Monthly Archives: December 2017

The Text Adventures of 1991

Coming exactly halfway between the shuttering of Infocom and the release of Graham Nelson’s landmark epic Curses, 1991 was the most exciting year of the little-remembered interstitial between interactive fiction’s commercial era and its supposed rebirth as an endeavor of dedicated hobbyists. The games of 1991 show what a misnomer the word “rebirth” really is in this case; the text adventure never actually went away at all. The tools available to amateur authors were certainly rougher than they would be in years to come, design standards as well less thought-through, but the fact remains that not a single year has gone by since Adventure first took the computing world by storm in 1977 when at least one or two worthy text adventures haven’t been written. In fact, hobbyists did considerably better than that in 1991. Amidst its blizzard of activity, that year yielded the four games I’ll be writing about today: two classics, one enjoyable journeyman, and one heart-breaker which came that close to being one of the finest text adventures ever written. Not bad for a dead form, eh?


It seemed like such a good idea at the time.

As a self-employed computer consultant, working at home was the logical decision: no more long commutes, no expensive office to lease, no boss. Unfortunately you also have no secretary, no janitor and no weekends. Your living room has become a glorified break room and your only human contact is by Electronic Mail.

Thank God for your computer.

It is 3:30pm on Friday the 7th of September, 2001. Everyone else in your time zone is finishing up work and looking forward to a relaxing evening. But you, R.J. Wright, overtired undernourished overeager programmer that you are, have promised to deliver a finished program to a client by 8:00 tomorrow morning.

Time to get to work....

I’ve always been entranced by the personal aspect of so many vintage text adventures of the amateur stripe. In this respect, they stand apart from almost all other games of their era, which preferred to emphasize the science-fictional, the fantastical, the epic. Even when those other games aren’t demanding that we leave the world we know for some strange new one, they almost always prefer the macro to the micro: wars and battles lost and won, the rise and fall of civilizations, the grand sweep of history. But text adventures, by contrast, can give us an intimate view of a single person’s life, whether said life be that of a PhD student at Harvard or a community-theater volunteer in a small California town. To state the case in literary terms, these are the quieter novels of ordinary people that, for some of us at least, become far more interesting than the latest swords-and-sorcery doorstops as we get older.

Cosmoserve, a game written in AGT by a Celtic harpist and sociologist-to-be named Judith Pintar, has the added advantage of providing a window into a world I’ve just spent the last two months on this blog doing my level best to capture: that of the commercial online services of the 1980s and early 1990s, which pioneered so much of what has since become daily life on the Internet. The game’s title is of course playing on that of CompuServe, the most popular of these services for well over a decade, and a service to which Pintar herself was an active subscriber for many years.

Cosmoserve is ostensibly set in 2001, but we can’t award Pintar too many points for her skill at prognostication. She manages to simultaneously underrate and overrate the pace of technological change to come. In her version of 2001, the wide-open World Wide Web never came along to bury the closed ghettos of the commercial online services, Windows never entirely replaced MS-DOS, and Intel never abandoned their old “x86” nomenclature for their microprocessors. And yet, at the same time that Pintar’s fictional universe was progressing more slowly than ours in all these respects, full-on online virtual realities — the sort of thing that’s just starting to become imaginable for us in 2017 — had already become a thing there by 2001.

But then, prognostication isn’t the point of Cosmoserve. Rather than an extrapolation about computing’s future, what you’re actually getting here is a gentle satire of the computing present which Pintar knew as she was writing the game. If you were already a computer freak in 1991, you’ll find yourself chuckling at things that are barely remembered today but were a major feature of the landscape of those times. For instance, do you remember the way that Intel, thanks no doubt to some marketing genius of an MBA inside the company, used to release crippled “SX” versions of their latest chips — versions that in some cases actually performed worse than the chips of the previous generation? I barely did myself, until Pintar reminded me:

This is the newly-released Orfland 786SX. Most of the advanced features of the revolutionary 786 chip were factory-disabled for the SX model: it hasn't got enough memory to run a graphical interface and it chugs along at about 12Mhz, but hey, its still a 786!

The plot of Cosmoserve is a classic shaggy-dog story in text-adventure form, the same approach that would be used to more famous effect by Curses two years later. Playing the role of a harried free-lance computer consultant, you need to get a patch for Turbo Pascal — another blast from computing’s past; the Borland product was by far the most popular development tool in the world in 1991 — in order to complete an assignment for an important client. You should be able to get the patch on Compu… err, Cosmoserve. It’s when you fire up your computer to go online and fetch it that the game, after having started out as a slice of life set in your own house, begins to show its real cards.

Most of Cosmoserve plays as a simulation of that whizz-bang Orfland 786SX computer of yours. First you’ll have to navigate the DOS prompt to get yourself online; in the some-things-never-change department, remembering your password will pose a particular problem on that front. Then, once you do manage to get online, the simulation extends yet one level deeper, allowing you to roam the Cosmoserve service, visiting forums, chat rooms, file libraries, email — all the things I’ve spent so many recent articles describing — along with a few futuristic touches, like the virtual-reality area, presented in recognition of the fact that we’re allegedly in 2001. It eventually emerges that getting the Turbo Pascal patch and finishing your assignment will first require you to stop a computer virus that threatens to take over the world. For those who aren’t aware: yes, viruses were already a problem in 1991. I really could go on forever about this game’s palimpsest of the familiar and the obscure, how it constantly signals all the ways things have changed in computing and all the way they’ve remained the same.

As a purely technical achievement, Cosmoserve is remarkable, especially considering that Pintar wasn’t an experienced programmer. AGT was by the standards of text-adventure authoring systems to come a very primitive tool indeed, riddled with assumptions about the sorts of games it would be used to create that can be almost impossible to completely override. And yet Cosmoserve manages to push AGT farther out of its comfort zone than any other game I’ve ever seen. If its simulations of DOS, of a terminal program, of CompuServe/Cosmoserve itself — even of a text adventure within this text adventure which you can play from the DOS prompt — aren’t always perfect, they’re far better than they have any right to be.

From the standpoint of the modern player especially, Cosmoserve does have some drawbacks. It’s never an unfair game according to its own old-school lights, but it is a demanding one. If you’ve never used MS-DOS, or have forgotten everything you once knew, you’ll likely have to consult a reference manual in order to get anywhere at all. And you’ll certainly have to pay careful attention and make some notes if you hope to solve this one.

More controversially, Cosmoserve plays on a clock. Timing is tight, you have a lot to do, events happen online at specific times… meaning, yes, this is one of those try-and-try-again games which require you to make a series of losing reconnaissance runs to get the lay of the land before you put everything together for your victory dash. This design approach is absolute anathema to some people; if you’re one of those people, nothing I can say will persuade you otherwise. The good news, though, is that in this case at least I don’t really have to.

Judith Pintar revisited Cosmoserve in 1997, adding some further polish to the experience and, most importantly, greatly easing the time pressure. Which version you choose to play must be a reflection of your own preferences as a player. Personally, having been raised on the Infocom mysteries, I don’t mind the try-and-try-again approach overmuch, if it’s done within reason and if I know what I’m getting into. I thus actually prefer the earlier Cosmoserve, which feels like a purer expression of its designer’s original intent to me. But of course those of you who aren’t as old-school — masochistic? — as me should feel free to go for the later version.

Either way, I think you’ll find the experience worthwhile. Whether considered as a pure gaming challenge or as a cultural artifact of its very specific time and (virtual) place, Cosmoserve has a lot to offer; taken on either terms or both, the wit, humor, and humanity of its author shine through. It was given co-winner status in the 1991 AGT Competition, alongside a more traditional text adventure called The Multi-Dimensional Thief. The latter game, a confusing  mashup of Guild of Thieves and The Wizard of Oz that delights in insulting its player when it isn’t dead-ending her, hasn’t aged very well. Cosmoserve, on the other hand, has only become more essential as the online life it chronicles has faded into oblivion and its time-capsule qualities have come to the fore.

Although either the original or the updated version of Cosmoserve can be played most easily on modern computers using the AGT interpreter AGiliTy, you’ll lose much of the atmosphere provided by occasional sound effects, not to mention MS-DOS’s familiar old green text on a black background. I therefore recommend playing it under the original AGT interpreter, in DOS, to get the full effect. To make that as easy as possible for you, I provide versions of Cosmoserve and Cosmoserve 97 — take your pick — ready to run in the DOSBox emulator. Whether you have a Windows, MacOS, or Linux machine, just install the version of DOSBox for your platform and follow the instructions included in the zip file to get the game going.

The Dungeon of Dunjin

You are in a dark, mysterious and confusing forest. Tall fir-trees form a dark wall around you. A cold wind is blowing from the mountains, and in the far distance you can hear wolves howling. Faint trails lead east and south. To the north the forest seems to continue forever, and to the west the vegetation is so dense that it would be impossible to go in that direction.

If you’re anything like me, you may prefer the idea to the reality of the sprawling text adventures that ran on the big institutional computers of the 1970s. Still some of the largest works ever created in the medium of text and parser, games such as the original Zork and Acheton offer immense worlds of hundreds of locations and almost as many puzzles — worlds to get lost inside for weeks or months. They seem absolutely amazing at first. When you start to play them a little more, though, you come to realize that you just can’t trust these games. Standards of good and bad design simply didn’t exist at the time they were being made, meaning that they tend to be riddled with as many terrible puzzles as brilliant ones.

The Dungeon of Dunjin, written by a Swede named Magnus Olsson over the course of about five years, answers the question of what Zork might have been like if it hadn’t, as Robb Sherwin once so memorably put it, hated its player. The setup is as old-school as it gets: you, the nameless faceless adventurer, have arrived near the entrance to the titular dungeon with treasure on your mind. As you play, another plot line does begin to emerge, but it never feels all that compelling. At bottom, The Dungeon of Dunjin is best accepted as a game about looting a landscape and dropping your spoils in a repository for points — a concept that was beginning to feel a little retro already by the time Infocom left Zork behind in 1983. Olsson’s 1991 backward glance comes complete with a sprawling geography of some 180 rooms, filled with locations that in typically old-school fashion often fail to connect with one another in the expected ways; going south and then going north, in other words, isn’t guaranteed to return you to your starting position. (In light of this, you’ll wind up happy that the game engine doesn’t recognize secondary compass directions like northeast.) Needless to say, light sources, trolls, and dragons figure prominently in the puzzles and plot.

But the thing that separates The Dungeon of Dunjin from its legendary forebears is that all the really annoying old-school nonsense is blessedly missing. Olsson has clearly made a conscious, thoroughgoing effort to design a game that the motivated player can actually win, and without being bored to death by petty logistical problems in the process.

The game engine is home-grown, written in Turbo Pascal. (I did tell you it was everywhere in the early 1990s…) It’s nowhere close to the level of even the PDP-10 Zork, possessing only an extremely basic world model and a parser that’s for the most part limited to two-word verb-noun constructions. Many a designer forced to work with such an engine has wound up stretching it past the breaking point, stumbling into the territory of guess-the-verb puzzles and sheer logical incoherence in an attempt to make a more difficult game than the engine can really support. (I would argue that the entire Scott Adams catalog after the fifth or sixth game can be seen as extended proof of this thesis.) But Olsson is too smart to be caught in that trap: he knows how to work within his tools, avoiding puzzles — like those involving intricate mechanical manipulations — which his game engine just can’t handle. There’s enough that it can do, he realizes, to make a perfectly satisfying old-school adventure game.

The most unfortunate aspect of the writing actually comes right up front, in the horrid title. The Dungeon of Dunjin is no literary masterpiece — that’s hardly the point of a game like this one, is it? — but the writing acquits it surprisingly well for that of a non-native English speaker. (Olsson does poke a little fun at the thing many people still think of first when they think of Sweden by making one puzzle revolve around an Abba record.) Like Adventure and Zork before it, the game never takes itself too seriously, freely mixing contemporary culture with high fantasy, placing computer labs practically next door to slavering dragons. Sometimes a sly Zorkian wit peeks through, as when you find a human skull in the dungeon that’s made of plastic and has “Made in Taiwan” printed on the side. The dungeon itself, meanwhile, proves in the end to be a closed-down tourist attraction; shades of the bizarre postmodern endgame of Adventure.

Filled with little homages to its predecessors like these, but perfectly playable if you don’t know a rusty rod with a star on the end from a lonely white house, The Dungeon of Dunjin is one of the better old-school puzzlefests I’ve played in my time, consistently surprising and amusing, consistently challenging — not least as a result of the combinatorial explosion that stems from its considerable size — and yet never insurmountable and only very rarely actively annoying. I enjoyed playing it immensely, and fancy that any of you who are up for a big adventure that will absorb quite some hours of your time and who don’t mind making a map and checking it twice — thankfully, we have Trizbort these days! — may just do so as well. Being a native citizen of MS-DOS, it can only be played through an emulator on modern computers. I’ve therefore prepared a version for you that will make that as easy as possible. Just add DOSBox.

Save Princeton

According to the brochure that the admissions department gave you, Princeton University is one of the last bastions of intellectual pursuit, where students can engage in the quest for learning unencumbered by worldly cares. As far as you can tell, though, the place looks like any one of a thousand clones of Cambridge University that clutter the American academic landscape. Still, you figure there must be something at least vaguely interesting about it, considering the reputation that the place has managed to accumulate over the past 250 or so years.

You decide, therefore, to take an Orange Key Tour. Minutes into it, though, you realize that you have no interest in being shown buildings with cannonball scars from the Revolutionary War. So, as the guide leads you through yet another archway, you break off from the group and wander through a nearby door. You find yourself in an entryway, standing in front of a door labeled with the number 21. When you turn the handle, the door swings open, and you enter, hoping no one will catch you being a voyeur.

As you poke around, the sound of gunfire coming from outside in the courtyard startles you. You dive for cover beneath a desk and remain there, shaking, until the tumult dies down. When you come out, you can sense a tension in the atmosphere. Clearly, something strange has happened.

Written by a Princeton University student named Jacob Weinstein with some assistance from his fellow student Karine Schaefer — “she just helped plot it, and left the geeky stuff to Jacob” — Save Princeton is another entry in a weirdly overstuffed sub-genre of interactive fiction: the collegiate text adventure, a category that includes such earlier classics as The Lurking Horror and A Dudley Dilemma. This game isn’t on the same level as either of those, but it has its charms.

As soon as you start Save Princeton, you’re smacked in the face with how much some things have changed since 1991; a plot involving terrorists taking over a major university would never be treated so flippantly in these times of ours. Here, though, it’s just a mechanism for pushing you to explore the campus and lap up — or, more likely, scratch your head at — the endless in-jokes. While nothing really stands out about the game’s puzzles or construction, there’s nothing notably objectionable either; this is all pretty standard fare, albeit delivered in a pretty user-friendly way, without the inscrutable puzzles, mazes — well, there is a fake maze — or parsing issues that were still typical of most amateur text adventures of this era. Doubtless helping the game’s cause is the fact that it’s written in TADS, a much more powerful and polished system for programming text adventures than AGT, if also a much less popular one in 1991. The writing is actually more ramshackle than the technology or the puzzle design, with the tossed-off feel that was also so typical of early amateur text adventures.

But it isn’t Save Princeton‘s merits as a piece of timeless game design, much less as a piece of writing, that makes me want to cautiously recommend it. It rather comes down once again to that personal quality of so much amateur interactive fiction.

Weinstein fills his slice of life not just with the architecture of Princeton, nor just with the pop-culture detritus of 1991 — “there are posters of such charming items as Laura Palmer’s corpse” — but with himself, along with the friends he has at university. Go to the “Girls’ Common Room” and there’s Lisa, “working on the New York Times crossword puzzle”; there’s Melisande, “buried in an Orson Scott Card novel.” Go to the boys’ room and there’s Eric, “humming the violin part of The Rite of Spring“; there’s Otis, the “fairly accomplished computer programmer” who “won the Mr. Princeton bodybuilding contest his freshman year.” Eventually you’ll also meet Karine — yes, Weinstein’s alleged coauthor — sitting in the romantic glow of a lava lamp, dreaming of Anthony Hopkins of all potential heartthrobs, “making an acidic comment regarding the cultural inferiority of every city in the world except for New York.” Somehow I suspect that Jacob was crushing hard on Karine, to the point of giving her a dubious authoring credit on his game, only to be stuck permanently in the Friend Zone.

Of course, I don’t really know what was going on with any of these young people. Nor have I ever even been on the Princeton campus, meaning that every in-joke is utterly lost on me. And yet — and you can chalk this up to my going all American Graffiti in my middle age if you like — there’s something about this unassuming little game that I find almost unbearably poignant. It so happens that I’m almost the exact same age as the kids we meet in it, and I can’t help but feel a connection with all these entitled little dreamers, so full of grand plans for the future, so convinced that the meaning of life can be revealed to them by the right song, book, or film. Where have their lives taken them? If they were given this game to play today, would they be surprised to meet the people they used to be?

Call me a sentimental fool, but Save Princeton, patently envisioned by its author as just a light-hearted adventure game, kind of puts a lump in my throat. Your own mileage may of course vary, but it’s certainly not a bad little game even if it doesn’t prompt in you the same ruminations about the cycle of life. You can download it from the IF Archive and play it using any of the many freely available TADs interpreters.


You awake from uneasy dreams. Since you're no longer on easy street, maybe that's the way your dreams are going to be from now on. Exactly where you are becomes clear as you sort out the sounds of the river to the east, the rustlings of birds to the north and west, and the sweet scent of sleep-inducing poppies wafting down from the northwest. Apparently, after a day of determined walking about, you burrowed down next to the river and let consciousness drift.

What exactly induced this bout of walking? Well, two nights ago, Count Zero handed you your walking papers and extracted your latchkey to the museum in exchange (little does he know that you keep a spare hidden in the topiary). It's just as well that you were dismissed from the museum--your duties as combination custodian and librarian involved either re-shelving books and dusting off clocks or rewinding timepieces and dusting off books. However, you were onto something. Exactly what is unclear since the pieces of the puzzle seem to disconnect with sleep. You resolve not to sleep until you've recollected and reconnected their jagged edges. You can be just as calculating as the Count. You can even reach beyond the Zero . . .

I had one of the most magical gaming experiences of my life with T-Zero.

About fifteen years ago, I was working the graveyard shift at a computer-services firm in Dallas. From 7 PM to 7 AM, three or four nights per week, I’d sit in a nearly deserted data center babysitting the servers and mainframes, just in case anything should happen. Most of the time, it didn’t, meaning I had a lot of free time on my hands. Boredom was a big problem. There were enough curious eyes wandering about the place doing their system modifications and whatnot that playing anything that looked like a game would have been a really bad idea. Text adventures, however, were a different story. I could sit typing away into a window filled with text, looking for all the world like I was hard at work on something vital. Thus I played a lot of text adventures during this period, delving back into a lot of forgotten games — often justifiably forgotten! — from the early issues of SPAG magazine. One of the games I played was T-Zero.

I was playing it one night when a message popped up out of nowhere, apropos of nothing I was actually doing in the game at the time. “It’s a full moon tonight,” it read. “Go outside and take a look.” So, curious whether this already very old game knew what it was talking about, I did.

Well, the game did know what it was talking about. Outside a huge harvest moon hovered low over the warm night. I’d always loved the silence of the predawn hours, when the only sounds you could hear were the omnipresent Texas crickets. Now the peaceful scene, blanketed in the moon’s silvery sheen, seemed to fuse with the peculiar beauty of the game I’d just been playing. I stood there in front of my employer’s antiseptic corporate building for quite some minutes, marveling at the beauty that can visit us at the most banal times. As I turned to go back inside, I knew that I’d never forget this night. And, as this little reminiscence demonstrates, I was right.

T-Zero truly is a magical game in some ways; at its best, it almost attains the same heights as Trinity, my all-time favorite work of interactive fiction. Indeed, comparisons between the two works strike me as unavoidable. T-Zero at the time of its release had the most subtly textured writing that had been seen in a text adventure since Trinity. More than that, though, it resembles and even pays homage to Infocom’s finest hour in many respects: a sundial and a gnomon, to name an obvious if superficial example, figure prominently in T-Zero as well as in Trinity. Less superficially, both games share an abiding obsession with the mystery of time, and both have a smile-through-tears quality, a gentle whimsy laced with melancholia.

T-Zero was written by one Dennis Cunningham, a person about whom I know nothing beyond his description of himself as “a programmer with literary leanings.” He is or was obviously very talented in both fields. For someone like me who loves words, T-Zero is a source of constant delight. As an example of its love of clever wordplay, consider that you begin it by waking up in the location known as “River Bed” — or consider the “buxom bell” you’ll soon be ringing. Unlike so many self-consciously “literary” interactive works, which tend to get buried under the weight of their own aspirations, T-Zero‘s writing dazzles without ever seeming to try to do so; Cunningham’s writerly touch is light where the others are heavy. I can perhaps best convey his game’s atmosphere by borrowing a line from one of his room descriptions: “Either your vision is becoming near-sighted or this scene has all the pointillist charm of a Monet painting.” Like an Impressionist painter, Cunningham is more interested in the interplay of light and shadow than he is with concrete forms. Maybe that explains why the moonlight affected me so on that one magical night.

I hesitate to trample over the delicate poetry of T-Zero too much more with my leaden reviewer’s prose, but will note that it takes place in the slightly surreal landscape which surrounds a strange museum where you until recently worked as a low-grade custodian and librarian. You will eventually learn that the time is out of joint, and you will have to learn to travel through time to visit the same locations in other millennia, learning of the other inhabitants who dwelt and will dwell here. These inhabitants are not human; nor is it ever entirely clear whether you yourself are. Again, T-Zero isn’t concerned with the concrete. It’s a dream and a meditation, and it’s all the better for it.

T-Zero has a spirit of unabashed intellectualism to it — a complete disinterest in talking down to its player — which looks forward to Graham Nelson’s Curses. Cunningham peppers his game with allusions: to Miguel de Cervantes and his deluded knight, to Edgar Allan Poe and his bells ringing in the night, to Douglas Hofstadter and his Eternal Golden Braid, to the Beatles and their Walrus. This sort of thing can veer into rank pretension in a hurry. But, again like Nelson, Cunningham’s erudition reads as intriguing rather than off-putting, sending you scurrying off to Wikipedia to learn more about the references you don’t entirely get.

With so much going on at the literary level, it may seem almost belittling to focus on the technology that underpins the game, but I’d actually be doing its author a disservice not to mention it. Like Magnus Olsson, Dennis Cunningham chose to write his game from scratch rather than use a text-adventure authoring system like AGT or TADS. And here I have to break out the superlatives yet again: the engine he created is quite simply the best bespoke text-adventure engine I’ve seen. Ever.

Cunningham doesn’t just meet the Infocom standard that was still the aspirational ideal among amateur text-adventure makers in 1991, he actually exceeds it in a number of respects. The parser handles even the most complicated constructions with aplomb, and the game is rife with little conveniences seldom seen during its era: things like undo, like a generous command-history buffer, like a menu-based restore command that doesn’t expect you to remember the name of every save file you create. The world model is complex and coherent, and the addition of carefully chosen shades of color, rather than just looking gaudy as color so often tends to do in an all-text game, adds to the rich atmosphere.

Just look at this! T-Zero offers a menu for disambiguation as one of the conveniences it’s absolutely rife with.

But now, having praised T-Zero to the skies, I have to tell you about its one tremendous flaw: this game is just way, way too hard. A goodly chunk of the puzzles involve wordplay of the Nord and Bert variety, the sort of thing that delights some players and drives others — especially players who don’t have English as a first language — absolutely crazy. This in itself may thus be enough reason for some of you to reject a game, but we’re just getting started with the litany of barriers to solving it. T-Zero muddies the waters further than Nord and Bert did in that it doesn’t have discrete sections devoted to discrete kinds of wordplay; you never know, in other words, whether it’s looking for an idiom or an anagram or an allusion. Or, for that matter, whether it’s looking for something else entirely: there are also plenty of traditional puzzles here, grounded in real-world — or at least text-adventure — physics. And then we have to throw onto the pile the fact that this is a big game with a lot of locations to explore, and over several time periods at that. Because the descriptions of these intricate landscapes are drawn in such loving detail, it’s really, really hard to know for sure which locations contain puzzles waiting to be solved and which just exist for you to drink in on their own terms.

Not helping the situation is a tendency for the parser, so flexible in most ordinary tasks, to suddenly become needlessly persnickety in some specific situations, with failure messages that can be not just unhelpful but actively misleading. For instance, at one point you need to tear the flyleaf out of a book. You need to type it exactly like that: “tear flyleaf out of book.” If you try to “pull flyleaf out of book,” you’re told that “your pull is next to nothing when it comes to the flyleaf.” Far worse, if you just type, “tear flyleaf,” you’re told that there’s “no reason to play the vandal”; if you’re foolish enough to take the game at its word here, you’ll never solve it. There aren’t heaps of situations like this one, but there are more than enough to ruin an otherwise brilliant game for its player even absent the other questionable design choices.

That said, it must also be admitted that there is a partial solution to all these problems built right into the game. Among its other technical wonders, T-Zero includes a full-fledged adaptive hint system that keeps track of your progress and doles out context-specific hints for each location — the first such system I’m aware of in the history of interactive fiction. It breaks my heart, but I have to recommend to any of you who choose to play this game that you use it liberally, typing “hint” as a matter of course in each location you visit. Sometimes doing so gives away the full answer to the puzzle; sometimes it at least leaves a little for you to work out on your own. The former in particular is far from ideal, but what else can you do if you’d prefer not to beat your head for hours and hours against this brick wall of a game? The shame, of course, is that there are some very good puzzles here which you won’t be able to enjoy thanks to the bad ones. Ah, well… at least T-Zero‘s wonderful version of the maze-that-isn’t-really-a-standard-maze, almost as venerable a text-adventure tradition by this point as mazes of the old drop-and-plot variety, isn’t entirely spoiled by the hint system.

Having to recommend that you play T-Zero in this way really does pain me, not least in that it destroys all the critical goodwill I have toward every other aspect of the game. As you regular readers know, I’m deeply skeptical of the idea of the “great, as long as you have a walkthrough” species of adventure game. Adventure games are interactive works, and when their interactivity fails them it’s hard for me to see why one should bother with them. As I once put it, “an adventure game that cannot be solved unaided, or for that matter that can be solved only through sheer doggedness and refusal to give in to tedium, is a bad game.”

But I will say now that this particular bad game comes closer than any other to making me recommend that you go ahead and play it anyway using the hints, just to experience the prose and the beautiful environment it evokes. In the end, you’ll have to decide for yourself whether this failure that by all rights should have been numbered among the all-time greats is worth your time. Once again, you can download an almost-ready-to-play version from this site. The only other thing you need is DOSBox.


Posted by on December 29, 2017 in Digital Antiquaria, Interactive Fiction


Tags: , , , , , , , ,

Games on the Net Before the Web, Part 3: The Persistent Multiplayer CRPG

Black Dragon, CompuServe’s first CRPG, was popular enough that the service sold tee-shirts.

The first CRPG to go online with CompuServe was also one of the first entirely original games to appear on the service, following the initial glut of institutional-computing refugees. Black Dragon, written by a programmer of telephone switching systems named Bob Maples, was at bottom a simplified version of Wizardry — not a hugely surprising state of affairs, given that it made its debut in 1981, at the height of the Wizardry craze. The player created a character — just one, not a full party as in Wizardry — and then began a series of expeditions into the game’s ten-level labyrinth, fighting monsters, collecting equipment and experience, and hopefully penetrating a little deeper with each outing. Only the character’s immediate surroundings were described on the scrolling, text-only display, so careful mapping became every bit as critical as it was in Wizardry. The ultimate goal, guaranteed to consume many hours — not to mention a small fortune in connection charges — was to kill Asmodeus, the black dragon of the title, who lurked down on the tenth level. Any player who managed to accomplish that feat and escape back to the surface was rewarded by seeing her name along with her character’s immortalized on the game’s public wall of fame.

Those bragging rights aside, Black Dragon had no multiplayer aspect at all, which might lead one to ask why its players didn’t just pick up a copy of Wizardry instead; doing so would certainly have been cheaper in the long run. But the fact is that not every CompuServe subscriber’s computer could run Wizardry in those early days. Certainly Black Dragon proved quite popular as the first CompuServe game of its kind. Sadly lost to history now, it has been described by some of its old players as far more cleverly designed than its bare-bones presentation and its willingness to unabashedly ride Wizardry‘s coattails might lead one to believe.

Black Dragon‘s success told Bill Louden, the “games guy” at CompuServe, that subscribers had a taste for this sort of experience. In 1982 a second, somewhat more sophisticated single-player CRPG went up. Known as Dungeons of Kesmai, it was, as the name would imply, another work of the indefatigable John Taylor and Kelton Flinn — i.e., Kesmai, the programmers also responsible for CompuServe’s MegaWars III and, a bit later, for GEnie’s Air Warrior. Like so many of CompuServe’s staple games, both Black Dragon and Dungeons of Kesmai would remain on the service for an absurdly long time, until well into the 1990s.

But more ambitious games as well would come down the pipe well before then. A few years later after these first single-player online CRPGs debuted, CompuServe made the leap to multiplayer virtual worlds. As we’ve already seen in my previous article, MUD washed up from British shores in the spring of 1986 under the name of British Legends, bringing with it the idea of the multiplayer text adventure as virtual world. Yet even before that happened, in December of 1985, the CRPG genre had already made the same leap thanks to still another creation from Kesmai: Island of Kesmai.

Taylor and Flinn had originally hoped to make Dungeons of Kesmai something akin to the game which Island would later become, but that project had been cut back to a single-player game when Bill Louden deemed it simply too ambitious for such an early effort. Undaunted, Kesmai treated Dungeons as something of a prototype for their real vision for a multiplayer CRPG and just kept plugging away. They never saw nor heard of MUD when developing the more advanced game, meaning that said game’s innovations, which actually hew much closer than MUD to the massively-multiplayer games to come, were all its own.

Island of Kesmai demonstrated just how far games’ presentation had come on CompuServe in the four years of creeping advancement that had followed Black Dragon. While it was still limited to text and crude character graphics, the latest terminal protocols did allow it to make use of color, and to divide the screen into quadrants dedicated to different purposes: a status “window” showing the state of the player’s character, a pseudo-graphical overhead view of the character’s surroundings, a text area for descriptions of the environment, a command line for the player to issue orders. Island of Kesmai looked like a roguelike, a genre of hardcore tactical CRPG that was a bigger favorite with hackers than with commercial game developers. This roguelike, however, was a multiplayer game set in a persistent world, and that changed everything.

Island of Kesmai used the ASCII graphics typical of roguelikes. Here “>” represents the player’s character; “A” is a monster; “B” is another player’s character; “@@” is a spider web; and “$” is a treasure or other item. The brackets are walls, while “–” represents a closed door and “/” an open one.

As with British Legends, up to 100 players could share Island of Kesmai‘s persistent world at the same time. Yet Kesmai’s creation was a far more coherent, far more designed experience than the cheerful insanity that was life on MUD. Players chose a class for their characters, along with an alignment, a gender, and even a land of origin. As befitted the game’s grounding in CRPG rather than text-adventure tradition, combat was a far more elaborate and tactical affair than in MUD. You had to reckon with the position of your character and your opponents; had to worry about initiative and fatigue; could be stunned or poisoned or even fumble your weapon. The magic system, too, was far more advanced and subtle than MUD‘s handful of ad-hoc spells that had often been added as much for comedic value as anything else.

The Island that gave the game its name was divided into five regions, comprising in total some 62,000 discrete locations, over which roamed some 2500 creatures in addition to one’s fellow players. The game was consciously designed to support differing levels of player engagement. “A person can play casually or seriously,” said Ben Shih, a “scenario designer” hired by Kesmai to continue evolving the game. “He or she can relax and take out frustrations on a few goblins or unwind by joining other players in hunting bear and griffin. But to become a superstar, a ‘mega-character,’ takes time.”

Ben Shih, John Taylor, and Kelton Flinn of Kesmai.

Scenario designers like Shih added content on a regular basis to keep the game fresh even for veteran players, sometimes giving a unique artifact to the first player to complete a new quest. Kelton Flinn was still excited about adding new stuff to the game three years after it had first gone online:

We don’t feel we’re designing games. We’re designing simulations. We create a world and then we let the players roam around in it. Of course, we’re always adding to our view of the world, fiddling with things all the time, creating new treasures, making things work better. I suppose at some point you have to call a halt and say, “Let’s see if we want to make a clean break and try something bigger.” But we haven’t reached that stage yet.

For all the changes the game went through, the ultimate achievement in Island of Kesmai remained always to kill the dragon, the toughest monster in the game. Players who did so were rewarded with everlasting fame as part of the true elite. As for the dragon: he of course re-spawned in a few days’ time, ready to serve as fodder for the next champion.

Those who hoped to do well were almost forced to buy the 181-page manual for the game, available for the low, low price of $16.50 directly from CompuServe. A rather stunning percentage of the elements described therein would still ring true to any World of Warcraft player of today. There was, for instance, a questing system, a ladder of challenges offering ever greater rewards in return for surviving ever greater dangers. Even those looking for an equivalent to the endless stream of World of Warcraft expansions can find it with Island of Kesmai. In 1988, Kesmai opened up the new lands of Torii and Annwn, filled with “more powerful weapons, tougher monsters, and a variety of treasures.” Advanced players were allowed to travel there only after their characters had hit the old Island’s level cap, and weren’t allowed to return again after they passed through the magic portal, lest they wreak havoc among the less powerful monsters and characters they once left behind.

While play on the Island was much more structured than it was in The Land of MUD, it was still the other players who really made the game what it was. Taylor and Flinn went into the project understanding that, and even anticipating to an extraordinary degree the shape of virtual societies to come. “We fully expect that a political system will evolve,” said Taylor upon the game’s launch, “and someone may even try to proclaim himself King of Kesmai.” Much of the design was put in place to emphasize the social aspect of the game. For example, a conference room was provided for strategizing and conspiring, and many quests were deliberately designed to require the cooperation of several characters. The verbiage adopted by players in relation to the quest system still rings true to modern ears. For example, a verb was coined for those loners determined to undertake quests on their own: to “solo.”

Although player-versus-player combat was allowed, it was restricted to specific areas of the Island; an attempt to attack another character in a “civilized” area, such as the town where new players began their adventures, would be met by the Sheriff, an invincible non-player character guaranteed to grind the brawniest hero into dust. Alignment also played a role: a karma meter kept track of players’ actions. Actions like assault or theft would gradually turn a good character neutral, then finally evil. The last alignment was highly undesirable from many perspectives, not least in that it would prevent you from entering the town, with its shops, bars, and trainers.

And there were still other mechanisms for discouraging the veterans from tormenting the newbies in the way so many MUD players so enjoyed. Players were urged to report griefers who preyed excessively upon newbies, even if they only did so in the dungeons and other “uncivilized” areas where player-versus-player combat was technically allowed. If enough people lodged complaints against them, the griefers might find themselves visited by the wrath of the “ghods of Kesmai,” the game’s administrators — the alternate spelling was used so as not to offend the religious — who might take away experience points, steal back their precious magic items, or just smite them dead as punishment. The game thus tended to foster a less cutthroat, more congenial atmosphere than MUD, with most players preferring to band together against the many monsters rather than fight with one another.

A journalist from the magazine Compute’s Gazette shared this tale of his own almost unbelievably positive first encounter with another player in the game:

I desperately wish I could afford to buy a few bottles of balm sold by the vendor here in the nave, but at 16 gold pieces each they are far above my limited budget. Another player walks in from the square. “Hello, Cherp!” she says, looking at me. Taking a close look at her, I recognize Lynn, a middle-aged female fighter from my home country of Mnar.

“Howdy to you. Are you headed down into the dungeon? I’ve just arrived and this is my first trip down,” I tell her.

“Ah, I see. Yes, I was headed down, but I don’t think it’s safe for you to hunt where I’ll be going. Do you have any balm yet?” she asks as she stands next to the balm vendor.

“No, I haven’t got the gold to afford it,” I say hesitantly.

“No problem. I have a few extra pieces. Come and get them.”

“Thank you very much,” I say. Lynn drops some gold on the ground, and we wait as the vendor takes the gold and drops the balm bottles for us. I pick up the bottles and add them to my meager possessions.

“I can’t thank you enough for this,” I say. “Is there some way I can repay you? Perhaps we could meet here again later and I could give you some balms in return.”

“No,” she laughs, “I have no need of them. Just remember there are always other players who are just starting out. They may find themselves in the same position you are in now. Try to lend them a hand when you are sufficiently strong.”

At the risk of putting too fine a point on it, I will just note one more time that this attitude stands in marked contrast to the newbie-tormenting that the various incarnations of MUD always seemed to engender. At least one player of Island of Kesmai so distinguished himself through his knowledge of the game and his sense of community spirit that he was hired by Kesmai to design new challenges and serve as a community liaison — a wiz mode of a different and much more lucrative stripe.

But the community spirit of Island of Kesmai at its finest is perhaps best exemplified by Valis, one of the game’s most accomplished players. This online CRPG was actually the first RPG of any stripe he had ever managed to enjoy, despite attending university during the height of the Dungeons & Dragons fad: “I could never get into sitting around eating crackers and cheese doodles and arguing for twelve hours at a time. I can do as much in a half hour in Island of Kesmai as they did in twelve hours.” Valis became the first person to exhaustively map the entire Island, uploading the results to the service’s file libraries for the benefit of all. Further, he put together a series of beginners classes for those new to what could be a very daunting game. CompuServe’s hapless marketers advertised his efforts as an “escort service,” a name which perhaps didn’t convey quite the right impression.

We think we’ve come up with the perfect way of teaching a beginners class. We spend an hour or so in the conference area with a lecture and questions. Then we go on a “field trip” to the Island itself. I lead the beginners onto the Island, where we encounter a few things and look for some treasure. That usually is enough to get them started.

In many respects, the personal stories that emerged from Island of Kesmai will ring very familiar to anyone who’s been reading my recent articles, as they will to anyone familiar with the massively-multiplayer games of today. Carrie Washburn discovered the game in 1986, just after her son was born fourteen weeks premature. During the months the baby spent in intensive care, Island of Kesmai became the “link back to reality” for her and her husband. After spending the day at the hospital, they “would enter a fantasy world in order to forget the real one. The online friends that we met there helped pull us through.” Of course, the escape wasn’t without cost: Washburn’s monthly CompuServe bill routinely topped $500, and once hit $2000. Later she divorced her husband and took to prancing around the Island as the uninhibited Lynn De’Leslie — “more of a slut, really” — until she met her second husband there. Her sentiments about it all echoed those expressed by the CB Simulator fraternity on another part of CompuServe: “One of the great things about meeting people online is that you get to really known them. The entire relationship is built on talking.” (Appropriately enough for a talker, Washburn went on to find employment as the administrator of the Multiplayer Games Roundtable on GEnie.)

Kelton Flinn once called Island of Kesmai “about as complicated as a game can be on a commercial system.” Yet it deserves to be remembered for the thought that went into it even more than for its complexity. Almost every issue that designers of the massively-multiplayer games of today deal with was anticipated and addressed by Kesmai — sometimes imperfectly, yes, but then many of the design questions which swirl around the format have arguably still not been met with perfect answers even today. Incredibly, Island of Kesmai went online in December of 1985 with almost all of its checks and balances already in place, so thoroughly had its designers thought about what they were creating and where it would lead. To use Richard Bartle’s terminology from my previous article, Island of Kesmai was a “product” rather than a “program,” and it was all the better for it. While MUD strikes me as a pioneering work with an awful lot of off-putting aspects, such that I probably wouldn’t have lasted five minutes if I’d stumbled into it as a player, Island of Kesmai still sounds like it must have been fantastic to play.


One big name in the field of single-player graphical CRPGs took note of what was going on on The Island quite early. In 1987, a decade before Ultima Online would take the games industry by storm, Richard Garriott and Origin Systems began doing more than just muse about the potential for a multiplayer Ultima. They assigned at least one programmer to work full-time on the technology that could enable just such a product. This multiplayer Ultima was envisioned on a more modest scale than the eventual Ultima Online or even the current Island of Kesmai. It was described by Garriot thus: “What you’ll buy in the store will be a package containing all the core graphics routines and the game-development stuff (all the commands and so on), which you could even plug into your computer and play as a standalone. But with a modem you could tie a friend into the game, or up to somewhere between eight and sixteen other players, all within the same game.” Despite the modest number of players the game would support and the apparent lack of plans for a persistent world, Origin did hold out the prospect of a partnership with CompuServe. In the end, though, none of it went anywhere. After 1987 the idea of a multiplayer Ultima was shelved for a long, long time; Origin presumably deemed it too much of a distraction from their bread-and-butter single-player CRPG franchise.

Another of the big single-player CRPG franchises, however, would make the leap — and not just to multiplayer but all way to a persistent virtual world like that of MUD or Island of Kesmai. Rather than running on the industry-leading CompuServe or even the gamer haven of GEnie, this pioneering effort would run on the nascent America Online.

Don Daglow was already a grizzled veteran of the games industry when he founded a development company called Beyond Software (no relation to the British company of the same name) in 1988. He had programmed games for fun on his university’s DEC PDP-10 in the 1970s, programmed them for money at Intellivision in the early 1980s, been one of the first producers at Electronic Arts in the mid-1980s — working on among other titles Thomas M. Disch’s flawed but fascinating text adventure Amnesia and the hugely lauded baseball simulation Earl Weaver Baseball — and finally came to spend some time in the same role at Brøderbund. At last, though, he had “got itchy” to do something that would be all his own. Beyond was his way of scratching that itch.

Thanks to Daglow’s industry connections, Beyond hit the ground running, establishing solid working relationships with two very disparate companies: Quantum Computer Services, who owned and operated America Online, and the boxed-game publisher SSI. Daglow actually signed on with the former the day after forming his company, agreeing to develop some simple games for their young online service which would prove to be the very first Beyond games to see the light of day. Beyond’s relationship with the latter would lead to the publication of another big-name-endorsed baseball simulation: Tony La Russa’s Ultimate Baseball, which would sell an impressive 85,684 copies, thereby becoming SSI’s most successful game to date that wasn’t an entry in their series of licensed Dungeons & Dragons games.

As it happened, though, Beyond’s relationship with SSI also came to encompass that license in fairly short order. They contracted to create some new Dungeons & Dragons single-player CRPGs, using the  popular but aging Gold Box engine which SSI had heretofore reserved for in-house titles; the Beyond games were seen by SSI as a sort of stopgap while their in-house staff devoted themselves to developing a next-generation CRPG engine. Beyond’s efforts on this front would result in a pair of titles, Gateway to the Savage Frontier and Treasures of the Savage Frontier, before the disappointing sales of the latter told both parties that the jig was well and truly up for the Gold Box engine.

By Don Daglow’s account, the first graphical multiplayer CRPG set in a persistent world was the product of a fortunate synergy between the work Beyond was doing for AOL and the work they were doing for SSI.

I realized that I was doing online games with AOL and I was doing Dungeons & Dragons games with SSI. Nobody had done a graphical massively-multiplayer online game yet. Several teams had tried, but nobody had succeeded in shipping one. I looked at that, and said, “Wait, I know how to do this because I understand how the Dungeons & Dragons system works on the one hand, and I understand how online works on the other.” I called up Steve Case [at AOL], and Joel Billings and Chuck Kroegel at SSI, and said, “If you guys want to give it a shot, I can give you a graphical MMO, and we can be the first to have it.”

The game was christened Neverwinter Nights. “Neverwinter” was the area of the Forgotten Realms campaign setting which TSR, makers of the Dungeons & Dragons tabletop RPG, had carved out for Beyond to set their games; the two single-player Savage Frontier games were also set in the region. The “Nights,” meanwhile, was a sly allusion to the fact that AOL — and thus this game — was only available on nights and weekends, when the nation’s telecommunications lines could be leased relatively cheaply.

Neverwinter Nights had to be purchased as a boxed game before players could start paying AOL’s connection fees to actually play it. It looked almost indistinguishable from any other Gold Box title on store shelves — unless one noticed the names of America Online and Quantum Computer Services in the fine print.

On the face of it, Neverwinter Nights was the ugliest of kludges. Beyond took SSI’s venerable Gold Box engine, which had never been designed to incorporate multiplayer capabilities, and grafted exactly those capabilities onto it. At first glance, the end result looked the same as any of the many other Gold Box titles, right down to the convoluted interface that had been designed before mice were standard equipment on most computers. But when you started to look closer, the differences started to show. The player now controlled just one character instead of a full party; parties were formed by multiple players coming together to undertake a quest. To facilitate organizing and socializing, a system for chatting with other players in the same map square had been added. And, in perhaps the trickiest and certainly the kludgiest piece of the whole endeavor, the turn-based Gold Box engine had been converted into a pseudo-real-time proposition that worked just well enough to make multiplayer play possible.

It made for a strange hybrid to say the least — one which Richard Bartle for one dismisses as “innovative yet flawed.” Yet somehow it worked. After launching the game in June of 1991 with a capacity of 100 simultaneous players, Beyond and AOL were soon forced by popular demand to raise this number to 500, thus making Neverwinter Nights the most populous virtual world to go online to date. And even at that, there were long lines of players during peak periods waiting for others to drop out of the game so they could get into it, paying AOL’s minute-by-minute connection fee just to stand in the queue.

While players and would-be players of online CRPGs had undoubtedly been dreaming of the graphics which Neverwinter Nights offered for a long time, smart design was perhaps equally important to the game’s long-term popularity. To an even greater degree than Island of Kesmai, Neverwinter Nights strove to provide a structure for play. Don Daglow had been interested in online gaming for a long time, had played just about all of what was available, and had gone into this project with a clear idea of exactly what sort of game he wanted Neverwinter Nights to be. It was emphasized from the get-go that this was not to be a game of direct player-versus-player conflict. In fact, Beyond went even Kesmai one better in this area, electing not just to ban such combat from certain parts of the game but to ban it entirely. Neverwinter Nights was rather to be a game of cooperation and friendly competition. Players would meet on the town’s central square, form themselves into adventuring parties, and be assigned quests by a town clerk — shades of the much-loved first Gold Box game, Pool of Radiance — to kill such-and-such a monster or recover such-and-such a treasure. Everyone in the party would then share equally in the experience and loot that resulted. Even death was treated relatively gently: characters would be revived in town minus all of the stuff they had been toting along with them, but wouldn’t lose the armor, weapons, and magic items they had actually been using — much less lose their lives permanently, as happened in MUD.

One player’s character has just cast feeblemind on another’s, rendering him “stupid.” This became a sadly typical sight in the game.

Beyond’s efforts to engender the right community spirit weren’t entirely successful; players did find ways to torment one another. While player characters couldn’t attack one another physically, they could cast spells at one another — a necessary capability if a party’s magic-using characters were to be able to cast “buffing” spells on the fighters before and during combat. A favorite tactic of the griefers was to cast the “feeblemind” spell several times in succession on the newbies’ characters, reducing their intelligence and wisdom scores to the rock bottom of 3, thus making them for all practical purposes useless. One could visit a temple to get this sort of thing undone, but that cost gold the newbies didn’t have. By most accounts, there was much more of this sort of willful assholery in Neverwinter Nights than there had been in Island of Kesmai, notwithstanding the even greater lengths Beyond had gone to prevent it. Perhaps it was somehow down to the fact that Neverwinter Nights was a graphical game — however crude the graphics were even by the standards of the game’s own time — that led to it attracting a greater percentage of such immature players.

Griefers aside, though, Neverwinter Nights had much to recommend it, as well as plenty of players happy to play it in the spirit Beyond had intended. Indeed, the devotion the game’s most hardcore players displayed remains legendary to this day. They formed themselves into guilds, using that very word for the first time to describe such aggregations. They held fairs, contests, performances, and the occasional wedding. And they started at least two newsletters to keep track of goings-on in Neverwinter. Some issues have been preserved by dedicated fans, allowing us today a glimpse into a community that was at least as much about socializing and role-playing as monster-bashing. The first issue of News of the Realm, for example, tells us that Cyric has just become a proud father in the real world; that Vulcan and Dramia have opened their own weapons shop in the game; that Cold Chill the notorious bandit has shocked everyone by recognizing the errors of his ways and becoming good; that the dwarves Nystramo and Krishara are soon to hold their wedding — or, as dwarves call it, their “Hearth Building.” Clearly there was a lot going on in Neverwinter.

The addition of graphics would ironically limit the lifespan of many an online game; while text is timeless, computer graphics, especially in the fast-evolving 1980s and 1990s, had a definite expiration date. Under the circumstances, Neverwinter Nights had a reasonably long run, remaining available for six years on AOL. Over the course of that period online life and computer games both changed almost beyond recognition. Already looking pretty long in the tooth when Neverwinter Nights made its debut in 1991, the Gold Box engine by 1997 was a positive antique.

Despite the game’s all-too-obvious age, AOL’s decision to shut it down in July of 1997 was greeted with outrage by its rabid fan base, some of whom still nurse a strong sense of grievance to this day. But exactly how large that fan base still was by 1997 is a little uncertain. The Neverwinter Nights community insisted (and continues to insist) that the game was as popular as ever, making the claim from uncertain provenance that AOL was still making good money from it. Richard Bartle makes the eye-popping claim today, also without attribution, that it was still bringing in fully $5 million per year. Yet the reality remains that this was an archaic MS-DOS game at a time when software in general had largely completed the migration to Windows. It was only getting more brittle as it fell further and further behind the times. Just two months after the plug was pulled on Neverwinter Nights, Ultima Online debuted, marking the beginning of the modern era of massively-multiplayer CRPGs as we’ve come to know them today. Neverwinter Nights would have made for a sad sight in any direct comparison with Ultima Online. It’s understandable that AOL, never an overly games-focused service to begin with, would want to get out while the getting was good.

Even in its heyday, when the land of Neverwinter was stuffed to its 500-player capacity every night and more players were lining up outside, its popularity was never all that great in the grand scheme of the games industry; that very capacity limit if nothing else saw to that. Nevertheless, its place in gaming lore as a storied pioneer was such that Bioware chose to revive the name in 2002 in the form of a freestanding boxed CRPG with multiplayer capabilities. That version of Neverwinter Nights was played by many, many times more people than the original — and yet it could never hope to rival its predecessor’s claim to historical importance.

The massively-multiplayer online CRPGs that would follow the original Neverwinter Nights would be slicker, faster, in some ways friendlier, but the differences would be of degree, not of kind. MUD, Island of Kesmai, and Neverwinter Nights between them had invented a genre, going a long way in the process toward showing any future designers who happened to be paying attention exactly what worked there and what didn’t. All that remained for their descendants to do was to popularize it, to make it easier and cheaper and more convenient to lose oneself in a shared virtual world of the fantastic.

(Sources: the books MMOs from the Inside Out by Richard Bartle and Gamers at Work: Stories Behind the Games People Play by Morgan Ramsey; Online Today of February 1986, April 1986, August 1986, June 1987, January 1988, August 1988, September 1988, and February 1989; Computer Gaming World of June/July 1986; The Gamers Connection of September/October 1988; Compute!’s Gazette of July 1989; Compute! of November 1991; the SSI archive at the Strong Museum of Play. Online sources include Barbara Baser’s Black Dragon walkthrough, as preserved by Arthur J. O’Dwyer; “The Game Archaeologist Discovers the Island of Kesmai” from Engadget. Readers may also be interested in the CRPG Addict’s more experiential impression of playing Neverwinter Nights offline — and be sure to check out the comments to that article for some memories of old players.)


Posted by on December 22, 2017 in Digital Antiquaria, Interactive Fiction


Tags: , , , , , ,

Games on the Net Before the Web, Part 2: MUD

You are stood on a narrow road between The Land and whence you came. To the north and south are the small foothills of a pair of majestic mountains, with a large wall running round. To the west the road continues, where in the distance you can see a thatched cottage opposite an ancient cemetery. The way out is to the east, where a shroud of mist covers the secret path by which you entered The Land.

— the first text players saw upon entering the University of Essex MUD 

During the time when the original PDP-10 Zork was the hottest game in institutional computing, a fair number of overseas visitors managed to access the machine that ran it at MIT’s Dynamic Modeling Group. One day in 1980, one such visitor, from the University of Essex in Britain, left a strange message on the Zork mailing list: “You haven’t lived ’til you’ve died in MUD.”

When the folks inside MIT investigated, they learned that the spirit of hacker oneupsmanship which had caused them to beget Zork as a response to Crowther and Woods’s Adventure had finally come back around to bite them. Zork had perfected the Adventure formula at the same time that it exploded it, applying a much more advanced parser and a much more detailed and coherent world model to a game that was several times as big. Now, MUD — the Multi-User Dungeon — was taking the next step, applying the innovations of its predecessors to the world’s first shared persistent virtual world. The creators of MUD had first encountered Zork in the form of an unauthorized port which bore the name of Dungeon; thus the name of Multi-User Dungeon made the challenge to the existing state of the art in text adventures all too plain.

Later in 1980, Dave Lebling of the Dynamic Modeling Group penned an article for Byte magazine about Zork which, not coincidentally, included this musing about possible future directions:

Another area where experimentation is going on is that of multiplayer CFS [computerized fantasy simulation] games. Each player (possibly not even aware how many others are playing) would see only his own view of the territory. He would be notified when other players enter or leave the room, and could talk to them.

This would not, however, be the road which Lebling and his colleagues would ultimately choose to go down. Instead they would focus their energies on crafting the most polished, compelling single-player text adventures they possibly could, forming the company known as Infocom to publish them on the microcomputers of the time — a story regular readers of this blog already know well.

And yet the idea of the multi-player text adventure — and with it the idea of a text adventure that was a persistent world to be visited again and again rather than a single game to be solved and set aside — wasn’t going to go away either. On the contrary: a direct line can be traced from Adventure and Zork through MUD and its many antecedents, and on to the functioning virtual societies, with populations in some cases bigger than many small countries here on our real planet, that are the persistent online worlds of today.

As has been the case for more seminal developments in computing history than some might like to admit, MUD was spawned less by a grand theoretical vision than a technical quirk in the computer which ran it. In fact, it was the very same quirk as that which would lead Sandy Trevor over at CompuServe in the United States to create the equally seminal CB Simulator, the world’s first real-time chat program.

The big DEC PDP-10 computer, a staple of institutional computing and with it of hacker culture during the 1970s, might host dozens of simultaneous users, many of whom might be running the same program. It would be absurdly wasteful of precious memory to copy this program’s code over and over into each individual user’s private memory space. Therefore each user was given access to two pools of memory. One, used for program code, was shared among all users on the system — so that if, say, many of them were running the same text editor then the code for that text editor would have to exist in the computer’s memory in only one place. The other area was reserved for the unique data — in this example, the actual file being edited — that each user was working with; this private space could only be accessed by her. In itself, none of this constituted a quirk; it was just good system design, and as such is still used by most computers today.

Roy Trubshaw

But something that was a little quirky about the PDP-10 was noticed by a University of Essex student named Roy Trubshaw in 1977: the fact that nothing absolutely demanded that only static program code, as opposed to dynamic data of other sorts, reside in the shared memory space. With a bit of trickery, the PDP-10 could be convinced to use the shared space to facilitate real-time communications between users who would otherwise be isolated inside their own bubbles. Trubshaw’s first experiments along these lines were as basic as could be: he wrote a couple of programs to cause one user’s terminal to echo text being typed into another’s. “It might seem odd to someone who wasn’t there,” he remembers, “but the feeling of achievement when the line of text on one teletype appeared as typed on the second teletype was just awesome.” From such experiments would eventually spring MUD — for, Trubshaw would realize, there was nothing preventing the shared memory space from containing an entire virtual world rather than just lines of typed text.

The road from typed text to virtual world would, like so much else in gaming history, pass through Adventure. The year of 1977 was also the year of Will Crowther and Don Woods’s pioneering creation, which fascinated Trubshaw as much as it did all of the other hackers who encountered it. The source code to Adventure fortuitously popped up on the University of Essex’s PDP-10 the following year, while Trubshaw was still tinkering with his shared-memory experiments. Taking inspiration from Crowther and Woods’s code — but not too much; he considered the game’s implementation “by and large a giant kludge” — Trubshaw developed a markup language for describing a shared world, to be brought to life by an interpreter program. He named his new language MUDDL, for Multi-User Dungeon Implementation Language. (MUDDL was an amusing if coincidental echo of MDL — pronounced “muddle” — the language the Dynamic Modeling Group at MIT had used to implement Zork. Ditto Infocom’s later MDL-based in-house language, ZIL: the Zork Implementation Language.) The first MUDDL-built shared world to go online modeled, in the grand tradition of countless other first text-adventure creations, the house where Trubshaw had grown up and where his parents still lived. While it’s difficult to anchor these developments precisely in time, the project may have reached this point as early as late 1978, and certainly by 1979.

Richard Bartle

Trubshaw’s most enthusiastic fan and assistant was an undergraduate named Richard Bartle with a taste for Tolkien, Dungeons & Dragons, and the single-player text adventures, like Adventure and Zork, which echoed them. With that perfect resume to hand, Bartle began to function as the world designer for the nascent MUD. In the spring of 1980, shortly after Trubshaw and Bartle together had posted that cryptic message to the Zork mailing list, Trubshaw graduated from university — he blamed the time spent working on MUD for having finished with a second rather than a first — and moved to Belgium, bequeathing all of his old code unto Bartle. Trubshaw wouldn’t do any more serious work on MUD for the next several years, and would only very rarely visit it as a player. From this point on, it would be Richard Bartle’s baby. (There is an interesting parallel here with the original Adventure, which was started by Will Crowther largely as a coding experiment, only to be abandoned by him and developed into a full-fledged game later by Don Woods. The notable difference is that Trubshaw and Bartle, unlike Crowther and Woods, did know one another, and actually worked together on the project for a while.)

Bartle was now working earnestly on making the world of MUD, which he called simply The Land, into a place worth visiting. From its modest beginnings as a facsimile of Trubshaw’s parents’ house it would grow over the course of years to become an immense place, with some 600 rooms, encompassing everything from the expected Tolkienesque fantasy to an area based — without authorization, of course — on Jim Henson’s Fraggle Rock television program.

At first, The Land was inhabited only by a select group of University of Essex students who could locally access the PDP-10 on which it ran. The same PDP-10 was, however, also accessible by a privileged few outside the university who had managed to finagle, by means legal or extra-legal, access to the early Internet. One of the first such outsiders to drop by was a precocious 17-year-old named Jeremy San who spent his days prowling the Internet, looking for interesting things. MUD, he knew immediately, was very interesting indeed. He became its greatest propagandist among the tiny British modem fraternity of the time; a huge percentage of the people who wound up playing MUD did so thanks to his encouragement. (Jeremy San took the handle of “Jez” on MUD, a nickname which leaked out of its virtual context to become his professional name. As Jez San, he would go on to become a major figure in British game development, responsible for Starglider among other hits.)

Enough outsiders like San were soon playing MUD to prompt the division of players into “internals,” meaning people playing from within the university itself, and “externals,” meaning people logging on from outside its hallowed halls. The university’s administrators proved rather astonishingly generous with their computing resources, allowing players from all over the country and, eventually, all over the world to log on and play during evenings and weekends. But make no mistake: doing so was a tall order in the Britain of the early 1980s, where electronic communications in general were still in a far more primitive state than the United States of the period, where even simple local phone calls were still charged by the minute and modems were rarer than hen’s teeth among home-computer owners. To add to the logistical difficulties, MUD didn’t even become available on the university’s PDP-10 until 1:00 AM, after everyone else had presumably finished doing their real work on the machine. Nevertheless, enough people combined the requisite privilege, cleverness, and dedication that the software’s limitation of no more than 36 simultaneous players quickly began to frustrate.

A typical MUD session

For all that even MUD‘s very name implied it to be nothing more than a multiplayer version of Zork/Dungeon, the move from a single-player game to be solved and set aside to a persistent world to be inhabited by the same players for months or years changed everything. The designs questions which confronted Bartle were completely different from those being debated contemporaneously inside the early Infocom. In fact, they had far more in common with those still being debated by the makers of the massively-multiplayer games of today than they did with those surrounding the single-player games of their own time. In a single-player game, the player is the star of the show; the (virtual) world revolves around her. Not so inside The Land. Most traditional text-adventure puzzles made no sense at all there. The first person to come along and solve a puzzle might have fun with it, but after that the shared world meant that what was solved for one was solved for all: the door remained unlocked, the drawbridge remained lowered, etc. This was not, needless to say, a good use of a designer’s energy. MUD did include some set-piece puzzles which could be solved by simply typing an answer, without affecting the environment — riddles, number sequences, etc. — but even these became mere pointless annoyances to a player after she had solved them once, and tended to be so widely spoiled by the first player to solve them that they too hardly seemed a good use of a designer’s time.

The most innovative puzzles in The Land were those which took advantage of this being a multiplayer text adventure to demand cooperation; think of a heavy portcullis which can only be lifted by several players straining together. Perhaps the cleverest example of this species of puzzle was The Swamp, a huge maze with an immensely valuable crown hidden at its center. Because it was a swamp, the tried-and-true drop-and-plot technique for solving mazes was a nonstarter; items dropped would just disappear below the surface. The only way to defeat the maze was rather with a team of a dozen or more players, leaving one player behind in each room to mark its identity.

Yet puzzles — even brilliant puzzles like this one — were never really the heart of MUD‘s appeal; even the Swamp maze became a fairly rote exercise as soon as the first group figured out how to approach it. Although it presented itself in the guise of a text adventure, MUD often played more like a CRPG. Replacing the deterministic behavior of Adventure and Zork was a focus on emergent behavior. Each player had scores for strength, stamina, and dexterity, which determined the outcome of many actions, including combat with the many creatures which dotted the landscape. These attributes increased as one’s in-game accomplishments grew.

Like most traditional text adventures, MUD had a system of points for measuring the player’s achievements. Less typically, but very much in keeping with the CRPG side of the game’s personality, this score translated into levels, which among other things allowed players to attach honorifics (“Hero,” etc.) to their names. Points became a sort of currency within The Land; it was possible to transfer some of your points to another player by “hugging” or “kissing” her, and this was often done in exchange for goods and services.

The most straightforward way of scoring points, however, was at first glance lifted straight from Adventure and Zork: treasures were scattered about The Land, which players could retrieve and drop into a designated location. But this scheme alone was obviously unsuitable to a multiplayer context, at least if Bartle didn’t want to spend all his time hiding new treasures for players to discover. Instead he came up with a scheme, always more tolerated than liked by even the most dedicated players, by which The Land was periodically “reset,” with all treasures moved back to their original starting locations. Such resets, besides being something of a blight on MUD‘s very identity as an allegedly persistent online world, were an imperfect solution in that the more experienced players were the ones who knew where the treasures lay after a reset; they thus could rush over to claim them before the newbies had a chance. It would often take the veterans no more than five minutes to scarf up all the treasures during a “reset rush hour,” an exercise as unchallenging for them as it was baffling for the newbies.

But then, the same newbies who were left nonplussed by the reset rush hour soon found much worse to complain about. Not only did MUD feature permadeath, but it allowed its experienced players to kill the new ones. Indeed, it even encouraged such behavior by rewarding the griefers with one-twenty-forth of the points their victims had accumulated over the course of their careers. Combined with the cliquey nature of MUD‘s culture, it could make The Land a hugely off-putting place to visit for the first time.

Tormenting the newbies was a favorite pastime among the regulars. Richard Bartle tells the sad story of one of MUD‘s first two externals, who managed to connect from all the way over in the United States in 1980 and got nothing but grief for his trouble:

Also in the game was Niatram, one of the system operators (who can’t spell his name backwards). He decided to loom up on this [newbie] character, follow him around a bit, then kill him. This he repeated several times, gaining plenty of points in the process. Finally, the newcomer was at his wits’ end.

“Who’s this Niatram character?” he asked. “He keeps following me around and killing me!” “Yes, he’s done that to me before,” came the reply. “I think he may be dungeon-generated!” At this point Niatram appeared, and out of despair his victim quit, rather than be killed yet again by this “artificial person.”

Today’s virtual worlds are generally wise enough to prevent this sort of thing by one means or another.

Player agency was another area where MUD was dramatically different from the massively-multiplayer games of today, but in this case the difference was, at least from some points of view, a more positive one. Once ordinary players, or “mortals” in MUD speak, reached a certain level, they became “wizards” and “witches” — known by the unisex term of “wizzes” — who were perhaps better described as gods, given that they were granted a stunning degree of control over The Land. “Wiz mode” was originally simply the debug mode used by Trubshaw and Bartle themselves, but it wound up becoming a key part of the game, Bartle’s answer to the expert player’s question of “Well, what now?”

Becoming a wiz required one to amass 76,800 points without getting killed, no mean feat in the cheerfully genocidal realm of The Land. The powers wiz mode conveyed extended to the point of being able to crash the game at will. Tired of watching wizzes invent ways to do so to prove their cleverness, Bartle actually added the verb “crash” to the wizzes’ vocabulary in order to convey the message that, yes, you have the power to do this; you therefore don’t have to bother inventing more convoluted methods for doing so, like, say, plucking the rain from the sky of two separate rooms and mashing the lot together. Safe in the knowledge that they could crash the game at will, effortlessly, most wizzes did indeed see no reason to bother actually doing so. Which didn’t, of course, prevent all the accidental crashes: it was always a sure sign that someone new had attained wiz status and was experimenting with her powers when the game just kept going down over and over.

Once mastered, though, the wizzes’ powers truly were extraordinary. They could snoop on mortals, watching the commands they typed and what appeared on their screens in response; could move objects from any room to any other room without touching them; could move themselves or other players from any room to any other room; could use a “finger of death” spell to instantly kill any mortal, permanently; could grant instant wiz status to any non-wiz they chose.

It seemingly defies all logic — defies everything we know about human nature — that a game willing to grant some of its players such powers could have sustained itself for a week, much less the years during which MUD ran at the University of Essex. It was all indicative of what a different sort of game MUD really was. Running on a university’s computer, free to access by anyone with the wherewithal to make a connection, MUD was a very different proposition from a commercial game, more a joint creative experiment than a product. Bartle:

MUD is an evolving game, and so indeed it should be. It has been incremented gradually, with new ideas put in to be instantly tested by a horde of willing wizzes, or mortals if it was something that they could use (the various “injury” spells — BLIND, DEAFEN, CRIPPLE, DUMB, and CURE — for example). This is one of the great strengths of doing MUD at a university; it’s all research. If a commercial company were to put up a game riddled with bugs, the players would be justifiably upset when it crashed on them. Here, though, it’s free for them to play and they actually like finding mistakes, because it gets them one over on me (and occasionally gets them some points for their honesty!). And it’s also good because we don’t have to pay people to play-test, either — plenty will do it willingly in their spare time for free!

So there will always be a place for MUDs at universities, simply so that research into them can proceed. Universities can have “programs,” whereas commercial companies must have “products.” Products don’t crash (well, not often!) and they are nice and stable. Programs crash like nobody’s business and you never know from one day to the next whether some terrible new command has been added which you don’t know about, but which someone who does is about to use on you. Products are fun, but they don’t change until everything has been thoroughly tested; programs are exciting in their volatility.

Perhaps there is a place for the “not fully tested” system. Even if I as a player did have to put up with a crash every 20 minutes (MUD needs to be reset once a night on average), I think that experiencing the excitement of seeing things evolving and of being among the first to use the novel commands would make me happy to play the program, not the product. Fortunately, enough people think the same way to make debugging that much easier, and to encourage new additions to make the game even more fun for generations of adventurers to come.

Every rule in the game existed, it seemed, to be exploited; mechanically, the place was a leaky sieve that poor Richard Bartle was constantly rushing around to patch. MUD‘s players were endlessly creative when it came to finding exploits. One of Bartle’s favorite anecdotes involves two players who were each about halfway to wiz status. Deciding they’d had enough incrementalism, they hatched a plan to put themselves over the top. First, one of the players kissed the other some 1400 times in succession to transfer all of his points to her, giving her enough to make wiz. Then the newly minted wiz used her special powers to grant instant wiz status to her helper. In response to that, Bartle had to add a rule that players could only use hugs and kisses to transfer points to those with a smaller point total than their own. Just another day at the office…

For all that exploits were a way of life among The Land’s denizens, the absolute power granted to the wizzes proved not to corrupt absolutely. In fact, just the opposite. For all that there was a definite hierarchy in place among the inhabitants of The Land, for all that tormenting the newbies was regarded as such good sport by virtually everyone, a certain attenuated but real code of fair play which Bartle himself did nothing to institute took hold within MUD. And it was the wizzes, the players with the ability to wreck The Land almost beyond redemption if they chose to do so, who came to regard the code as most sacred. Bartle:

The wizzes were once mortals themselves. Wizzes know all too well what it’s like to be summoned to a cold, dark room and left alone with the word “hehehe” ringing in your ears. They know the disappointment in forging through the swamp for half an hour only to find that someone has swapped the incredibly valuable crown in the centre for a fake one. They’ve felt the pangs of outrage when you’ve been attacked by a souped-up bunny rabbit which took you 15 minutes to kill. In short, they know when to stop.

Wizzes make the game. They rule it, they stamp their personalities on it, and they give mortals something to aim for, a goal, a purpose, something which explains why they’re in there hacking and slaying. If MUD does nothing else for multi-user adventure games, for evolving the concept of a wiz it should always be remembered.

By 1984, 53 players had made wiz status: 40 from England, 5 from Scotland, 4 from Wales, 1 from Ireland, 1 from the United States, 1 from Czechoslovakia(!), 1 from Malaysia(!!). The creative powers granted to them made MUD a self-sustaining community. After Bartle built The Land and made it available, he needed do no more. The place was perfectly capable of evolving without him.

“The people,” Bartle noted, “are the game” of MUD. The sense of shared ownership could make The Land a downright cozy place when the inhabitants took a break from killing one another. There was a bizarre craze for Trivial Pursuit for a while, with wizzes shouting out questions which mortals could answer for in-game rewards. And there was at least one MUD wedding, between Frobozz the Wizard and Kate the Enchantress, with the happy couple taking up honeymoon residence in the coveted “Wizard’s Chamber.” (“What interaction occurred thereafter is unknown,” wrote the journalist covering the story.)

Christmas was always a special time of the year, when the wizzes would institute a strictly enforced ban on player-versus-player combat, scatter the landscape with holly, snowmen, wreaths, and plum puddings, and plant Christmas trees in the forests, where one was now apt to encounter a wandering Santa Claus with his reindeer in lieu of the usual monsters. The wizzes desisted from sporting with the newbies to vicariously enjoy their surprise and delight when they logged on to find The Land transformed into a winter wonderland, complete with snow. (“What’s all this snow?” “I don’t know. I just saw Father Christmas go by, and someone has given me this cracker…”) Players would band together to sing Christmas carols, swigging all the while from a shared bottle of rum and diligently role-playing their resulting intoxication.

But who, you might wonder, were all these people who flocked to The Land every night? Like the CB Simulator fraternity on CompuServe, it was a diverse demographic who enjoyed this glimpse of an online future, ranging from teenage hacker whiz kids like Jeremy San to academics in their forties. Perhaps the most dedicated player of all was Sue, the first woman to make wiz, who slept for just three hours or so each night before MUD went online, then played for the full six hours it was available every morning before heading off to her ordinary office job.

Just as on CB Simulator, some of the other people on MUD who claimed to be women weren’t really women at all. Among this group was one Felicity, who was eventually found to be the avatar of a dude named Mark. (Felicity had the reputation, for what it’s worth, of being the “kindest, most responsible wiz of all time.”) Once Felicity/Mark’s deception was uncovered, players claiming to be women were routinely greeted with suspicion. A favorite tactic was to enlist bona-fide females like Sue to engage the latest suspect into a discussion of, say, dress sizes. Thereby were the imposters generally discovered quite quickly.

A wiz called simply Evil gained the reputation of being the most eccentric of all time — quite an achievement with this group. Bartle:

If you wanted to get to any room from any other, no matter how far away, he could give you the shortest route instantly. This was despite the fact that he laboured under a tremendous disability: east-west dyslexia.

It is for this that Evil is best known. His entire in-the-head map of MUD, and all those he wrote down on paper, were flipped east for west. His misapprehension extended to commands, so if he wanted to go west from the start, which is to the left, he’d think it was to the right, and that the command for going to the right was west. So he’d get it correct, but in the wrong way! So absolutely everything was inverted, in a kind of “Evil through the looking glass”. Indeed, when I finally found out about his error I put a looking glass in MUD to celebrate!

He didn’t realise his mistake for years after he’d made it to wiz, and if people used left/right descriptions instead of west/east, he just thought they were barmy. Only when I drew a map of MUD on a blackboard did he finally discover his gaffe, and to this day he thinks a subtle change in the physics of the universe caused everyone in the world to swap east for west in their heads except for him, who remained unaffected due to his enormous and obvious intelligence…

While MUD‘s personality as a game may have been molded more by the cast of eccentrics who inhabited The Land than by Richard Bartle himself, the latter would always remain by far the most important person associated with the game. MUD was, after all, his baby at the end of the day. Even after finishing his degree, he remained at the University of Essex as an artificial-intelligence researcher, with the ulterior motive of continuing to further the cause of MUD. And indeed, from the standpoint of publicity as from so many others, he continued to prove himself to be its greatest asset. As charming and articulate in person as he was an easy and prolific writer, Bartle was able to attract far more press attention for his odd experiment running in the middle of the night on an academic computer than one might expect. He became a fixture in the British computing press of the 1980s, penning long articles about life in The Land for magazines like Practical Computing and Micro Adventurer. His efforts in this regard have proved a goldmine to modern historians, who have come to recognize in the way that his contemporary peers couldn’t possibly be expected to what a landmark creation this first persistent multi-player virtual world really was. It’s largely Bartle’s old articles that form the basis of this new article of mine. With so much in online-gaming history already lost to the ephemerality of the digital medium, we can be thankful that Bartle was as prolific as he was.

Still, preserving MUD culture for posterity wasn’t Richard Bartle’s main motivation for writing these articles. This academic researcher didn’t want MUD to remain a tiny research project at a second-tier university forever; had a keen interest in moving it beyond the confines of the University of Essex. Early on he gave copies of the software to universities in Portugal, Sweden, Norway, and Scotland, sparking the transformation of the name MUD from a designation for a specific virtual world to a generic tag applied to all parser-driven textual virtual worlds. Despite the identical mechanics, the people who played each incarnation of the software which Bartle shared so freely gave each version of The Land a distinctive character. The MUDs in Scotland and Norway, for instance, were much more easygoing than the one at the University of Essex; delegations from the former expressed horror at the sheer amount of killing that went on in the latter when they came for a visit.

Bartle had no doubt that MUDs represented a better model for adventure gaming, the direction the entire genre by all rights ought to be going:

Like it or not, in the next few years multiplayer games like MUD are going to become the dominating factor in adventure games. The reason for this is quite simple — MUDS are absolutely fantastic to play! The fact that it’s You against Them, rather than You against It, adds an extra electricity you just can’t experience in a single-player game. If you like adventure games already, MUD will absolutely slay you (often literally!).

This quote dates from 1984. As it would imply, Bartle by that year had decided the time was right to start commercializing his research experiment. He made a deal to bring MUD to Compunet, a pioneering online service for British owners of Commodore computers, initially funded largely by Commodore’s own innovative United Kingdom branch. The first service of its kind in Britain, Compunet would muddle along for years without ever achieving the same success as QuantumLink, its nearest equivalent in the United States. The tiny staff’s efforts were constantly undone by the sheer expense of telecommunications in Britain, along with a perpetual lack of funding to set up a proper infrastructure for their service; Compunet had, for instance, only a single access number for the entire country, meaning that the vast majority of potential customers had to pay long-distance surcharges just to access it. Although it was available on Compunet for several years, MUD too just never managed to make much of an impression there.

MUSE hosted a gathering of wizs from the University of Essex MUD to mark the launch of BT MUD. Richard Bartle is the mustachioed fellow standing near the center of the group.

But even as the Compunet MUD was failing to set the world on fire, the game had attracted an important patron. Simon Dally was a longtime gamer, first of tabletop and later of computer games, who had done very well for himself in the book-publishing trade.  When he stumbled across MUD for the first time, it was love at first sight. Dally was certain that MUD could change the world, and that it could make him and its creators very rich in the process. He, Bartle, and Roy Trubshaw — the last had recently returned to England from Belgium — formed Multi-User Entertainments, or MUSE (not to be confused with the American software publisher of the same name), to exploit what Dally, even more so than the always-enthusiastic Bartle, believed to be the game’s immense commercial potential. (Trubshaw was more skeptical, and seemed to have agreed to work on MUD once again more for the programming challenge than for its supposed potential to make him rich or to change the world.)

The first project MUSE undertook was to completely rewrite MUD, with Trubshaw once again doing the low-level architecture and Bartle building a world upon these technical underpinnings. Whereas the old MUD had been inextricably tied to the rapidly aging DEC PDP-10, what became known as MUD2 was designed to “run on just about any minicomputer or mainframe in the world” with a minimum of porting; its first incarnation ran on a DEC VAX, a much newer and more powerful piece of kit than the PDP-10. Among many other improvements came the inevitable increases in scale: the old limit of 36 active players on the original MUD became 100 on MUD2, and The Land itself could now be twice as large. As befitted his day job as an artificial-intelligence researcher, Bartle was particularly proud of his new non-player creatures — known as “mobiles” in MUD speak — which were able to talk and trade with human players whilst pursuing goals of their own.

A proud Simon Dally announces the BT MUD.

Dally took the new-and-improved MUD to British Telecom and convinced them to make it available as a standalone dial-up service. Playing MUD in its latest commercial incarnation wouldn’t be cheap: it cost £20 to buy the starter pack, plus £2 per hour to play, all on top of the cost of the phone call needed to dial in. With prices like these, the new MUD endeavored to shed its scruffy hacker origins and project an upscale image, beginning with a gala launch at The London Dungeon. Dally used his connections to get a glossy book published:  An Introduction to MUD, penned by one Duncan Howard. “I know the BT venture is the just the start of something truly enormous,” Dally said. “Our MUD development language — MUDDL — will allow anyone to come to us with an idea for an interactive type of game, and allow us to implement it quickly and cheaply. We are certainly ahead of the States, where MegaWars III, a rather limited interactive game, is going down well, and we have high hopes of selling MUD to the Americans.” Those hopes would come to fruition in remarkably short order.

But in neither country would the game achieve the “truly enormous” status Dally had so confidently predicted. The BT venture proved particularly disappointing. One step Bartle took which may not have been terribly wise was to convince much of the Essex MUD old guard to migrate over to the new game, pulling strings to get them online at reduced or free rates. Thus he imported the old game’s murderous culture and set up an upper class and a lower class of players at a stroke. A writer for Acorn User magazine who surveyed the newbies found that they “don’t much like MUD. They can’t find any treasure, don’t know what to do, and spend their time waiting for the next reset or chatting to each other in the bar. Typical answers to my inquiries included phrases like ‘bored’ or ‘where’s all the T?'”

In addition to the culture clashes, the BT MUD was also a victim, like the Compunet MUD before it, of all the difficulties inherent in telecommunicating in Britain at the time. MUSE could measure their falling status within British Telecom by the people they were assigned as account managers. “It was a gradual decline,” remembers Bartle, “from speaking to board-level directors to being signed off by a youth-opportunities employee.” The next wave of adventure games in Britain the BT MUD most definitely didn’t become. It continued to run for years, but, after the initial publicity blitz puttered out, it existed as little more than another hangout for the same old small in-group it had attracted at the University of Essex. The BT MUD was finally closed down in early 1991, when British Telecom decided they’d had enough of an exercise that had long since become pointless from their perspective.

MUSE’s efforts did fare significantly better in the United States. Dally convinced no less of an online player than CompuServe to take the game as an offering on their service. It went up there in the spring of 1986 under the name of British Legends. Perhaps assuming that what had worked for Lord British of Ultima fame could work equally well for them, CompuServe’s marketers emphasized the Anglo connection at every opportunity. They declared over-optimistically that “in England, the game is a sensation,” with “thousands of players.” A couple of years after British Legends made its debut, CompuServe began offering their service in Britain as well the United States, facilitating the first meetings between large numbers of British and American players in the very same MUD. “I’ve learned more about the United Kingdom from the British players than I learned in all my college courses,” enthused one American player.

Mechanically, British Legends remained unchanged from its other incarnations, including the wizzes running roughshod over the place. Still, it tended to be a less cutthroat world than had the Essex MUD, with more time given over to socializing and cooperating and less to fighting. And of course it was blessed with an initial group of players who were all starting alike from scratch, and thus largely avoided the class conflicts which plagued the BT MUD. It would remain available on CompuServe for more than thirteen years, becoming by far the most long-lived of all the MUD incarnations licensed by MUSE. It was successful enough that virtually all of CompuServe’s competitors either made their own licensing deals with MUSE or came up with an alternative MUD of their own — or, in some cases, did both. Some of the latter were very innovative in their own right. GEnie’s Imagine Nation, for example, strained to be a kindler, gentler sort of MUD, eschewing combat and even goals entirely in favor of providing an environment for its denizens just to hang out and socialize, or to invent their own games in the form of scavenger hunts and trivia contests. In later years MUDs in general would increasingly go in this direction. (See, for instance, the modern interactive-fiction community’s longstanding, beloved IF MUD, where people congregate to play text adventures together, to discuss game design, and just to chitchat rather than to chase one another around with virtual swords.)

Yet the relative commercial success MUDs enjoyed in the United States shouldn’t be overstated. Even British Legends, probably the most popular single MUD incarnation ever, was always a niche product even within that small niche of the public with the money and the interest to access a service like CompuServe in the first place. The online services made MUDs available because they could do so fairly cheaply, and because, once they were up, they tended to produce a self-sustaining community of hardcore players which required virtually no nurturing — that is to say, they required virtually no further financial investment whatsoever, thanks to the self-sustaining genius of wiz mode.

But in the big picture, Richard Bartle and Simon Dally’s mutual passion was destined to be far more influential than it would be profitable. The death of the dream of MUDs as the dominant adventure-game form of the future came in a tragically literal fashion in 1989, when Dally killed himself. It’s impossible to say to what extent his suicide was prompted by his disappointment at the failure of MUD to achieve the world domination he had predicted and to what extent it was a product of the other mental-health issues from which he had apparently been suffering, as manifested in behavior his friends and colleagues later described as “erratic.” There can be no doubt, however, that his death marked the definitive end of the MUD’s flirtation with mainstream prominence.

Any attempt to explain why MUDs remained a niche interest must begin with their textual nature. MUSE had been formed just after the text adventure had reached its commercial peak and was about to enter its long decline, over the course of which it would gradually be replaced on store shelves by the graphic adventure. A game that consisted of nothing but text was doomed to become a harder and harder sell after 1985. But MUDs had their problems even in comparison to single-player text adventures — problems which Richard Bartle was always a bit too eager to overlook. Many players of adventure games loved puzzles, but, as we’ve seen, puzzles didn’t really work all that well in MUDs. Many craved the experience of seeing a story through from beginning to end, knowing all the while that they were the most important character in that story; MUDs couldn’t provide this either. Ask many who had tried to like MUDs, and they’d tell you that they were just too capricious, too unstructured, too difficult to get into, even before you started wrestling with a parser that sometimes seemed willfully determined not to understand you. MUDs had invented the idea of the persistent online virtual world, but the millions and millions of players who would later come to live a good chunk of their lives in such places would have a very different technological window onto them.

The way forward, in commercial terms at least, would be through more structured designs attached to cleaner interfaces, eventually using graphics instead of text wherever possible. While the hardcore who loved MUDs for the very things the casual dabblers hated about them complained — and by no means entirely without justification — that something precious was being lost, game developers would increasingly push in this more accessible direction. Instead of making multi-player text adventures with CRPG elements, they would build their persistent worlds on the framework of the traditional CRPG — full stop.

(Sources: the books MMOs from the Inside Out by Richard Bartle and Grand Thieves and Tomb Raiders: How British Videogames Conquered the World by Magnus Anderson and Rebecca Levene; Byte of December 1980; Micro Adventurer of September 1984, October 1984, November 1984, December 1984, January 1985, February 1985, and March 1985; Practical Computing of June 1982 and December 1983; Popular Computing Weekly of December 20 1984, February 28 1985, and May 23 1985; Commodore Disk User of November 1987; Your Computer of September 1985; Sinclair User of November 1985; Computer and Video Games of November 1985; Commodore User of February 1986; Acorn User of July 1985, October 1985, April 1987, and June 1987; New Computer Express of March 9 1991; Online Today of February 1986, January 1988, and March 1988; The Gamer’s Connection of September/October 1988; Questbusters of October 1989. Online sources include the article “CNET — Moving with the Times” from Commodore Apocalypse and Gamespy’s history of MUDs.

You can still play MUD1 today by telnetting to, port 27750. You can play MUD2 by telnetting to, port 27723.)


Posted by on December 15, 2017 in Digital Antiquaria, Interactive Fiction


Tags: , ,

Patreon About-Face

As many of you have doubtless heard by now, Patreon has abandoned all plans to institute the billing changes which have caused such chaos and consternation over the past week. CEO Jack Conte has posted a public apology, and it’s a very good example of the genre: no making excuses; no placing the blame partially on us with weasel words like “if you felt like you were wronged”; no asking for forgiveness that will have to be earned, not given. There’s not even a single “awesome!” in sight; unlike most of Patreon’s missives, it doesn’t read like it was written by an over-enthused Valley girl. Patreon says that they still want to address the problem which allegedly prompted the changes in the first place — a problem which was never a problem at all for this blog — but they promise to do so in consultation with the community instead of unilaterally, and they promise to find ways to protect the small donors on which this blog and many other projects depend. “We are nothing without you,” Jack Conte writes. He’s correct; hopefully Patreon will never forget that again.

All of this leaves me, not for the first time this week, with a tough decision on my hands — and just after I’d settled on what seemed the right direction forward at that.

On the one hand, if not actively angry anymore I’m still highly irritated. This last week has cost me stress and restless nights, and I got the joy of devoting last weekend almost entirely to working out the technological bits and pieces that would be needed to roll out a more localized funding solution relying on Memberful. And of course this has all cost me a few dozen patrons, many of whom had been with me for a long, long time. I don’t blame them for jumping ship. Not in the least: I blame Patreon. So, there’s a part of me that still wants to wave my middle finger in Patreon’s direction and tell them exactly what to do with their apology.

Yet there is the cutting-off-the-nose-to-spite-the-face factor in doing so. I’m actually kind of proud of the Memberful solution I was pulling together, but it has its drawbacks. It relies on a lot of WordPress hacking to work just the way I want it to — the kind of thing that all of my experience with software tells me would become a source of constant headaches, needing to be tinkered with as new releases of WordPress come down the pipe. And I’m sure there are bugs which I would spend lots of time chasing, and then there’s the confusion inherent in offering dueling pledging systems for the same blog. I would also be forced to take on the role of customer-support guy: figuring out why people’s credit cards were rejected by Stripe, why they couldn’t get into their accounts, etc. I was prepared to take on that burden if there was no other alternative, but I’d prefer to avoid it. While I like to think I’m a decent programmer, I believe I’m a very good to excellent writer. (This serves as the universe’s way of compensating me for the staggering number of everyday things at which I’m freakishly terrible.) I’ve worked many years in IT, but now, at 45 years old, I want to be doing the thing I’m best at as often as possible. In other words, I’d like to just be a writer, and to let someone else — like Patreon — take care of the technical and customer-support stuff. That was actually working out pretty well before last week. So, my decision, arrived at not without some agonizing, is to stay with Patreon as this blog’s primary support mechanism.

Now we get to the heart of the matter. What does this mean for those of you who are or were patrons, but took action of your own in response to Patreon’s boneheaded move?

Well, those of you who front-loaded your pledges onto the first article each month to avoid the extra fees can change things back to the way they were in a few weeks. Because pledges are once again aggregated at the end of the month, just as they’ve always been in the past, I’ll pay the exact same amount in processing fees either way.

Those of you who jumped ship entirely have a harder decision. Obviously I would like to see you come back — would like for you to give Patreon another chance — but only you know whether that feels right. I’ve already seen the gamut of sentiment expressed in your comments over on my area of the Patreon site, from “I’ve lost all trust in Patreon” to “Sites backing down in the face of user outrage needs to be encouraged, not discouraged.”

If you can see your way to coming back, I’d be thrilled to have your support again. But if you just can’t justify it, I totally understand. This decision wasn’t an easy one for me either. Perhaps somewhere down the road, if Patreon continues to behave as they promise to from now on, you’ll feel that they’ve earned your forgiveness and your business. Either way, I know where the blame for the loss of your support resides. Hint: it’s not with you. Thank you for all of your support in the past.

And with that all said, I’m going to spend the next couple of days working on writing articles instead of stressing over Patreon and/or Memberful, publish a long and (I think) interesting article tomorrow, and then enjoy a weekend spent visiting the Christmas markets and putting up a tree with my wife instead of sitting hunched in front of the computer putting the final pieces of a new pledging solution in place. I figure I deserve it; it’s been one hell of a last seven or eight days. I hope you all have a similarly relaxing weekend in the offing.

Thank you for your past, current, and/or future support, as the case may be, and happy holidays! See you tomorrow with more piping hot digital antiquaria!


Patreon Update

I’ll be rolling out a new pledging system for this site next week. Built on a platform called Memberful, it will let you pledge your support right from the site, without Patreon or anyone else inserting themselves into the conversation. The folks from Memberful have been great to communicate with, and I’m really excited about how this is shaping up. I think it’s going to be a great system that will work really well for many or most of you.

That said, my feeling after much vacillation over the last several days is that I won’t abandon Patreon either. Some of you doubtless would prefer to stay with them, for perfectly valid reasons: for high pledge amounts, the new fee schedule is much less onerous; some of you really like the ability to pledge per-article rather than on a monthly basis, which is something no other solution I’ve found — including Memberful — can quite duplicate; some of you really want to keep all of your pledges to creators integrated on the same site; etc. And of course it’s possible that Patreon will still do something to mitigate the enormous damage they did to their brand last week. At the risk of introducing a bit more complication, then, I think the best approach is just to clearly explain the pros and cons of the two options and leave the choice in your hands.

For right now, those of you who are current Patreon patrons don’t need to do anything. The article which I’ll publish on Friday will still be bound by the old rules; the new ones won’t go into effect until December 18. On or about December 18, I’ll introduce you to the new pledging system, and you can decide then what works best for you.

Thanks for your relentless positivity during what was a pretty stressful few days — and a big welcome to the few of you who defied all the conventional wisdom by signing on with Patreon in the midst of this whole brouhaha. The stress is vastly less now: the path ahead is becoming clear, and, best of all, I continue to be blessed with the best readers in the world.

More news in a week!