address in 1980-class PCs.”

Emily says, “Oh, sure: but I think it’s naive to invoke raw power and assume that that’s going to solve what’s essentially a game design problem as well as
a programming problem. Facade had buckets of power, it accepts all input, and it’s deeply unsatisfying because it’s not clear how you’re steering the thing.”

Nämmyung says, “Galatea, on the other hand, did a pretty darn good job at handling unexpected / ‘lateral’ player interaction using z-code.”

Emily says, “(many people would disagree about that, but thanks.)”

Moriarty says, “Emily: I violently agree. Simply increasing the range of inputs that can be understood does not solve the MUCH bigger problem of doing something meaningful and entertaining with those inputs.”

Emily says, “I see a limited application for what you’re asking for, in the sense that broadening acceptance of basic questions, etc., does seem to improve player experience, and people like Aaron Reed have done some tests with novice audiences on this and produced better libraries for input handling.

“Even beyond the massive programming problem of “how do I handle all these inputs” however, there’s the game design problem of how the player should know what to type. ‘Type anything’ is a TERRIBLE kind of game guidance. It’s confusing and boring. The parser feedback informs the player about what’s possible and reasonable in the world. And sometimes, admittedly, that’s done badly, but I think it is in fact a needed aspect of any text-input game, not a flaw.”

Moriarty says, “Let me suggest that actually LIMITING the player’s inputs in some novel way might be a fruitful avenue? In playing some recent works, especially those by Aaron, I sometimes felt I was using the wrong technology — they felt more like hypertext novels.”

Emily says, “Now I’m confused about what you’re advocating.”

Moriarty says, “Emily: We agree again. The problem, for the NON-SAVVY player, is that the input line itself STRONGLY SUGGESTS that anything can be typed, leading to frustration and disillusionment when the fraud is discovered. Is there some other technique that would not feel too limiting?”

zarf says, “Limiting as in not using a text parser, or only accepting one-word inputs, or only accepting words highlighted in blue… ?”

Emily says, “My approach lately has been to try to teach the command possibilities more explicitly.”

Moriarty says, “Zarf: Those are old techniques. Is there something BETTER than all this power we have makes possible?”

zarf says, “I don’t know.”

Jota says, “What do you feel would be better?”

zarf says, “But my basic view is that the old-fashioned IF input line *is* a limiting input mode, and its flaw is not how it works, but making the limitations apparent to the player in a smooth and playable way. If that’s what you’re saying, then the lack of power *isn’t* the problem, because the limits are a feature.”

Gunther says, “hmm, auto-complete of commands?”

tramp says, “auto-complete of commands would be interesting to try, but I already hate that on my cellphone.”

Emily says, “ADRIFT does auto-complete and it generally drives me up a wall.”

Gunther says, “The problem is, if you want to make the parser *seem* to accept anything, you can just program it to reply to “frooble snooblewitz” with “You frooble the snooblewitz!” etc., but that’s not meaningful, and in fact makes it *worse* to find out what *does* do something”

Taladan says, “Unless you’re talking about something like bash’s tab-completion.”

Nämmyung says, “a part of it also probably depends on the writing. If the player can get a sense of clear choices on what he should be doing at a given point in the game, then the parser probably becomes less of an issue.”

Taladan says, “If you let the player have a choice on whether to complete or not, I could live with that.”

Gunther says, “Taladan, I mean it as a help for novices, to see what kinds of commands are available. So, strictly optional, yes.”

Moriarty says, “Zarf: Here I must complain. We ALREADY KNOW that players can be ‘trained’ to speak in 1980-era Infocomese using a text input field. I’m just afraid that the community has settled into a false sense that this is sufficient.”

zarf says, “I think we have some awareness of the problem. 🙂

“(Managing community effort is a different problem and I think it would be distracting to get into it here)”

Moriarty says, “The tools available are confining us horribly. Have you seen some of the beautiful animated poems that people are making with Flash? I simply think the community needs access to much broader range of aesthetic possibilities.”

tramp says, “we want the interface to not be an impediment to the potential fullness of the experience.”

Gunther says, “Anyone can write text. Not anyone can make graphics, music, and animation.”

Emily says, “I think there are some ways to add to player expressiveness, but they’re a bit orthogonal to what you’ve been saying so far. One is to add better handling for non-physical commands (such as conversation); another is to do better with sort of broad commands that entail many actions, like CHECK INTO HOTEL or similar. And those things are in fact a lot easier to handle now than they were a few years ago, I would say (though obviously I have a lot of bias in the game)”

Gunther says, “IF also isn’t an animated poem, because you can interact with it in a hopefully meaningful fashion.”

Jota says, “Wouldn’t something that took its inspiration from Flash animated poems be a completely different field? As in, wouldn’t this not really be the thing that current IF authors are interested in authoring?”

Moriarty says, “AnneLions: I teach my students with Flash. Believe me, I am depressingly aware of its limitations. :)”

Moriarty says, “Jota: Well, if authors are content simply writing games for each other, which are never seen by a larger audience, they can simply go on doing what they’re doing. But I recall a pleasant time when people could actually GET PAID to write text games. I would love for that time to return.”

zarf says, “Man, I never thought about that!”

Moriarty says, “Anyway, I seemed to have stirred the pot a bit. This, as a Professor, warms my cockles. :). Thank you all again. And please remember, it’s not my intent to offend anyone, or denegrate any of the wonderful (and in some cases AMAZING) work that has been done in the past decades to keep text games alive.”

Emily says, “I’m not exactly offended, but I think you’re underestimating the amount of discussion there has already been on a lot of these topics, and in particular on what the parser is useful for, and how. The current state of the discipline isn’t the result of complacency, exactly.

“On the presentation side, my own experience suggests that a moderately attractive and reliable browser-based interpreter does more than anything else to attract new players, and one of the things we’re working on with I7 is a way to publish with a playable website with attractive stylesheets”

Moriarty says, “Emily: I’m a daily lurker on RAIF, and I’ve followed it all with keen interest. Your contributions in particular, which are often in the AMAZING category I noted earlier. :)”

Nämmyung says, “Regarding many of these issues, let me just pose another question: how much difference would it make if we introduced more player interaction in the form of more players? Is there room in IF for side-by-side reading / acting?”

Zebranky says, “Co-op, effectively, Nammyung?”

Emily says, “There’s Floyd, which allows people to play a game with one PC but talk about it as they go; there’s also Guncho, which is designed for multiplayer IF. But that hasn’t seen a lot of development yet. Guncho is a bit like a mud but programmable in Inform 7”

Moriarty says, “Nammyung: Adding more players is the heavy magic, of course. Here’s another irony: Most of the top roleplaying sims in Second Life allow ONLY text chat. Voice chat is disabled. :)”

Nämmyung nods, “yeah, most MMOs are really just text games in disguise, but rather than the hack+slash we get in many muds, what other kinds of gameplay could we get? Just co-op roleplaying is not quite the kind of crafted narrative fiction we can get in IF. In my ideal world it could be something that feels like IF for each player, but in whole looks like a boardgame.”

zarf says, “Yeah. True multi-player IF, whether on Guncho or a MUD platform, is barely explored. It’s one of the things I’d like to work on if I had three more lives”

Emily says, “I have a half-finished Guncho project I keep meaning to get back to, but it’s buried under a heap of conversation system work”

Moriarty says, “Did any of you know that there was a project inside Infocom to create a commercial multi-user Zork?”

zarf says, “I don’t think so. I imagine it went less smoothly than Cyan’s effort to make a commercial multi-user Myst. (A disaster that I watched from close-up.)”

Moriarty says, “It died when Activision closed its jaws. But it was actually working in prototype.

“Wow, it’s fun to talk about text games with a group of people that really cares! I wish RAIF had a better signal-to-noise ratio.”

Emily says, “Heh, I was worried I was coming off as defensive. RAIF is pretty bad these days, unfortunately. A lot of the theory discussion has moved off onto blogs and Planet-IF”

Moriarty says, “Emily: Naw. You know what this communitu needs? A real meatspace conference. Zarf, can’t you get MIT to sponsor something? ;)”

Emily says, “There’s been some discussion of having a bunch of IFfers get together at Boston PAX, but I don’t know if that will come off”

Moriarty says, “Hey, a PAX meetup could work. I’ll bring a bottle of Scotch. ;)”

zarf says, “I am totally in for Boston PAX IF activity. Also, may I interject an ad for the Boston monthly IF meetup. It’s a small bunch, but we have ambitions”

Moriarty says, “I kind of need to go, myself. My email is [obfuscated – ed.] by the way, if anyone wants to contact me. I’d be very keen to know if a meetup actually gets scheduled for PAX.”

Zebranky says, “Thanks again for coming!”
AnneLions waves.
Sketch claps, cheers.
zarf says, “Thanks for speaking”
Nämmyung says, “Thanks for the presentation! It was a lot of fun, and it certainly gave much to discuss.”

Moriarty says, “Zarf: I’ll try to come sometime, if my schedule permits. It sounds wonderful.”
tramp applauds.
Emily says, “Thanks”
Chaz applauds! ^^

Moriarty waves to the audience. For a moment, you see him as a sinister old man in an Edwardian longcoat. Then the vision fades, and he is gone.
Moriarty has disconnected.


Advertisements

6 comments so far

  1. Matt on

    An interesting read. A couple comments on the main points from a MUD perspective:

    Parser development:
    This is a non-issue for MUDs, IMO. Emily had the right of it in that log.

    Presentation:
    You don’t need to have animations and graphics to make a MUD more visually appealing, you just need to present the text in the most attractive way.

    Most people these days are used to reading text in a web page and may never have used a terminal screen so I think more MUDs should emulate web fonts, colour schemes and hypertext links. Most MUDs still look like a terminal window and I’m sure that’s off putting to a lot of modern gamers.

    Using MXP in your MUD enables you to have clickable links, full html colours and custom right click context menus. You can also use it to pass data to the client (for stat bars for example) or to specify multiple windows for output display.

    I think MXP on a custom web client is the way to go if you want to jazz up your interface but don’t want to lose compatibility with existing clients, not to mention blind gamers.

  2. georgek on

    Matt, I think you have the issue well in hand for muds. Over the last couple of years I’ve moved almost completely away from using a dedicated mud client and only play muds with browser interfaces, they’re just more enjoyable to play in a casual manner.

  3. Pacian on

    When it comes to the natural language parser stuff: although being able to type ‘now I want to stick the cog in the clock’ may make the game a lot more accesible at the bottom level, is it really a significant leap forward from a game where you’re typing ‘take cog, put cog in clock’?

    I think it isn’t so much the complexity of the input that needs to be ramped up, so much as the nuance of your interactions in the world. (I don’t really understand what Curveship is doing, but it seems like the kind of thing I mean.)

  4. […] the results (Starship Titanic, Façade). Brian Moriarty and I (and various other community members) argued about the need for NLP for interactive fiction a few months ago. Since then my own ideas about it have been shifting a bit, though it’s not the case (as […]

  5. […] pretends to understand all player input, but that’s a natural direction to wonder about; see this old chat with Brian Moriarty, who, I think we can agree, has more of an insider view on the problem than Blow ever […]

  6. […] improving its natural language processing capabilities? Brian Moriarty, former Infocom implementor, sees NLP as a near-necessity for IF to be better. And in the wake of Watson’s victory, others have wondered why IF parsers don’t take […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: