How long is the coast of Britain? Would you be surprised if I told you that it’s infinitely long? Infinite coastlines are just one of many byproducts of the strange mathematics of fractal geometry.
Why is geometry often described as “cold” and “dry?” One reason lies in its inability to describe the shape of a cloud, a mountain, a coastline, or a tree. Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line.
— Benoit Mandelbrot
For centuries mathematicians and philosophers were dogged by a nagging problem. Classic Euclidean geometry, with its regular lines and planes and solids, was elegant and useful, but fell down when it came to describing the natural world. It was ill-equipped to describe the seemingly endless spatial complexity of a leaf — or, for that matter, a section of coastline. Indeed, it’s largely the very presence or absence of such chaotic complexity that we subconsciously use when we decide whether something is natural or human-made. The world humanity has built is created in the image of our geometry, while the universe that birthed us, even the very forms that enclose us, is defined by its refusal to conform to our ideals of coordinates and shapes and angles. To believe that the universe is somehow corrupted or deformed because of this refusal would be the height of hubris. “These are not natural irregularities, but with respect to our fancy only; nor are they incommodious to the true uses of life and the design of man’s being on earth,” wrote Richard Bentley in the time of Swift and Pope. Yet how to reconcile the riot of natural beauty with the cooler constructions of Euclid? Until very recently we literally lacked a mathematical language to describe the world around us.
In 1975 the Polish/French mathematician Benoit Mandelbrot provided such a language when he published his theory of fractal geometry. Mandelbrot discovered that certain solution sets to certain equations generate what he called “fractals” — patterns that are infinitely detailed. Such sets were much better suited to describe natural forms than anything heretofore present in the geometric arsenal.
It’s this seeming paradox of infinite detail that’s the thorniest conceptual part of natural forms and fractal geometry alike. To begin to get a handle on it, consider the “snowflake” curve, an idea first proposed by Helge von Koch in 1904. Imagine that you start with a line forming a simple geometric motif.
Now you replace every straight-line segment with another copy of the original motif.
You do it again…
And you keep on doing it over and over and over, to infinity.
As you look at a coastline or a leaf at ever greater levels of magnification, there are always more details, more curves and changes to discover. This continues forever — or at least until you hit the level of individual particles of matter, the sort of resolution limit of the universe itself. Thus that aforementioned infinite — or at any rate unmeasurable — coastline. “Coastline length turns out to be an elusive notion that slips between the fingers of one who tries to grasp it,” wrote Benoit Mandelbrot. “All measurement methods ultimately lead to the conclusion that the coastline’s length is very large and so ill-determined that it is best considered infinite.” What to make of such a logical monstrosity?
Mandelbrot found a way out in some theories of dimensionality espoused by the controversial German mathematician Felix Hausdorff more than half a century before. The easiest way to describe conventional notions of dimensionality mathematically is to say that an object’s number of dimensions equates to the number of numbers needed to describe a single point on it. Thus a line (X) is one-dimensional; a plane (X and Y) two-dimensional; a solid (X, Y, and Z) three-dimensional. Hausdorff, however, floated the counter-intuitive notion that dimensions could be fractions as well as whole numbers. Consider a line. As it gets more wiggly, ever more complex in its gyrations, its dimensionality increases, until it becomes infinitely wiggly, and completely fills — or, perhaps better said, becomes — a plane. The pattern labelled 1a below, for example, is made up of infinite repetitions (within the bounds of the resolution of the picture) of the hook-like line segment labelled 1b. According to Hausdorff’s formulation, it has a dimensionality of 1.8687.
This idea of fractional dimensionality is a strange one, but seeing complex lines in terms of dimension rather than distance does allow us to dodge the paradox of infinite coastlines, restoring a mathematical order in which the coast of Britain is again longer than that of, say, the island where I live these days, Funen.
The most interesting fractal equations are not those like the ones shown above that simply repeat a pattern, but rather those that echo themselves but never exactly repeat, to infinity. The classic example is one of the first discovered by Mandelbrot himself, the appropriately named Mandelbrot set, consisting of all solutions to (z(n) ← z(n – 1)2 + c) which converge toward zero rather than expanding toward infinity. (Whilst trying to avoid getting too far down into the weeds here: this notation represents a potentially infinite series of iterations, in which n represents the current iteration. The result of the previous iteration — represented as z(n – 1) — is used as input for the next. The variable c is assumed to be a complex number, meaning it can be plotted onto the X and Y axes of a Cartesian coordinate system.)
The Mandelbrot set produces the most famous and immediately recognizable visual pattern in the field of fractals.
At this point I’d like to give you a chance to explore the Mandelbrot set for yourself, via the little program embedded below. You can pan around by holding down the left mouse button and dragging in the desired direction; drill down deeper into the image by holding down the right button and dragging to select the region you’d like to magnify; and undo each step you’ve previously made by clicking the middle button. Notice how this infinite space is made up of similar patterns that are nevertheless never quite exactly the same.
Update: as of January 10, 2023, the website which originally hosted this toy is no more. What you see below is an archived version. You can still play with it, but it may be a bit slow to load, and the source code is no longer available.
As useful and oddly beautiful as fractal lines like the ones produced by the Mandelbrot set can be, they’re only the beginning. Fractal planes are also possible, existing in some fractional limbo between two and three dimensions. Just as a plane is in this formulation an infinitely complicated line, a solid is an infinitely “wrinkled” plane. And just as fractal lines are perfect for representing coastlines and leaves, fractal planes can easily become mountains. The striking pictures shown below were computer-generated by Richard F. Voss in 1983 using only fractal equations.
It’s also possible — mathematically, that is — to go further, to produce fractals that strain toward a fourth (or higher) dimension. However, representing them becomes a problem given our three-dimensional world. We can only show a three-dimensional slice of these fractals. Doing so produces some strange shapes indeed, like the “dragons” of Alan Norton.
Benoit Mandelbrot published his magnum opus, The Fractal Geometry of Nature, in 1982; it still remains the definitive book on the subject. It’s a strange sort of mathematics text, in which Mandelbrot devotes at least as much space to philosophical digressions into the implications of his discoveries as he does to proofs and equations, and that much space again to lots of beautiful color slides of the fractals themselves. With the book’s publication, the theory of fractals became an example of a rare phenomenon indeed: a development in abstract higher mathematics that was taken up and trumpeted widely and excitedly by not just the likes of Scientific American (which devoted many, many articles to the subject) but also mainstream magazines, newspapers, even television broadcasts. As is all too typical of any new media plaything, fractals were hyped as useful for everything short of curing cancer — and I suspect that some wouldn’t put that past them either. Not only could physicists use them to understand the motions of molecules and biologists to understand the growth of plants, but some researchers claimed that fractal music was a possibility, while others claimed that they could help you get rich by revealing the “hidden patterns” that govern the stock market. In a 1984 article for Byte magazine, Peter R. Sørensen waxed effusive: “their uses range from physics, biology, and sociology, to art and even motion-picture scene simulation.” While I’m hardly qualified to speak to their uses in the former three scientific disciplines, I do feel fairly confident in claiming that they have had the greatest impact on those latter two, mushier categories, especially if we preface “art” with the word “computer” and allow computer games to slip in under that label.
Tellingly, Mandelbrot was employed at IBM’s Thomas J. Watson Research Center at the time he wrote The Fractal Geometry of Nature; all of the images found within it were generated by the computers there. Many fractal equations, including some of the most beautiful, aren’t complex at all in themselves, but the need to iterate through them so many times to produce their solution sets means that the science of fractals could exist only as a theory without the aid of computers. The first primitive computerized fractal visualizations had been produced by Robert W. Brooks and Peter Matelski in 1978, just three years after Mandelbrot first codified his theories. Virtually from that time forward fractals were inseparable from the computers needed to properly generate and study them.
Many people got their first glimpse of fractals before the hype started in earnest, in a remarkable sequence of computer-generated special effects included in 1982’s Star Trek II: The Wrath of Khan.
The creators of the sequence were the Lucasfilm Graphics Group, a collection of software and hardware engineers whom George Lucas had started putting together in 1979. These folks, whose numbers included such soon-to-be computer-graphics legends as Alvy Ray Smith, designed and built all of the hardware and software used to create sequences like this one. In 1986, the Graphics Group would be spun off to become an independent company whose name you might know: Pixar.
When Lucasfilm elected to start a computer-games division shortly after the sequence shown above was created, it was natural for it to draw inspiration, technologies, and even personnel from the Graphics Group. One of Lucasfilm’s first two games, Rescue on Fractalus!, proudly bears its graphical underpinnings and its status as the first game to make significant use of fractals right there in its title. (By way of continuing the theme, the alien enemies you fight are called the “Jaggies,” a reference to the ugly pixelized artifacts that are one of a computer artist’s worst enemies.) The game sparked great interest from its first press preview in May of 1984, thanks to its endlessly varied mountainous terrain, generated on the fly using fractal algorithms.
In the wake of Rescue on Fractalus! fractals were suddenly everywhere in the computing press; I’m not sure there was a single magazine that didn’t publish at least one big feature article on the subject over the next few years. The appeal of fractals to programmers was obvious, and had little to do with Mandelbrot’s esoteric philosophies about geometry and nature. Irregular, natural-looking landscapes had previously been dauntingly hard to craft for games — hard not only because they were a pain to draw but also and predominately because they were so un-amenable to compression algorithms and, indeed, to virtually all of the many innovative techniques programmers had discovered to store graphics data using minimal memory. Thus the distinctly regular, blocky graphics that were the norm, and the shock that was Rescue on Fractalus! Much like the Fibonacci-derived galaxies of Elite, fractals let programmers create whole landscapes generatively, from only an equation and a few seed numbers. They had their limitations, not least the amount of processing power that had to be allocated to generating them — Rescue on Fractalus! runs at all of 5 or 6 frames per second — but for many applications they seemed like magic. It was not a coincidence that after 1984 virtual landscapes started to become markedly richer and more natural.
But fractals did more than just make games prettier; they opened up whole new realms of possibility for them. The bit of video below, from a remarkable game that will be the topic of another article I’ll be publishing soon, may not look like much in contrast to some of what you’ve already seen today. Yet consider that, thanks to the magic of fractals, the planet being landed on is one of thousands to be topographically mapped in its entirety, and that you can land on it absolutely anywhere you like, zooming in from space to touch down where you will just as you can zoom and pan and explore the Mandelbrot set via the toy embedded into this article. Each of these unique worlds is generated using just a few numbers, a bare handful of bytes of precious memory.
Fractals aren’t quite the media darlings they once were; many other shiny objects have come and gone since the 1980s. And while they remain a valuable tool in many branches of science, they’re no longer viewed there either as the revolution they once were. You certainly don’t hear much anymore about fractal music or using fractals to play the stock market. Likewise, they’re now just another item in a game programmer’s bag of tricks. Still, they retain a fascination and beauty all their own. In that spirit, have fun in any further explorations you undertake, and if you discover any interesting patterns using my little toy above, or if you create any enhancements on the Studio Sketchpad site that hosts it, by all means let me know.
(As noted in the article proper, old computer magazines are an embarrassment of riches when it comes to information on fractals. Particularly good articles are found in the Byte of March 1984, September 1984, and June 1986; the 80 Microcomputing of December 1984; the Ahoy! of April 1987; the Amazing of March 1989, July 1989, October 1989, January 1990, June 1990, and April 1991; the Compute! of January 1983; and the A.N.A.L.O.G. of January 1986.
And since we’re in a multimedia sort of mood today, check out this song about the Mandelbrot set, sent to me by reader Rick Reynolds.)
There are at least two stories to tell about the way that George Lucas’s Star Wars movies changed the world. One is the tale of the impact the films themselves had on the culture of movie-making and movie-going. For better or for worse, the first Star Wars film ushered out the brief New Hollywood era of auteur-driven American film-making that had followed the collapse of the old studio system in the 1960s, whilst ushering in, with a strong assist from Lucas’s buddy Steven Spielberg’s Jaws, the era of the action-packed, escapist blockbuster that still persists to this day. And of course for the nerdier culture of 1980s gaming Star Wars became nothing less than a third great holy text to join Star Trek and The Lord of the Rings, which does much to explain why it keeps showing up around these parts.
Yet there’s also another Star Wars story that’s less appreciated. During the production of the first film and especially when the millions began to pour in after its release, Lucasfilm, Lucas’s production company, forever changed many of the technologies behind media creation and consumption. To say they did so “quietly” would be overstating the case. Some of the names associated with the technological side of Lucasfilm — Industrial Light and Magic, THX, Pixar, Skywalker Sound — are well known to just about everyone. But the actual nuts and bolts of the new developments, even of those pieces that hide behind one of those big names, can be difficult to appreciate for anyone who isn’t a professional working in one of the industries whose practices they revolutionized. I don’t propose to tell the full story of Lucasfilm the technology incubator here. (That’s actually already been done, and much better than I possibly could at that, in Michael Rubin’s Droidmaker.) I do, however, want to tell you about what it meant to the world of computer games. Like a surprising number of things at Lucasfilm, game development just seemed to happen of its own accord, as something the guy who made Star Wars just really ought to be involved in. It wasn’t initiated by Lucas or any of his cronies, but rather by Atari, who came for a visit in 1982, just as both companies were at the peak of their power, wealth, and influence.
That meeting would mark the beginning of Lucasfilm’s direct association with computer games, but their association with computers in general stretches back considerably farther, to the immediate aftermath of Star Wars‘s release and massive success. Made using traditional mechanical, analog techniques — scale models, stop-motion photography, blue-screening, etc. — Star Wars had been an exhausting film to shoot and edit, so much so that it had sent Lucas to the hospital on one occasion with a stress-induced panic attack. With plans already afoot to make many more films in the series, he was naturally eager to find ways of easing the burden of unglamorous, mind-numbing labor that still was film-making of any stripe — much less a special-effects-driven science-fiction epic — in the 1970s. He started collecting talented computer people to help with that process. This collecting was made much easier by the fact that Lucas, who had evinced a visceral loathing for the Hollywood machine since the days when their trade unions had denied him work as a camera operator whilst he was still a student, had chosen to center his film-making operation in Northern rather than Southern California, much closer to Silicon Valley than to Hollywood. He wasn’t much interested in computer-generated graphics in the beginning. For a guy like Lucas, who had never darkened the door of a computer-science department in his life, the notion barely existed. What he really wanted was a way to do optical printing — the overlaying of separate shots onto one piece of film — more easily and without the degradation that resulted from analog techniques; a replacement for the huge, hot, noisy machines that editors had to use in conjunction with razor blades, glue, and thousand-page notebooks to — literally in those days — cut movies; an easier way to mix sound; even a good accounting system to keep track of all his millions. The people he found to help with all that and much more would set the world on the path to a digital revolution in filmmaking, creating amongst other things the predecessors to modern digital-compositing software and video-editing programs.
More relevantly to our purposes today, however, his dragnet also scooped up some of the best pure computer-graphics minds in the country — people like Ed Catmull and Alvy Ray Smith, the eventual co-founders of Pixar. Soon joining them was one of the pioneers of fractal graphics, Loren Carpenter, whom they lured away from of all places Boeing after he brought down the house at the 1980 SIGGRAPH computer-graphics conference with the short film you see below.
Brilliant as they were, the little Lucasfilm Graphics Group that coalesced around Catmull, Smith, and Carpenter didn’t often have a lot to actually do in those earliest years. The second Star Wars movie, 1980’s The Empire Strikes Back, made no use of computer-generated imagery. Raised on traditional film-making techniques, Lucas could see the value of other new computer-based technologies that made things like editing and compositing easier, but wasn’t quite sure what to make of images that were born inside a computer. And so, like a number of research groups at Lucasfilm, Graphics tinkered away in benign neglect, refining their techniques and waiting for their big break, showing up at the occasional conference with something amazing, which prompted a steady buzz in magazines like Byte about the groundbreaking things that were apparently happening somewhere within the secret bowels of George Lucas’s Star Wars empire.
Their big break ironically came not from Star Wars but from that other big science-fiction franchise Star Trek. Industrial Light and Magic had been hired by the producers of Star Trek II: The Wrath of Khan to do many of the special-effects shots for that movie. This they mostly accomplished using their traditional models, composites, and stop-motion photography. But there was one big effect, an animation illustrating a terraforming “Genesis device” on an in-film computer screen, that stumped them. The script read simply, “And then the planet transforms… EFX sequence here.” It amounted to a creative blank slate for the Graphics Group to show the world what they could do — or perhaps to show just one man. “This is a sixty-second commercial to George Lucas,” Alvy declared, “to show him what he’s got.” They threw it all in: fractals, 3D modeling, texture mapping, fluid animation. Steven Spielberg, who tended to drop by the Graphics Group far more often than Lucas, loved it, saying it was a “great time to be alive” in a world that had such wonders; Star Trek II producer Harve Bennett was elated; Lucas himself called it a “great camera shot,” which by his laconic standards was gushing praise. Released in June of 1982, Star Trek II became in a sense Pixar’s public debut, thirteen years before Toy Story. Lucas was impressed enough to give Graphics some work to do on the third Star Wars film, 1983’s Return of the Jedi, although they weren’t given the opportunity to make any showstoppers like their Genesis sequence. Undaunted, a prescient Catmull insisted, “We’re going to be making entire films this way someday. We’ll create whole worlds. We’ll generate characters, monsters, aliens. Everything but the human actors will come out of computers.”
In the meantime, there were suddenly games. It was when the Graphics Group had just finished the Genesis sequence that Atari came to their nondescript offices in San Rafael, California, for a visit. Flush with even more cash than Lucasfilm at the time, Atari had quite a variety of research projects in progress, even if they would prove remarkably awful at turning them into finished products to supersede the aged Atari VCS games console. Thus it was natural for them to want to visit another company’s cutting-edge graphics research facility and see what they were up to. The two companies were hardly strangers; Atari had just released a licensed VCS game based on Raiders of the Lost Ark, Lucas’s hit cinematic collaboration with Spielberg.
What Atari saw in San Rafael blew them away. The delegation bombarded the bemused Catmull, Smith, and Carpenter with a million questions about how their stuff worked and, most importantly, how it might be adapted to videogames. Only Carpenter showed any real interest at all in such an endeavor. They were used to working on big workstations and minicomputers, not primitive micros, which they viewed with a certain contempt, dubbing their programmers mere “bit twiddlers.” They had high standards for their visuals: their graphics had to be good enough not to look out of place projected on a huge movie screen surrounded by other imagery shot on pristine 35-millimeter film. Their greatest enemies were what they called the “jaggies,” visibly blocky, pixelated areas that tended to lurk at the margins of what should be smooth, flowing curves. While the jaggies could be held at bay using the state-of-the-art, processing-intensive anti-aliasing techniques that the Graphics folks had spent years developing at Lucasfilm and elsewhere, those techniques weren’t much applicable to an 8-bit games console. Catmull and Smith at least wanted no part of that action; Carpenter was intrigued but also ambivalent, certainly not willing to entirely give up his film work for a game project.
Yet Atari persisted. Even Manny Gerard, the Warner Brothers executive who had orchestrated that company’s purchase of Atari, got involved, saying that “we ought to be in business” with Lucasfilm’s Graphics Group. Finally Atari offered to flat-out give Lucasfilm $1 million to set up a Games Group, for the products of which Atari would receive “right of first refusal” as publisher. An offer like that was hard to refuse. The deal was announced at the Summer Consumer Electronics Show in June of 1982. “We’ll be developing new forms of electronic entertainment,” said an excited Atari, “and the term ‘electronic entertainment’ is carefully chosen.” Don’t, in other words, just call them yet more videogames. Lucasfilm and Atari would continue to work hard to cultivate this rarefied image of Lucasfilm Games as artists of interactivity rather than mere game programmers over the months and years to come.
Had Atari not so aggressively forced their hand, one might be tempted to characterize the whole undertaking as something of a bait and switch on the part of Lucasfilm; no one currently in that Graphics Group whose work had so impressed Atari was earmarked to start working full-time on games. Still, the initial organization of an entirely new Games Group was tossed into the lap of the Graphics Group’s Ed Catmull, who found a fellow named Peter Langston at a Wall Street law firm to head it. Langston was a Unix hacker from way, way back who in 1971 had written Empire, an elaborate multiplayer strategy game played on a global scale, a forerunner to Civilization and other games of its ilk. (A later offshoot of the original game was even named Civilization.) An accomplished musician, he was very interested in the application of music theory to computers and vice versa and was, in the judgment of Catmull, just a visionary “star” in general, the perfect guy to take Games off his hands. When Langston, who was quite happy in New York, proved reluctant, Catmull continued to sweeten the deal, going so far as to offer to fly him back to New York for a couple of weeks out of every month if he liked. It was, once again, an offer that was hard to refuse. Catmull got his guy, and with a sigh of relief turned his attention back to film graphics.
Peter Langston surrounded by typically artsy trappings in his office.
Langston was an unusual choice for leader and administrator, a conceptual rather than an altogether practical thinker with a somewhat dreamy disposition. Working at the unhurried pace that would be typical of the young Games Group, he put a little team together to join him. He held the zap-em blast-em world of typical videogames in little more regard than did his colleagues in Graphics, and thus purposely avoided programmers with a lot of experience in the industry; he was after people like him, people who were “a little bit visionary.” He hired one David Fox because he admired Computer Animation Primer, a book Fox had recently written for which he had actually met and interviewed many members of the Graphics Group. It took him weeks more to settle on a very eager David Levine as a third team member. His biggest claim to fame was having designed the first add-on graphics board for the original Altair kit computer, which Langston thought was great. However, he’d also already done quite a lot of videogame programming, which maybe wasn’t so great. In the end Langston decided to give him a shot in spite of his surfeit of experience. A fourth employee, Charlie Kellner, late of the Apple Macintosh development team (another musician, he had programmed the pleasant little beep the Mac made at startup), would arrive still later. Like Graphics and many other teams benefiting from the Star Wars millions inside Lucasfilm at the time, Games walked a shadowy, largely unsupervised line somewhere on the intersection of a pure research group and a commercial proposition expected to deliver actual, tangible products. Certainly Games was nothing like the quickie projects being started by many other big companies to cash in on the videogame fad. Nor would they try to trade on the Star Wars name. In fact, they wouldn’t even be allowed to make Star Wars games.
Whilst negotiating with 20th Century Fox the deal that would lead to Star Wars, George Lucas had agreed to license the right to make Star Wars “toys and games” to Kenner Products. At a time when videogames still largely meant Pong, everyone interpreted “games” in this context to mean board games, which would be issued by Kenner’s subsidiary Parker Brothers, whose stable already included family perennials like Monopoly, Risk, and Clue. When videogames exploded a year or two later in the wake of Space Invaders (whose popularity was itself fed by Star Wars and the craze for all things science fiction that it engendered), Parker Brothers found themselves gifted with a golden goose for the ages, as was amply proved when they released an Atari VCS cartridge based on The Empire Strikes Back that spent many months in the top ten. The Star Wars videogame-licensing rights would be tied up in knots for years to come, the interested parties including not only Lucasfilm and Parker Brothers but also Atari, whom the latter agreed to allow to make standup arcade games and eventually console and computer games of their own under the name. (One way or another, everything involving videogames in the early 1980s always seemed to come back to Atari.) The end result was a circular tangle the likes of which only corporate America could create. Lucasfilm, the owner of Star Wars, had a games division that wasn’t allowed to make Star Wars games, while Atari had such a license thanks to kicking some money back to board-game manufacturer Parker Brothers, but chose to create those games in-house, even though they’d just paid Lucasfilm $1 million to set up the aforementioned games division for the purpose of making games for them. The upshot, however, was simple: no Star Wars for Langston, Fox, and Levine, nor for that matter for any of the others that would join them over the course of the rest of the decade. Many who were at Lucasfilm Games during this period have since remarked on what a blessing in disguise this really was, forcing the developers as it did to come up with original game concepts, original game fictions.
Even if they were barred from working directly with the Star Wars intellectual property, it was a damn good gig just to work for Lucasfilm, flush with cash, with all of the best equipment, with few or no hard deadlines, and right there on the close periphery of where the movie magic happened. If George Lucas himself seldom poked his head in the door, there was every day the possibility that he would. And there were other famous faces who were a more common sight, like Steven Spielberg, who spent quite some hours in between Industrial Light and Magic effects shots for Indiana Jones and the Temple of Doom camped out in the Games Group’s offices playing the Star Wars standup arcade machine Atari had been kind enough to send them. One day Spielberg recruited the team to become the screams of dying bad guys in Temple of Doom. Whatever else competing games studios might offer, they couldn’t offer experiences like these.
Of course, being constantly in the shadow of Star Wars and Indiana Jones could also be exhausting in its own way. David Fox tells the story of attending an early trade convention where a showgoer, excited by the games on offer, for once failed to make the connection, asking what else Lucasfilm had made: “I said, ‘We made the Star Wars films.’ Watching the look on his face was hilarious as he made the connection. And it was nice for once to not be in the shadow of the films!”
As for what they were actually working on in those early days… well, that wasn’t always entirely clear, at least for those looking in from the outside. They had unilaterally decided quite quickly not to develop their first games for the technologically antiquated Atari VCS games console, as Atari had anticipated, but rather to target the Atari 8-bit line of home computers and the new, more advanced Atari 5200 console that had largely the same internals as those computers — this even though neither platform was selling in anything like the quantities of the old VCS. Langston pronounced himself “amazed” when his research into the ways that games were typically made for home computers revealed most developers to be working “in a basement on a system with too little memory, too few floppy disks, no reasonable way to make a backup, and few if any debugging tools.” Determined to change all that, the trio spent months on an elaborate development system which ran on their big DEC VAX minicomputer. Conceptualized largely by Langston himself, it allowed them to write code in a LISP-like script, compile it, download it to an Atari 8-bit machine, and debug it as it ran there from their terminals. Similar systems, eventually capable of compiling down to a whole range of other microcomputers, would remain the core of the Games Group’s development methodology for years to come.
Their efforts to create actual games slowly coalesced around two projects, both of which had started as “throwaways,” learning exercises to work the kinks out of their development system and help the team gel. (That very description says much about Peter Langston’s academic style of management, and why Lucasfilm’s own management would in time start to find it kind of infuriating.) David Levine’s baby Ballblazer was the more artsy of the pair, a surreal 3D soccer game played between opposing spacecraft, with realistic if otherworldly physics and, most impressively, jazzy generative music provided by Langston that’s sometimes been compared to that of John Coltrane. The music was created using fractal algorithms pioneered by the Graphics Group that just seemed to be in the air in those San Rafael offices. Speaking of which: David Fox’s Rebel Rescue got a huge assist from Loren Carpenter, the only real games fan in the Graphics Group, who helped him to implement an admittedly jaggies-replete fractal landscape for his more grounded game that cast the player as a futuristic search-and-rescue pilot, trying to rescue downed pilots from a planet’s surface and bring them back to the mothership whilst fighting off hordes of invading aliens that are still swarming the atmosphere. Deciding they might as well have a sense of humor about the thing, they decided to call these aliens the Jaggies. Rebel Rescue would actually have made a darn good Star Wars game; its inspirations, including not only the name itself but also the X-Wing-like spacecraft you fly and the pilots you rescue in their distinctive orange flightsuits, are pretty hard to deny.
George Lucas, for whom games held little personal appeal, sat down with his Games Group exactly once during their first couple of years of existence. And yet his single visit had a huge impact on Rebel Rescue. The little team, idealistic as they were, took pride in the fact that both of their games were basically nonviolent. In Rebel Rescue you could avoid the enemy aliens or, if you were skilled, trick them into flying into mountains, but you couldn’t shoot at them. “Where’s the fire button?” Lucas asked. Fox explained. “Is there no shooting because of gameplay reasons or philosophical reasons?” Philosophical. “Great. Put in a fire button. I want to shoot at things.” It was also Lucas who suggested what would prove to be Rebel Rescue‘s second most memorable feature after the fractal terrain itself: some of the downed pilots are aliens in disguise, whom you have a split-second to zap before they kill you. This was deliberately left out of the game’s manual, thus nearly giving many players a heart attack when it first happened a few levels in.
Even as the team tinkered away with Ballblazer and Rebel Rescue, the Great Videogame Crash of 1983 was happening outside their ivory tower. Nevertheless, at year’s end the Games Group at last delivered working prototypes of both games to an Atari who had declared themselves bloodied but determined to fight on, who were “going to reignite the consumer’s love of videogames.” Just weeks later the Games Group was horrified to see their babies spreading like wildfire across the worldwide network of pirate BBS systems. It seemed that someone at Atari hadn’t been able to resist sharing these cool new games with a friend or two, and the thing had just exploded from there. Soon copies of the games started to show up for sale in flea markets and the less scrupulous software shops, decked out in homemade packaging invented by enterprising quick-buck artists. A still buggy Rebel Rescue in particular seemed to be in the collection of every Atari 8-bit owner on the planet, one of the most popular games on the system. Lucasfilm and Atari had a hit on their hands, but it was a hit they weren’t getting paid for. It was questionable whether Langston and his idealistic cohorts were more upset about the potential purchasers it was costing them or the fact that the games everyone was playing weren’t finished yet.
Most of the Lucasfilm Games Group, mid-1984: Charlie Kellner, David Levine (seated), Peter Langston, David Fox, Loren Carpenter (visiting from Graphics), Gary Winnick
Lucasfilm was understandably less than thrilled at Atari’s failure to protect their games. Atari, however, was also less than thrilled with Lucasfilm. It had now been eighteen months since their $1 million investment and, especially in light of their straitening financial circumstances, they wanted to see some finished, polished games in return. They were particularly unhappy that the Games Group had failed to deliver on a promise to give them something ready to show at the 1984 Winter CES. By now Lucasfilm management as well had decided that something had to give. In January of 1984, they hired one Steve Arnold to join Peter Langston as an awkward sort of co-manager. Arnold came from Atari, where he had been responsible for the Atarisoft line of ports of standup-arcade games to home computers, one of the few financial bright spots at the company during 1983. He’d rolled out an astonishing 49 separate conversions in five months as head of Atarisoft, so he certainly knew how to ship product. But a few Lucasfilm higher-ups wryly noted that his best qualification could be his PhD in psychology, or perhaps the time he’d spent years before as program director at a boy’s camp. Hopefully he could find a way to make the unruly Games Group toe the line without spoiling what made them unique in the first place. Also coming aboard around this time were two more team members. One was Noah Falstein, an established videogame designer and programmer late of Williams Electronics, where he’d worked on standup arcade games like Sinistar. The other was the Games Group’s first full-time visual artist, a veteran commercial artist named Gary Winnick.
With these new, somewhat more practical-minded additions, the Games Group did indeed start making progress more quickly; the new can’t-miss-it deadline promised to a still skeptical Atari was now Summer CES in June. Rebel Rescue was renamed Rescue on Fractalus! largely for legal reasons, to make it clear that it was not (officially) a Star Wars game despite the Lucasfilm logo on its box. The folks at Industrial Light and Magic built and photographed model spacecraft for the boxes, designed and built a cockpit model for the “Valkyrie” fighter the player flew in Rescue on Fractalus!, and even made flightsuits for the entire development team to wear in a grand photo spread. David Fox got to play the starring role, as a weary pilot trying to straggle home in his battered Valkyrie on the back of the Rescue on Fractalus! box. The whole effort cost at least $30,000. Yes, working for the company that made Star Wars did have its perks.
Final versions of both games, complete with packaging, were delivered to Atari well before the latest deadline. Lucasfilm held a lavish press conference to unveil them in May, presenting the games via slick videos created with the aid of professional voice actors and Lucasfilm’s general movie-making know-how.
A beleaguered Atari determined to press on (or still in denial) came to Summer CES with a new slogan: “The Day the Future Began.” The “Atari-Lucasfilm” games got a very positive response from press and public alike, and the partners put on the final touches for a July release. Yet practical questions still surrounded them. The Atari 5200 console had proved to be a flop, had already been discontinued, while Atari’s line of 8-bit home computers was still on the market but overshadowed by the cheaper Commodore 64. Meanwhile Atari themselves were still in financial free fall. And then, overnight, everything changed once again.
On July 3, 1984, Warner Communications announced that Jack Tramiel, late of Commodore, was buying Atari’s home-videogame-console and home-computer operations, surprising no one more than the people inside Atari who suddenly had their company sold out from under them. Looking on from the outside as Tramiel axed employees by the thousands in the weeks that followed, the Games Group wondered if July 3 should be called “The Day the Future Ended.” No one seemed to quite know in the midst of all the chaos where it left the Lucasfilm/Atari partnership. Tramiel himself didn’t seem to know much about their companies’ agreement and didn’t much seem to care. And so Steve Arnold set off to try to free the games from Atari’s exclusive clutches. He returned from his one and only meeting with Tramiel with a less than positive personal impression, saying that the latter reminded him of no one so much as Jabba the Hutt of Return of the Jedi fame. Sure enough, pictures of the two were soon hanging up around the offices of a very frustrated Games Group: “The Hutt Brothers: Jabba, Jack.” (Poor Jack just couldn’t win; the common comparison inside Atari itself was to Darth Vader.) But in this one case at least Tramiel’s bark was worse than his bite. Busy with other legal battles and the travails of rebooting Atari, he let Lucasfilm move on.
Arnold settled on Epyx as publisher out of a crowded field of suitors, signing a four-game deal that was announced with considerable fanfare at Winter CES in January of 1985. Ballblazer and Rescue on Fractalus! became widely available for purchase at last shortly thereafter. These, the Games Group’s first actual products, had taken two-and-a-half years to come to fruition, an eternity in an era when most videogames were still churned out in a matter of a few months.
Thankfully for the Lucasfilm brass, it looked likely that the next games wouldn’t be so long in coming. Peter Langston had bowed out at last in the fall of 1984, taking with him his rather abstract approach to game development and freeing Arnold to continue to refine the Games Group’s operations along more practical lines. And there was suddenly plenty of practical work to do. No longer beholden to Atari, they were now free to port Ballblazer and Rescue on Fractalus! from the fading Atari 8-bit line, every owner of which seemed to already have them anyway thanks to the leaked demo versions, to stronger platforms like the Commodore 64, where they ended up selling far more copies. Rescue on Fractalus! in particular became a hit, not a blockbuster and certainly not enough to justify the time and money poured into it absent Atari’s initial $1 million beneficence, but a solid piece of groundwork that established Lucasfilm Games as a maker of classy but accessible action fare. If it seemed just slightly underwhelming in light of the years it had been in production and all of the flashy promotion that surrounded it — seemingly every magazine in the industry published a big Lucasfilm Games feature article around this time; such was the cachet of the house that Star Wars had built — well, just about any game realizable on an 8-bit computer would have. It didn’t help that the year that had passed between the leak of those demo versions and the arrival of the finished games on store shelves had allowed lots of other programmers to start experimenting with fractal graphics, making Rescue on Fractalus! look far less revolutionary than it otherwise would have.
The next two games were very much designed to build on the technical as well as the commercial groundwork laid by the first two; both started with the graphics engine from Rescue on Fractalus!. Designed by Charlie Kellner, a newcomer who had been hired just before Langston left, The Eidolon had the player piloting a machine through networks of underground tunnels inside the protagonist’s own mind — fractally generated, naturally — full of dangerous “guardians of the id.” Koronis Rift hewed still closer to Rescue on Fractalus!: this time you were flying above an alien-infested, fractally-generated planet trying to collect technological relics rather than downed pilots. Both were once again well-reviewed when released in late 1985 after comparatively reasonable one-year development cycles, going on to sell modestly well, if not to match the sales of Rescue on Fractalus!. In the span of 1985 the Games Group had increased their catalog from zero to four solid action games, one of them a genuine hit, and were now largely self-sustaining.
At the same time, though, a slight sense of underachievement clung to the Games Group, who had failed to completely deliver either the revolutionary gameplay experiences for which Peter Langston had been hired or the blockbuster sales figures one might expect from the company of Star Wars. They were still something of an odd duck in the industry, their huge cachet still largely based on that name on their boxes rather than the actual contents of the disks therein. Yet, even after Peter Langston’s departure, the sense of artistic idealism he’d worked so hard to engender remained alongside Steve Arnold’s determination to actually ship games on a semi-regular basis. “I think in general we’ll be moving away from the concept of games,” said Kellner shortly after the release of The Eidolon and Koronis Rift, echoing some of the verbiage Atari and Lucasfilm had used when first announcing the new venture three-and-a-half years before. “We’re trying to produce an experience that’s like being part of a film, rather than just being part of a game.” The fact that they were still having to promise to move beyond mere “game” in the future could be read as an admission that visionary software had proved to be a bit more difficult to develop than expected. On the bright side, their next project would be by far their most audacious and, yes, visionary yet.
(Sources: the book Droidmaker by Michael Rubin; Byte of March 1984; A.N.A.L.O.G. of August 1984, April 1985, August 1985, March 1986; Antic of August 1984, December 1985; Commodore Power Play of August/September 1986, October/November 1986; Compute!’s Gazette of August 1985; Compute! of August 1982, November 1982; Creative Computing of March 1982; Enter of September 1984; Family Computing of August 1986; Game Developer of December 1994; K-Power of September/October 1984; Zzap! of February 1986, March 1986; Retro Gamer 27, 44, 116; the website LucasFans, now available only via the Wayback Machine; Peter Langston’s paper on the early Games Group’s development system, available from his website.)
Shortly before the departure of Peter Langston from the Lucasfilm Games Group, he and Steve Arnold hired one Chip Morningstar as a tools programmer. The latter was coming off a stint spent working for Ted Nelson, a sort of philosopher of information management who had coined the terms “hypertext” and “hypermedia” back in the 1960s. Morningstar, who says that “few people could actually work with Ted for more than a day or two without becoming clinically insane,” was initially thrilled to just hunker down making practical tools for game development on the Games Group’s Unix workstations.
But then, on one of his final days at the office, Langston tried to interest everyone in a joint playthrough of his own old multiplayer grand-strategy classic Empire, for the purpose of, according to Morningstar, “getting us to think about alternative modalities of game design.” They all played for many hours and then, with Langston’s encouragement, picked apart what they liked and didn’t like about the experience. Morningstar found he didn’t much like the very thing that seemingly made Empire a game: the cycle of build-up, increasing conflict, and the final triumph of one player — or, more frequently, the final nuclear annihilation of all of them. He would prefer it if “the world was much bigger and the player goals more open-ended.” He wrote up a proposal for an online portal to host a “10,000-player computer game” that showed the influence of Ted Nelson’s visions of networked consciousness as well as the cyberpunk science fictions that he’d been reading a lot of lately — works like True Names by Vernor Vinge and William Gibson’s brand new, seminal Neuromancer. The poetic language he employed in his proposal rather self-consciously echoes Gibson’s own descriptions of cyberspace.
Picture, if you will, a network, an intricate web of knots and threads spanning thousands of miles. The knots are machines, made of silicon, metal, and plastic. The threads are metal wires. It is a computer network. The machines are computers, sitting in homes, schools, and offices across the continent. The wires are telephone lines, tying the hundreds upon hundreds of individual processors into a single, unified whole. At each of these machines sit people. People of all kinds and all ages… they experience the signs and sounds of that which exists only in the wholeness of the web, and in their own minds.
Steve Arnold made a habit of filing away all of the ideas that poured out of his charges, even seemingly outlandish ones like this one. After all, you never knew what might be worth considering if the right partner and/or situation came along.
We haven’t had much occasion to talk about it yet on this blog, but PCs had always been seen by hackers as, amongst other things, tools of communication. One might even say it was in the PC industry’s very DNA. Well before Apple, for instance, Steve Jobs and Steve Wozniak had gotten their start as businessmen by selling “blue boxes” for purposes of “phone phreaking,” hacking the telephone system to make free long-distance calls. (The widespread application of telecommunications technology to illegal activities would also be an ongoing theme of the young industry.) The first modems were available almost as soon as the first PCs, using hardware designs adapted from bigger institutional computer systems.
To say that using a computer for telecommunications was challenging in the early days hardly begins to describe the situation. “Modem” stands for “modulator/demodulator.” It converts data into sound and transmits it over a phone line, the “modulating” part of the process. What sounds like an unholy racket to human ears can be converted — “demodulated” — back into data that a computer can understand by another modem at the other end of the line. A modem monopolized the phone line, meaning that unless you lived alone or could afford to spring for a second line, negotiations on its use with other householders were likely to be tense ones. Once made, modem connections were notoriously unstable. In many areas even the “call waiting” tone resulting from an incoming call would knock you off the line while simultaneously leaving you no way to answer the second call, a useless double whammy if ever there was one. And Modems were slow. A 300-baud modem, for many years the standard, could transmit or receive at about 35 characters per second, meaning a single line of text on an 80-column screen took well over two seconds to print, a full screen of 25 lines almost a full minute.
Assuming you were willing to put up with all these annoyances, whom should you actually call using your computer? The World Wide Web was still many years in the future, the Internet itself largely restricted to users at big universities and other research institutions, discussing esoteric subjects using specialized messaging software. Your realistic options could thus largely be divided into two categories: the so-called computerized bulletin-board systems, or BBSes, and, in time, bigger commercial dial-up services.
The typical BBS operator was a guy not that different from you who had elected to set up a little online presence with a spare computer — or a primary computer when he wasn’t actually using it — and a second phone line. Since these systems usually only had one modem and one phone line, only one user could actually be online at a time. Accessing a popular system was thus generally a matter of setting up a “war dialer,” a program to dial the BBS’s number again and again until the busy signal was replaced with a modem’s answering tone — all the while, of course, tying up your own phone line with the effort. Once online, you could read and post messages and upload and download software, within reason; you didn’t want to be That Guy staying online for hours and blocking everyone else out of the system. Indeed, popular systems often strictly limited your time to make sure that didn’t happen. Many of the BBSes were engines of software piracy. Thanks to phone phreaking, the necessary numbers and techniques for which were also widely distributed on pirate BBSes, pirates could call systems all over the country and the world, while more legitimate users unwilling to spend a fortune in long-distance fees were limited to those in their own hometowns. BBSes, even — especially! — the legitimate ones, were thus by their very nature very limited affairs, bound to a certain geographic area and, with only 24 hours in a day and only one phone line, sharply restricted in how many users they could realistically support. Many old-timers today will tell you that that was a big part of their charm.
Still, there was obvious potential for larger services that could bring more people together. In the late 1970s one Bill von Meister had an epiphany that would establish the model for such services for many years to come. When corporate America shut down for the evening, he realized, millions of processing hours on time-shared institutional computers went unused, as did much of the telecommunications infrastructure that linked all of those machines together. He founded an online service called The Source to take advantage of this excess capacity by essentially selling it to ordinary computer users at home, who could access his system via modem using local phone numbers that served as entrance ramps to normally business-focused packet-switched networks like Tymnet and Uninet. After The Source was announced at a gala event in June of 1979 — special guest/paid spokesman Isaac Asimov was there to declare it to represent nothing less than “the start of the information age” — similar online services began to spring up in considerable numbers, their user counts to climb steadily. CompuServe, the second entrant in the burgeoning field and always the largest and most lavishly promoted, had 5000 subscribers in 1980; 145,000 in 1984; 300,000 in 1986; 460,000 in 1988.
What people did on these services was in some senses mostly the same as what they do on the Internet today. By mid-decade CompuServe offered discussion forums and chat rooms on a huge variety of topics; an online encyclopedia; electronic editions of newspapers and magazines; online banks, stockbrokers, and other financial services; worldwide weather forecasts; online shopping malls, airplane tickets, and hotel reservations; and of course your very own private email address. Games were also popular; amongst the expected slate of traditional board and card games one could find the occasional standout, like CompuServe’s MegaWars, an elaborate and addictive multiplayer strategy game played on a universal scale, not all that far removed from Peter Langston’s Empire. The first accredited online university, called simply The Electronic University, already had 10,000 students by 1985, offering seven degrees from an Associates in the Arts to an MBA. The mainstream media began to publish the occasional skeptical report on a new phenomenon known as “online personals,” noting with shock that some people had supposedly ended up marrying others that they had first met online. Indeed, the amount of sharing that took place online was a constant source of surprise to the unwired. An apocryphal tale made the rounds, winding up eventually even on PBS’s Computer Chronicles program, of a woman who had announced on a chat room that she had had enough, that she was about to kill herself right there and then, only to be talked down from the brink by her interlocutors. Less positively but more inevitably, politicians began to fret about online pornography and child predation, making it an explicit crime to use a computer to traffic in child pornography — it was anyway, but that’s politics for you — with the Child Protection and Obscenity Enforcement Act of 1988.
It’s easy enough to see these services, described as I just have in the abstract, as essentially equivalent to the World Wide Web of today. That, however, would be a mistake. I really need to emphasize just how limited and primitive online services of the 1980s were by modern standards. Without the bandwidth to send pictures and with few computers capable of displaying them with any fidelity at all anyway, the online services were made of nothing but text, laboriously transmitted page by page at speeds of 300 or at best 1200 baud. There were no hyperlinks, no mouse support, no color, no sound, no fonts, no page layout, no windows or columns, just walls of slowly printing monospaced monochrome text separated by menu prompts in the form of blinking cursors. And each of these services was a world unto itself: if you signed up for CompuServe, you could only use those facilities that CompuServe offered, could only chat or send mail to other CompuServe subscribers. The interoperability and interconnectedness that define the World Wide Web of today didn’t exist, making your choice of which service to splurge for a fraught one indeed.
Finally, these services were expensive, staggeringly so by modern standards. In 1985 it cost $40 just to sign up with CompuServe, then $6 per hour at 300 baud or $12.50 per hour at 1200 baud during non-business hours. As for usage during business hours, that was so expensive that you didn’t even want to think about it — which was, after all, kind of the point. The other services charged similar rates. Small wonder that avid users devised and shared a multitude of techniques, from command shortcuts for bypassing layers of menus to ways of collecting digests of messages for reading offline, to minimize their time spent connected. And small wonder as well that, despite the services’ steadily increasing user bases, for many years relatively few computer owners had the requisite combination of financial wherewithal and patience to make use of them. At mid-decade it was estimated that less than 5 percent of active home-computer users subscribed to one of the commercial online services, while less than 40 percent of them even owned modems.
Those numbers represented opportunity. That, anyway, was how they were seen by Steve Case, marketing director for a heretofore underwhelming would-be purveyor of telecommunications services called Quantum Computer Services. Case came up with a scheme that would address some if not all of the reasons that most users still stayed away from the big online services. His QuantumLink would work only on Commodore 64s. Rather than using generic text-oriented terminal software, it would be given away as a user-friendly, colorful, point-and-click-driven application that would nevertheless interface with Quantum’s online databases via modem. And, aware that the Commodore market was the most price-conscious in computing, Case made QuantumLink dramatically cheaper than any other service: a flat fee of $10 per month, plus $3.60 per hour for certain “premium services,” whether delivered at 300 or 1200 baud. The scheme did come with its drawbacks, like the fact that major additions to the service would require the mailing of a new disk to every customer, and of course the fact that it would be limited to Commodore users, but on the whole Case believed the advantages would far outweigh the disadvantages.
Critically, Case was able to convince Commodore themselves to sign on as sponsor and principal investor, making QuantumLink the “official” online service for Commodore owners. QuantumLink “telecommunications starter kits” were soon being inserted into every Commodore modem package, not to mention everywhere else Case could find to stick them. He was determined to make QuantumLink the easier, friendlier online service, suitable for the non-technical. “For the average guy,” he says, “we needed to offer the market something that was a little easier and cheaper and more useful.” Case’s “average guys” would sign up with QuantumLink to the tune of about 50,000 members one year after the service made its debut in November of 1985 — not bad numbers at all for a brand new, platform-specific online service that was starting from scratch.
QuantumLink’s point-and-click main menu.
An online chat on QuantumLink.
QuantumLink featured a fairly typical slate of offerings, including plenty of games. The Commodore 64 was after all the premier gaming computer in the country, and the QuantumLink software had actually been built from the remnants of an earlier, failed venture called PlayNet, an attempt to create an online service revolving exclusively around games. QuantumLink’s games were largely inherited from that effort, including old standbys like backgammon, chess, checkers, hangman, and Go, as well as an online casino that dealt in perks rather than money. There were also heaps of public-domain games to download and play offline, plus demo versions of commercial games. Still, right from the outset QuantumLink and Commodore looked for an online game that was newer, bigger, and more exciting, something the likes of which had never been seen before. It was this quest that brought Clive Smith, a vice president of strategic planning at Commodore, out to Lucasfilm Games one day in the summer of 1985, well before the QuantumLink service was planned to actually go live. Smelling a chance to do something unprecedented on someone else’s dime, Steve Arnold pulled Chip Morningstar’s old proposal for a networked virtual world out of his ideas folder. Somewhat to everyone’s surprise, Smith immediately loved it. He quickly secured for Arnold and Morningstar a chance to pitch it in person to Quantum at their headquarters in Vienna, Virginia.
For his presentation, Morningstar dropped all vestiges of an “outer space/conquer the galaxy” type of game like Empire, or for that matter CompuServe’s MegaWars, in favor of a game that “looked kind of funky and suburban,” a “game” in name only that would ultimately be all about “the social dimension — people interacting with other people.” It was an idea so unprecedented that even its originator had difficulty describing or even completely envisioning it. The closest analogues would be MUDs — “multi-user dungeons,” essentially elaborate online text adventures in the spirit of Adventure and Zork that could be occupied by dozens of players at a time. Yet, while there was certainly a strong social element at play there, most MUDs also placed a heavy emphasis on dungeon-crawling, monster-killing, and leveling-up, whilst taking place in exactly the sorts of fantasy or science-fiction worlds that Morningstar had so definitively rejected. And of course MUDs consisted solely of text, while Morningstar’s world was imagined from the start as a richly graphical environment. Groping for the right vocabulary to describe his ideas, he found himself coining new jargon of his own, some of which has persisted to this day. Most notably, he borrowed the term “avatar” from Hinduism, where it represents a deity’s earthly incarnation. Morningstar used it to describe the onscreen figure that each player — the god in the machine, the deity pulling the strings — would control. That said, her avatar wouldn’t quite be her, at least not all the time, for Morningstar’s world would amongst other things be a space for role play, a space for trying new identities on for size. Richard Garriott applied the same term to a game even earlier than Morningstar, during the development of Ultima IV: Quest of the Avatar, but his usage applies specifically to the ethical quest that forms the plot of that game: the player’s goal is to become an “avatar of virtue.” The more generalized use of the term in videogames is better attributed to Morningstar, even if Garriott’s usage probably better reflects its real implications in Hinduism. Whatever else you could say about “avatar,” it at least sounded more dignified than “puppet.”
The project, first dubbed Universe and then MicroCosm, was a hard sell. For some months Steve Case and the others at Quantum remained intrigued but uncertain. The contract wasn’t finalized and signed at last until December of 1985, after QuantumLink itself was already up and running. By this time the name had been changed yet again, this time to simply Habitat. Hoping to make a splash, the new partners rented the Palladium, the hottest new nightclub in New York City, to officially announce the project in mid-1986, despite the fact that there was still lots of work to be done before it could possibly go live. “It was quite an anomaly to have this essentially hip nightclub where people dressed in black… and then all these computer geeks showing off their multiplayer computer games,” admits Steve Arnold. “It was a little bit of a mismatch between PR positioning and target audience.”
Undaunted, Quantum continued to hype Habitat quite heavily in their advertising. It was perpetually “coming soon,” first in the summer of 1986, then in the fall, then at some undetermined point in 1987. An initial spate of intrigued articles in the Commodore trade press dissipated as the months went by and the project started to look more and more like vaporware. Through it all Chip Morningstar struggled to actually build his monster with the help of a partner, another recent Lucasfilm hire named F. Randall Farmer. Only gradually did it dawn on them just what they had gotten themselves into. Making Habitat come alive was going to be hard. Really hard. Farmer has called the Habitat project the most complicated single thing ever done with a Commodore 64. While he’s hardly unbiased, it’s also difficult to think of a more ambitious rival. To help you appreciate Habitat‘s scope, I’d like to give you a description of the experience it was meant to provide for the player.
When you signed up for Habitat, your first task must be to create your new avatar, choosing your sex, your hair color, the shape of your head and your facial features. Your avatar was then given a room of his own to live in, complete with a dresser for storing things and a cat for cuddling. (Yes, dog lovers were out of luck — and no, it wasn’t possible to kill your cat.) You commanded your avatar by moving the cursor about the screen and tapping the joystick button, which yielded a radial menu of four items: “go” (walk to where the cursor points), “do” (manipulate some object or machine to which the cursor points), “get” (pick something up), and “put” (drop something, possibly inside a container). You could, for instance, select “do” over the dresser to open a drawer and reveal its contents, which you could then manipulate via “get” and “put.” Your avatar could normally only carry one thing at a time, but you could use a container, like the handy sack seen in the screenshot above, to carry many more.
The discrete areas, called regions in Habitat terminology, were linked with one another to form a grid or map, like the individual rooms of a text adventure. This, however, was one huge adventure game. There were many thousands of public regions, plus a “Turf Sweet Turf” for every player. By “going” to the door of your home turf, you could access the grand world outside your avatar’s humble abode. It was there that you’d begin to meet others.
Due to technical constraints, only six avatars could occupy a region at the same time, but you could communicate with all of your region-mates simply by typing whatever you liked on the keyboard. The text you typed appeared as comic-style thought bubbles over your avatar’s head for all to read.
And that’s largely all you needed to know in order to interact with Habitat. Yet that’s enough to allow for a huge scope of social possibility, a fact of which Morningstar and Farmer were well aware and of which they were determined to take full advantage. Your room contained a telephone with a functioning telephone number other players could use to call you. Likewise, you could call them by looking them up in a virtual telephone book and dialing. If you preferred the dying art of letter-writing, every avatar had a mailbox before her domicile, with a functioning post system for delivering letters. There was a bank; every player got a stipend of 100 tokens every day she logged in, which she could spend as she would. There was a travel agency and resort destinations to which it could send you, stores to buy clothes and gadgets, bars and galleries and theaters. To facilitate large-scale events like plays, concerts, and poetry readings despite the six-avatars-per-region limit, Lucasfilm implemented something called “ghost mode,” which let you peer into a region without your avatar being actually embodied there — ideal for being a passive member of an audience. There was a “teleport” system for getting around the thousands of regions as quickly and easily as possible, and hotels for overnight trips to far-flung corners of the world. For those needing a definite external goal to work toward, an “Oracle” found in a park near the center of the world assigned adventure-game-like quests: “find the mystic orb of Xebop and return it to the Temple of Zak.” But even they were envisioned as social rather than solitary affairs, often requiring multiple avatars to pull off.
Perhaps the most amazing thing about Habitat is that it actually worked at all on a purely technical level, on client machines that Morningstar himself describes as little more than “toys,” with 1 MHz 8-bit processors, 64 K of memory, and an unstable connection running at as slow as 300 baud. The servers that housed this virtual world and managed all of the clients did have a few more resources at their disposal: QuantumLink, and thus Habitat, ran on a cluster of Motorola 68000-based computers built by a company called Stratus. Still, managing a virtual community proved to be far more expensive in both human and computer hours than anyone had anticipated. Lucasfilm’s contract with Quantum called for an environment capable of supporting 20,000 users, with the possibility of scaling it up easily to 50,000 if necessary. By the time the active user base reached 50 employees and insiders, Morningstar admits, Lucasfilm was already starting to feel “over our heads.”
We needed things for 20,000 people to do. They needed interesting places to visit — and since they can’t all be in the same place at the same time, they needed a lot of interesting places to visit — and things to do in those places. Each of those houses, towns, roads, shops, forests, theaters, arenas, and other places is a distinct entity that someone needs to design and create. Attempting to play the role of omniscient central planners, we were swamped.
Morningstar and Farmer nevertheless continued to plug away. In the first weeks of 1988 QuantumLink finally began a public beta test consisting of about 500 players who had been lucky enough to wind up with membership packets containing a special “early-access pass” for Habitat. It’s here that the Habitat story gets really fascinating, as Morningstar and Farmer bring the world’s first massively-multiplayer virtual community fully online. Many of the questions they were soon being forced to address, the situations they confronted, will sound familiar to anyone who’s ever played in Ultima Online, World of Warcraft, or Second Life, or for that matter just read about them. Here, then, are some dispatches from the front of an 8-bit Second Life that lived for just a few months in 1988.
From the bureau of things that never change: some users promptly became addicted, which was a real problem for them given that this was a paid beta test, billed at the usual QuantumLink “premium” rate of $3.60 per hour. Some were soon racking up monthly bills of $200 or more, corresponding to well over 50 hours of play. One managed to hit $1000 in one month, despite warnings sent to his email address at $300 and $600 that he might want to “check out his usage in the billing section.” Horrifying as this was on one level, Lucasfilm and QuantumLink couldn’t help but note that in theory they would only need twenty more users just like him to cover all of their operating expenses and make Habitat profitable.
Seeking to make Habitat a believable place, Lucasfilm included richer and poorer areas, discount shops and luxury boutiques with largely the same goods but very different prices. Trouble began when a few players realized that they could actually pawn items in a rich area for more money than it cost to buy them new in a poor. They spent an entire night trekking back and forth, buying low and selling high. By morning they had effectively wrecked Habitat‘s economy, inflating the money supply by a factor of five and making themselves almost inconceivably rich in contrast to everyone else. This nouveau riche began to usurp power for themselves, getting others to do their bidding for trifling (to them) amounts of money, dispensing bread and circuses to the masses in the form of games and treasure hunts. With little outlet for their immense fortunes in spite of all their best efforts to spend them, the superrich ended up establishing a lucrative trade amongst themselves in what became Habitat‘s most ostentatious symbol of conspicuous consumption: custom heads for their avatars.
Weapons could be purchased in Habitat and player-versus-player conflict was allowed in the so-called “wilderness areas” outside of city limits, although the consequences of “death” were relatively mild: everything the dead avatar was carrying would be dumped onto the ground where she had been standing, and she would then be teleported back to her home turf, once again intact. Of course, it took about five minutes for someone to start randomly shooting people in order to take their stuff. This led to…
Law and Order Must Be Imposed!
The inhabitants decided that a police department must be established, and held elections for sheriff. Inevitably, the guy who won by a landslide was one of the one-percenters, who could afford a campaign on a scale of which the other candidates could only dream. Did someone say something about the role of money in politics?
Christianity Under Siege!
The first virtual church was opened by a Greek Orthodox minister. Those who wanted to join his flock were forbidden from stealing or engaging in any sort of violence. Unfortunately, whenever the minister and his flock weren’t around other players would march in, strip the church bare, and pawn the lot. The minister finally had to appeal directly to Lucasfilm for a special dispensation: a lock for his church.
While there is no record of any relationships formed inside Habitat escaping into the real world, there were at least three virtual-world weddings, all taking place in that aforementioned church. Lucasfilm helpfully joined the newlyweds’ turfs together for cohabitation. The first virtual divorce followed the first virtual marriage by just two weeks.
It was possible for players to “sleep over” in other players’ turfs rather than their own, if invited inside. Soon con artists started finagling such invitations from naive players, then logging in while the victim still slept blissfully and absconding with everything in the room.
It’s About Ethics in In-Game Journalism!
A couple of enterprising players founded a newspaper, The Weekly Rant, consisting of as many as fifty pages full of news, fiction, classified advertisements, and announcements (including news of weddings and divorces). Absolutely everyone in Habitat was soon using it as an essential resource until, after a dispute about editorial content — the publisher wanted a shorter newspaper with less fiction — the editor abruptly quit. Habitat felt the loss keenly for all of its remaining days.
For a special area they were creating called “The Dungeon of Death,” two players convinced Lucasfilm to build them special “elephant guns” that could kill another avatar in one shot instead of the usual twelve or so, on the condition that they would use them only in the person of their alter egos “Death” and “The Shadow” who lurked within the dungeon. Embarrassingly, one day while playing Death on loan Randall Farmer himself managed to get himself killed by another player, who promptly scooped up the gun. An ordinary player, unbound by any strictures whatsoever, now had this massively destabilizing weapon in her hot little hands. After threats and negotiations, a deal was struck: 10,000 tokens to buy the gun back. Echoing a thousand Hollywood thrillers, the two parties met on the grounds of Habitat‘s largest public park to make a tense exchange through a neutral intermediary; one can’t help but imagine their respective posses lurking tensely in the bushes all around in case trouble started.
The immediate transplantation of real-world societal structures and, one might say, societal woes into Habitat might be read as depressing. On another level, though, it was a sign that Habitat worked, that it had become a real community. “In a real system that is going to be used by real people,” wrote Morningstar and Farmer later, “it is a mistake to assume that the users will all undertake the sorts of noble and sublime activities which you created the system to enable. Most of them will not. Cyberspace may indeed change humanity, but only if it begins with humanity as it really is.”
In time, Morningstar and Farmer came to categorize the inhabitants of Habitat into five categories that perhaps apply almost equally well to the real world.
The Passive: Easily 50 percent of the number of users fall into this category, but they probably use only 20 percent of the connect time (rough estimates). They tend to show up for events ad-hoc and when the mood strikes. This group must be led by the hand to participate. They tend to want to “be entertained” with no effort, like watching TV.
The Active: This group is the next largest, and made up the bulk of the paying user-hours. The Active user participates in two to five hours of activities a week. They ALWAYS have a copy of the latest paper (and gripe if it comes out late).
The Motivators: The real heroes of Habitat. The Motivators understand that Habitat is what they make of it. They set out to change it. They throw parties, start institutions, open businesses, run for office, start moral debates, become outlaws, win contests.
The Caretakers: The Caretakers are “mature” Motivators. They tend to help the new users, control personal conflicts, record bugs, suggest improvements, run their own contests, officiate at functions, and in general keep things running smoothly.
The Geek Gods: The operator’s job is most important. It really is like being a Greek God from the ancient writings. The Oracle grants wishes and introduces new items/rules into the world. With one bold stroke of the keyboard, the operator can create/eliminate bank accounts, entire city blocks, or the family business. This is a difficult task as one must consider the repercussions of any “external” events to the world. Think about this: would you be mad at “God” if one day suddenly electricity didn’t work anymore? Habitat IS a world. As such, someone should run it that has experience in that area. A Geek God must understand both consistency in fictional worlds and the people who inhabit it.
In the beginning, Morningstar and Farmer tried to micromanage the world, but it quickly became clear that this would be impossible, even with just 500 players. Designing a massively-multiplayer game is a fundamentally different discipline than designing for a single player. A player in a single-player game expects and deserves to have the world revolve around her; a player in a massively-multiplayer game is but one among many would-be heroes. The reality of this difference, a difference which Morningstar and Farmer were amongst the first people in the world to confront, became clear with their first attempt at a large-scale community treasure hunt, the so-called “D’nalsi Island Adventure.” Farmer spent weeks designing and building the quest and the 100-region island that would house its goal, the lost “Amulet of Salesh,” hidden away in a remote, seldom-visited corner of the world. After announcing the quest at a special community meeting held in the county courthouse, Farmer and Morningstar sat back to watch the players go to work, anticipating it would take them some days to return with the amulet. As it turned out, someone found the island within fifteen minutes, then recovered the amulet therein within eight hours. Most players were never even aware that the quest was happening before it was all over.
Clearly, Habitat could not depend on such externally imposed quests, quests which the most knowledgeable and the most powerful were always destined to win. What happened inside Habitat would instead have to be driven by the players themselves. By the time the beta ended, Morningstar and Farmer had shifted their thinking entirely, realizing that the success of this virtual world would depend on them being able to move enough users along each step of the continuum outlined above, from curious but Passive non-participants to Geek Gods who had the competency and the interest to start to play a major role in the underlying workings of the world itself. (One might also say that this is the main goal a real-world society has for each succeeding generation.) From the standpoint of game design, this would prove to be the number-one lesson of Habitat — a sort of transcendence of game design. Morningstar and Farmer made a conscious effort to begin to think more like facilitators than game designers. When players came to them asking for an elephant gun or a church, they asked them what they planned to use it for and, if all seemed kosher, did their best to provide.
And then, just like that, it was suddenly all over. The beta ended and Habitat went dark forever, leaving a million might-have-beens in its wake. Steve Arnold:
We found that we had pushed the C-64 to its limits, and, if we had to do it all over again — and I’m not trying to insult C-64 owners — we really would have developed the Habitat program on a different system. Once we finished Habitat and really sat down and looked at it, we realized there was a lot more technology needed to do multiplayer gaming — a lot more than we could do effectively on the C-64.
Habitat just didn’t scale well, wasn’t going to work with 20,000 or 50,000 players. In addition to the problems on the client side, Quantum noted that Habitat, even with just 500 players, had already consumed an inordinate amount of processing power on their servers during the beta period. They claimed that trying to deliver a full-scale Habitat would require a massive investment in infrastructure that they just weren’t in a position to pay for.
That said, there were doubtless other factors at play in the mutual reluctance of both Quantum and Lucasfilm to commit yet more resources to making Habitat live for real. They were now well into 1988, and it was becoming clear that the Commodore 64, a computing evergreen for so long, was beginning to fade at last. This reality left neither partner eager to commit more resources to expensive long-range projects for the platform. Lucasfilm was shifting to MS-DOS as their front-line development platform, while Quantum was now pouring most of the revenue they were earning from QuantumLink into new online services targeting other platforms rather than continuing to make major improvements to the old service.
Lucasfilm washed their hands of Habitat in mid-1988 by selling the whole technology package to Fujitsu, who used it to create Club Caribe, a stripped-down environment of a relatively compact 500 regions, on QuantumLink in 1989. It essentially functioned as an amusing chat interface, with most of the elements that made Habitat a real, functioning community stripped away. “Don’t get the idea that Club Caribe is a half-baked attempt to implement a vision that was too far ahead of its time,” wrote one magazine reviewer eager not to lose QuantumLink’s advertising dollars but also not quite willing to completely avoid the truth. That, of course, was exactly what it was, although even as Club Caribe the technology was still more than able to impress those who weren’t aware of the full scope of Chip Morningstar’s original dream. Habitat‘s technology continued to live on to one degree or another in other Fujitsu projects spanning much of the next decade, such as the Habitat Japan that was launched in 1990 and another virtual world called WorldsAway that was launched on CompuServe during that service’s twilight years of the mid-1990s. Yet it wouldn’t be until Ultima Online in 1997 that an online virtual world of quite the same scope and complexity as the original Habitat would be attempted again, and not until Second Life in 2003 that a functioning virtual community of quite the same player-driven character would be dared again. (Second Life also has a more direct connection to Habitat: F. Randall Farmer was a consultant on the later venture.) Such spans are practically millennia in the fast-paced world of computers. Habitat was perhaps doomed to fail from the beginning, but if you’re going to fail you might as well make it a failure for the ages.
QuantumLink, and the Club Caribe to be found within it, persisted as what might be most charitably described as a “legacy service” for a surprisingly long time, not finally going dark until November 1, 1994, just a few days shy of its ninth birthday and after the death of Commodore itself. By that time, however, it was little more than an afterthought for its parent company, now one of the biggest success stories of the nascent Internet boom. After some hiccups in the form of failed AppleLink and PCLink services, you see, Steve Case had finally hit paydirt with America Online, which became the new name of Quantum Computer Services itself in 1991. I’m going to guess that you might have heard of them. If not… well, stick around. We’ll get there.
(Sources: the book Droidmaker by Michael Rubin; Compute!’s Gazette of January 1985, May 1985, March 1986, January 1987, December 1987, and January 1989; Run of August 1986 and November 1989; Commodore Microcomputers of November/December 1986; Computer and Video Games of May 1987. Two episodes of the television show Computer Chronicles, “Modems and Bulletin Boards” from 1985 and “Online Services Part One” from 1987, are pertinent. The Internet Archive gives access to a now-defunct site created by Keith Elkin that was devoted to Habitat. The Museum of Art and Digital Entertainment’s site hosts a Lucasfilm-to-Fujitsu technology-transfer document dating from the summer of 1988 that’s full of fascinating information and insights. Finally, Chip Morningstar and F. Randall Farmer have a website full of still more invaluable information and insights.)
Richard Garriott applied the same term to a game even earlier than Morningstar, during the development of Ultima IV: Quest of the Avatar, but his usage applies specifically to the ethical quest that forms the plot of that game: the player’s goal is to become an “avatar of virtue.” The more generalized use of the term in videogames is better attributed to Morningstar, even if Garriott’s usage probably better reflects its real implications in Hinduism.
As part of a general rearranging of the deck chairs at Lucasfilm in late 1985, the Games Group got moved from their nondescript offices in San Rafael to nearby Skywalker Ranch, the “filmmaker’s retreat” at the very heart of George Lucas’s empire. They were housed in an ornate structure of Victorian brick called the Stable House, with crackling fireplaces in almost every room. Later, old-timers would tell newcomers stories of the Games Group’s time at Skywalker Ranch, which would last for just a few years, like legends from before the Fall: catching a sneak preview of a new David Lynch film in the company of Lynch himself in the Ranch’s beautiful 300-seat art-deco theater; hanging out on a regular basis with Steven Spielberg, who wanted to play everything the Games Group had in development every time he stopped by, sometimes for hours at a stretch; playing softball on the Ranch’s gorgeously manicured field with rock star Huey Lewis; hiking up to the observatory after a long day at the office to do another sort of stargazing; eating gourmet lunches every day at the Ranch’s restaurant for $5 a pop. They might not have been able to make Star Wars games, but they could surround themselves with its trappings: when first moving in, they were given a chance to rummage through an enormous warehouse full of old props and concept art for office decorations. It’s questionable whether any other game studio, ever, has worked in quite such a nerd Elysium.
Continuing to blow through Skywalker Ranch as they had San Rafael, however, were winds of change that had been steadily altering Lucasfilm’s expectations of their little Games Group. As the middle years of the decade wore on, the company was becoming a very different place from what it had been during the free-and-easy early 1980s, when money seemed to flow like water. Lucasfilm’s financial outlook had changed almost overnight in 1983 when, even as Return of the Jedi was doing the expected huge numbers in theaters, George Lucas announced that he and his wife Marcia were getting a divorce. An accomplished film editor in her own right, Marcia had been a huge contributor to the Star Wars movies, especially the first, for which she’d won an Oscar — something her ex-husband has never managed — for her editing work. Now her divorce settlement would cost Lucasfilm big, to the tune of $50 to $100 million (precise estimates vary). Lucasfilm’s financial advisers were able to convince her to take her settlement as a series of payments spread over years rather than the lump sum the initial agreement demanded, but those payments nevertheless put a tremendous drain on the company’s finances.
And soon the other side of the ledger, that of incoming earnings, also began to diminish. George Lucas had long since declared that Star Wars was to be but a single trilogy of films, that there would be no more after Return of the Jedi. The lack of new films inevitably meant not just the loss of box-office receipts but also diminished sales of the toys and other merchandise that had always been the franchise’s biggest cash cow. Meanwhile the Indiana Jones series, which had turned into almost as successful a franchise as Star Wars, fell into a five-year hiatus after 1984’s Indiana Jones and the Temple of Doom. Filling that gap for Lucasfilm were a series of middling disappointments — Labyrinth, Tucker: The Man and His Dream, Willow, some almost perversely low-stakes Star Wars television programs featuring R2-D2 and C-3PO and, God help us, the Ewoks — and at least one outright bomb big enough to have become a punchline for the ages in Howard the Duck. It seemed that Lucas, who could do no wrong in the eleven years between American Graffiti and Indiana Jones and the Temple of Doom, had suddenly seen his Midas Touch desert him.
Never much of a manager and certainly not a numbers guy, Lucas hired a no-nonsense sort named Doug Norby to become Lucasfilm’s president in 1985. “Do what you have to do,” he told him, “and I’m just going to stay out of it.” Norby declared that there needed to be a culture change. Every division would now be expected to justify their existence by earning money for the company rather than costing it money. Those who couldn’t see a way to do so would get the axe. Ditto individual personnel within departments that had become too bloated; Norby orchestrated the first significant wave of layoffs ever to sweep over Lucasfilm. As the conflict-averse Lucas had likely intended, Norby was blamed for all of the pain and chaos, became for some time the most hated name at Lucasfilm, while Lucas himself was largely given a pass, as if he somehow didn’t know about the changes underway in his own namesake company.
As part of the restructuring, it was decided that Lucasfilm would now engage in only two specific lines of business: providing production services to the film industry (Industrial Light and Magic, Skywalker Sound) and making mass-market entertainments. The old Computer Graphics Group that had awkwardly spawned the Games Group still hadn’t really proved themselves to belong in the former category, while the Games Group, at least if you squinted just right, pretty much did belong in the latter. Thus, while the Games Group got to remain at Lucasfilm, the Graphics Group in February of 1986 was spun off to a collection of investors that included many of their own current personnel as well as, as ringmaster of the whole proceeding, Steve Jobs, recently exiled from Apple. The old Graphics Group was now known as Pixar, selling a $135,000 graphics workstation which they had developed during their years with Lucasfilm. Most of the rest of Lucasfilm’s computer-oriented research was either cancelled outright or similarly packaged up and sold off. (Most notably, Lucasfim’s EditDroid digital-editing project became an independent company called Droid Works.)
Soon the old Games Group represented the only significant hacker presence left at Lucasfilm. It was during this period of colossal change that George Lucas took rare personal notice of Games for long enough to deliver his most oft-quoted piece of advice to Steve Arnold: “Stay small, be the best, don’t lose any money.” This commandment has often been taken to represent a sort of creative carte blanche for Arnold and his charges. Taken in the context in which it was uttered, however, it’s probably better seen as a warning. The Games Group was free to continue to trade on the Lucasfilm name and enjoy their gourmet lunches at the company cafeteria, but they’d have to start paying their own way from here on. Should they fail at that, their rope would not be a long one, for Lucas had little personal investment in their work.
Given this situation, when Lucasfilm’s brass decided to throw the Games Group a bone in the form of an actual piece of intellectual property with which to work Arnold certainly didn’t turn up his nose at the prospect. It wasn’t Star Wars or even Indiana Jones, but it was a much-anticipated film called Labyrinth, a fantasy adventure directed by Jim Henson and starring David Bowie that was to be released in the summer of 1986. Beginning in November of 1985, Arnold poured most of his resources into the project, Lucasfilm Games’s first adventure game. The Henson connection secured the involvement of Christopher Cerf, a Sesame Street stalwart and all-around Renaissance man of the arts who seemed to know everyone and be involved with everything in the world of entertainment. Cerf was a good friend of Douglas Adams, a frequent guest at his legendary gala dinner parties; it had in fact been Cerf who had largely brokered the deal with Infocom that had led to the Hitchhiker’s Guide to the Galaxy computer game. In January much of the Games Group flew to London for an intense week of consultation with Henson, Cerf, and their buddy Adams.
Labyrinth had been conceived from the beginning as a graphic adventure, a genre that was just beginning to emerge from the primordial muck thanks largely to the work of Sierra and ICOM Simulations. It was Adams who suggested the game’s brilliant cold open: it begins as an ordinary text adventure, and not a very good one at that, until you arrive at a cinema and get sucked into the movie playing there by a pixelated David Bowie. It’s a ludic version of that iconic moment in The Wizard of Oz when the film suddenly shifts from black and white to color. Some of his other subversive touches, playing as he loved to do with the artificiality of the medium itself, weren’t so easily implemented. The team particularly lamented that they wouldn’t be able to use Adams’s idea for a film-editing room found in the game. He suggested that you should be able to view the scraps of film to see snippets of your own previous adventures, maybe even forgotten tributaries down which you’d wandered before restoring the game to its current state. Alas, something like that just wasn’t going to happen on the likes of a Commodore 64.
Not really a bad game but also not quite a fully baked one, Labyrinth would prove to be something of a steppingstone on the way to a grand tradition of Lucasfilm adventure games still to come. Your character can be moved about using the joystick, but other commands must be constructed rather awkwardly, by using the arrow keys to cycle through two separate lists, one of available verbs and one of nouns. Notably, when a verb is selected the list of nouns is limited to only those which logically apply, thus making it at least theoretically impossible to construct a completely nonsensical “sentence.” Driving much of the design was a philosophy that adventure games should be friendlier, less tedious, and much less deadly than was the norm from competitors like Sierra. It is, for instance, almost impossible to get yourself killed in Labyrinth, and David Fox noted in contemporaneous interviews how he had strained to “eliminate the dead-end or ‘insoluble’ situation.” In years to come Lucasfilm Games would virtually define themselves in opposition to what they saw as the Bad Old Way of doing adventure games, as particularly personified by the games of Sierra. It’s an idea that would take some experience and some technology upgrades to reach complete fruition, but it’s interesting to note that it was present right from the beginning.
Labyrinth. Note the “slot-machine” verb-noun selector at the bottom of the screen.
Released in June of 1986, the movie version of Labyrinth thoroughly underwhelmed by the standards of an expensive would-be blockbuster, spending just one week inside the top ten in the United States and garnering mixed (at best) reviews. The odor of a flop inevitably clung to the game as well when it was released two months later. Despite lots of advertising and the usual free publicity garnered from journalists eager to come out to Skywalker Ranch and bask in the aura of Star Wars, it became on the whole a commercial disappointment. This was now becoming a depressingly common theme for the Games Group. They were perilously close to violating that last and most important of Lucas’s commandments.
Their savior would come from a much smaller, quieter project than the big Labyrinth tie-in — indeed, a project from which Steve Arnold seemed to have no real expectations at all. Its father had himself been heretofore one of the less noticeable employees of the Games Group, a friendly, unassuming fellow with a wry sense of humor and a great aptitude for programming. His name was Ron Gilbert, and he was motivated by that most compelling of all workplace impulses: he was just trying not to get fired.
Born in 1964 in the rural Oregon town of La Grande, Gilbert had been programming since 1977, when his father brought home the family’s first Texas Instruments programmable calculator. Soon after starting at his hometown Eastern Oregon State College in 1982, he bought his first Commodore 64, and immediately discovered one of that machine’s most conspicuous weaknesses: its BASIC interpreter had no support whatsoever for the very graphics and sound capabilities that made the 64 so special. Working with a buddy named Tom McFarlane, he developed a BASIC extension called Graphics BASIC to change all that, adding over a hundred new commands to the language. It was impressive enough that they were able to sell it to HESWare, one of the biggest publishers in software at the time. In fact, HESWare was so impressed with Gilbert personally that they offered him a full-time job as an in-house programmer. So, he dropped out of university to move to Brisbane, California.
It didn’t work out. HESWare turned out to be a flash in the pan that had made a ton of unwise financial decisions in their eagerness to rule the software roost. Within months of Gilbert’s arrival the company collapsed, well before releasing anything he had worked on. He was forced to return sheepishly to La Grande to contemplate re-enrolling at Eastern Oregon — luckily, his dad was the president there — and getting back to the real world of adult employment; maybe he could get a job as a programmer at a bank or something. Then, one day in October of 1984, the telephone rang just as he was leaving the house. Prompted by he wasn’t quite sure what, he decided to rush back inside and answer it. It was Steve Arnold from Lucasfilm Games. He and his colleagues had seen Graphics BASIC and heard about Gilbert’s talents through the grapevine, Arnold explained. They needed someone to help port their games, which had been originally developed for Atari 8-bit machines, to the Commodore 64. Would he be willing to come down to San Rafael to talk about a possible contract? Like most prospective employees Arnold spoke to, Gilbert didn’t have to think twice when the company behind Star Wars came calling. It was just an interview, and for a contract position at that, but he nevertheless packed all of his possessions into his 280Z and took off for California. He had no intention of coming back.
He didn’t need to; he got the job. Still, as a contractor rather than a regular employee he was left perpetually uncertain about how long he’d get to live the dream. His anxiety only increased after the Commodore 64 versions of the Games Group’s modest early catalog of four action games were all pretty much complete and nobody seemed to be giving him any clear information about what he was expected to do next. Working with a couple of the other guys, he came up with a fanciful game proposal for Arnold’s bulging ideas file: I Was a Teenage Lobot, a “science-fiction role-playing strategy adventure game.” (Better check again, guys; I think you may have missed a genre or two.) But then the big Labyrinth project came along, depriving him of his would-be partners. Ominously, Gilbert was one of the few people in the Games Group not earmarked to that game.
Whether Steve Arnold was really snubbing him or whether he saw something special in him and wanted to give him his own space to figure out for himself what that was is still an open question. What is clear is that Gilbert started toying with another idea to justify his existence there at Skywalker Ranch, involving a group of kids sent, Scooby-Doo-style, to explore a creepy old mansion.
Gilbert claims that he didn’t originally conceive of Maniac Mansion as an adventure game at all, perhaps because one of its central conceits had rarely been done in an adventure game before. From the beginning, he was determined that you should be able to control several kids rather than just one, each of whom would have her own personality and abilities. Much of the gameplay would hinge on coordinating the kids’ actions to achieve things none of them could manage on her own. And that was pretty much the whole idea; just about everything else about the design seemed to be up in the air. But then, visiting home for the Christmas of 1985, he saw his eight-year-old cousin obsessively playing Sierra’s King’s Quest. Gilbert loved the graphics, but didn’t care for Roberta Williams’s death-heavy philosophy of game design any more than he did for Sierra’s primitive parser, which made a particularly poor fit with a game that was otherwise so graphics-oriented. He decided that he wanted to do an adventure game “because I hate adventure games,” because he wanted to show the world how they could be so much better.
I hated that you died all the time. You’d be walking along and you would step somewhere and out of the blue you would die. That just seemed frustrating to me. I think a lot of designers must think that’s fun. But it’s not. It’s horrible.
And too often the game devolved into what Gilbert calls “second-guess the parser”:
You would see a bush on the screen, and you’d type, “Pick up bush,” and it would say, “I don’t know what a ‘bush’ is.” Then you’d type, “Pick up plant,” and it would say, “I don’t know what a ‘plant’ is.” Then you’d type, “Pick up shrubbery,” and it would say, “I don’t know what a ‘shrubbery’ is.” Pretty soon you’d type, “Fuck you,” and it would say, “I don’t understand what ‘fuck’ is.”
So, I’m looking at this bush or plant or shrub and I cannot figure out the word that the game designer is using for it. That’s very frustrating because I can see it right on the screen. Why can’t I just click on it?
And the next logical step is: if I can just click on objects on the screen, why can’t I just click on verbs as well? Really, despite what the marketing departments and the backs of the boxes were telling us, these games only understood a very small number of verbs.
Beginning from textual lists of verbs and nouns much like the interface of Labyrinth, Maniac Mansion evolved into a much more intuitive experience: a clickable list of verbs at the bottom of the screen, which can be combined with hotspots in the pictures proper to build commands. In its day it was simply the best, most elegant interface for graphical adventuring yet devised. One might call it a combination of the best traits of the two most prominent systems for graphic adventuring already extant at the time: Sierra’s AGI games that debuted with King’s Quest and the ICOM Simulations line of adventures that began with Déjà Vu. Like the former, you can see your avatar (or avatars in this case) and move them about onscreen, but like the latter you don’t have to wrestle with a parser, being able instead to simply click on verbs and objects in your inventory or in the environment proper to construct commands. It’s afflicted with neither the perpetual disconnect between textual parser and graphical worldview that can make the AGI games so frustrating nor the cluttered, cramped feel of ICOM’s overly baroque interface. Maniac Mansion would prove to be by far the most graphical graphical adventure of its time, willing to do most of its storytelling through visuals and the occasional well-chosen sound effect rather than the big text dumps that mark the Sierra and ICOM games. Tellingly, it devotes exactly one line of the screen to text messages.
On the job in Maniac Mansion. Note the selectable list of verbs (including the immortal “New Kid”) and the character’s inventory below.
Gilbert found a great supporter of his budding adventure game in Gary Winnick, the Games Group’s indefatigable visual artist. In between contributing much of the art found in both Labyrinth and Habitat, Winnick found time to brainstorm Maniac Mansion and to create heaps of sample art. Yet progress was painfully slow. Gilbert was trying to build Maniac Mansion in the same way that Labyrinth was being built, by coding it from scratch in pure assembly language. Problem was, he was trying to do it alone. As 1986’s midpoint approached, Steve Arnold was getting noticeably annoyed at his apparent lack of productivity and Gilbert was surer than ever that he would be sent back to La Grande any day now.
It was at this juncture that Chip Morningstar made the suggestion that would change the direction of Lucasfilm Games forever. Why didn’t he devise a high-level scripting language that could be compiled on the Games Group’s big Unix workstations, then run on the Commodore 64 itself via an interpreter? Morningstar even took the time to help him design the language, a sort of cut-down version of some of the tools he and Randall Farmer were using to build the virtual world of Habitat, and to write the first compiler. SCUMM — the Script Creation Utility for Maniac Mansion — was born.
It wasn’t precisely a new idea, but it was vastly complicated by the need a graphic adventure like Maniac Mansion had to do many things concurrently, in real time. Many different “scripts” would need to run at the same time, forcing Gilbert to code what amounted to a multitasking kernel for the whole system on the little Commodore 64. Even with Morningstar’s help, it took Gilbert a full six months to get the SCUMM system up and running. Meanwhile Gary Winnick’s art continued to pile up, looking for a home, and Gilbert continued to tremble every time Steve Arnold looked his way. At last at the end of 1986 SCUMM was complete enough that he could return to the game proper. Arnold, evidently beginning to feel that his work had real potential, allowed David Fox to join him as a SCUMM scripter. Winnick as well was now working virtually full-time on the project, contributing not only all of the art but also major swathes of design and story.
Gilbert credits SCUMM and the relative ease with which it let the programmer script interactions for making the world of the finished Maniac Mansion much more interactive and alive than it could otherwise have been. Not least amongst the little gags and Easter eggs SCUMM facilitated was a certain soon-to-be-infamous hamster-in-a-microwave bit. Gilbert insists that it was actually Fox and Winnick who came up with and implemented this particular piece of tasteless humor, so angry missives should be directed their way rather than to him.
Winnick drew the kids you control and the other characters that inhabit the mansion like bobblehead dolls, heads out of all proportion to their bodies, to make sure their personalities came across despite the low screen resolution of the Commodore 64. He had already used the same technique in both Labyrinth and Habitat and would continue to do so for some time to come; it would become the most instantly recognizable graphical trait of early Lucasfilm adventure games. Gilbert’s original plan had called for the kids to literally be kids — children. Realizing, however, that no one wanted to see children endangered and potentially dispatched in gruesome ways, Gilbert and Winnick decided to make the kids teenagers, which made a better demographic fit anyway with the teenage players who were the biggest audience for computer games. They form a group of seven broad high-school archetypes, sketched with just a hint of a satirical edge, from amongst which you choose three to see you through the game. Bernard is an electronics buff, physics champion, and all-around nerd; Wendy is a prim and proper “aspiring novelist” who seems to have been born at age 40; Michael is Yearbook Guy at the school, an ace photographer; Jeff is a surfer dude who seems to have wandered into Maniac Mansion whilst looking for California Games; and, betraying perhaps a slight flagging of the creative muscles, both Sid and Razor are would-be rock stars (Sid’s a new waver, Razor a punk, for whatever that’s worth). And finally there’s Dave, a Good Kid of the sort who runs for Class President. He’s the leader of the group and the one kid you have to play with. It’s his girlfriend Sandy — a cheerleader, naturally — who’s been kidnapped by Dr. Fred, the creepy owner of the mansion, to feed to aliens. Some real people found themselves immortalized inside these archetypical shells: Razor’s look was based on Winnick’s girlfriend, Wendy on an accountant (what else?) at the office, Dave on Ron Gilbert himself. All of the kids have unique talents, some expected, some less so; clueless Jeff’s inexplicable hidden talent for fixing telephones is actually one of the funniest gags in the game. The idea was that any combination of kids should be capable of solving the game.
The kids. From left: Dave, Sid, Michael, Wendy, Bernard, Razor, and Jeff.
It was an idea that would cause Gilbert and Winnick no small amount of angst. Neither had ever designed an adventure game before, much less a knotty tapestry like this with its combinatorial explosion of protagonists, and their design document consisted of little more than a map of the mansion and a list of objects and the puzzles to which they applied. They desperately wanted to create an adventure game that would be more friendly and forgiving than the typical Sierra effort, but, inevitably, their lack of experience and planning and time, not to mention play-testing — the Games Group’s testing department consisted of exactly one guy sitting in front of a Commodore 64 with a pad of paper — led to a game fairly riddled with potential dead ends and unwinnable situations despite its designers’ best intentions. Gilbert, a great and much-needed advocate for fairness in adventure design, still castigates himself for that to this day.
Both Gilbert and Winnick were fans of knowingly schlocky B-grade horror movies like the then-recent Re-Animator. Maniac Mansion was conceived very much as an homage to the genre. The actual plot, of the mad scientist who owns the mansion attempting to tap the power of a mysterious meteorite that fell on his property, was inspired by one of the vignettes in Creepshow, an anthology of short horror films. Other references, like the man-eating plant lifted whole cloth from Little Shop of Horrors, are even more obvious. Still, it was going to have to be a much more family-friendly affair if it was to bear the Lucasfilm name. When Arnold demanded that all traces of swearing be removed from the game, Gilbert and Winnick did so only under duress, and to the tune of plenty of grumbling about “artistic vision” and the like. If you can tell me exactly why Dave has to call Bernard a “shithead” at the outset of the night, said Arnold, you can keep it. No one could. Gilbert says that the lesson thus imparted about the pointlessness of gratuitous profanity has stuck with him to this day.
Better a tuna head than a shithead…
For the mansion itself, they found a fecund source of inspiration very close to home indeed: the big neo-Victorian “Main House” at Skywalker Ranch. The spiral staircase inside the library in Maniac Mansion is lifted straight from the “filmmaker’s research library” in the Main House. In the game, the staircase has an “out of order” sign on it and cannot be climbed under any circumstances. This was a subtle inside joke: George Lucas’s personal office was on the balcony at the top of those stairs in the real house, and nobody was allowed to go up there without an invitation.
Given that it was a game inspired largely by movies that was being developed at a movie studio, Gilbert wanted to give Maniac Mansion a cinematic flavor. He imagined little episodes that would “cut away” from the player’s current actions to advance the plot and show what the captive Sally, her captor Dr. Fred, and the other creepy inhabitants of the mansion were up to. He asked Arnold if there was a filmmaking term for this technique that he could employ. Arnold said that “cut scene” sounded more than good enough to him. Thus did a new term enter the gaming lexicon. Maniac Mansion was hardly the first game to employ them — there was Jordan Mechner’s 1984 classic Karateka and Sierra adventure games like Space Quest and even the old Ms. Pac-Man game in the arcades — but it had been left to Lucasfilm to finally give them a name. The concept was baked right into the SCUMM language, with a special kind of script called simply “cut-scene” that when triggered would automatically save the player’s state, play the cut scene as a little animated movie all its own, and then restore the player to control.
One ironic consequence of the cut scenes is to make the game harder in just the ways that Gilbert would have preferred to avoid. Most of them are triggered by simple timers. While some are just there for atmosphere or to convey information, others directly affect the state of the world, such as when a postman arrives with a package. There are often things you must do to react or to prepare for these dynamic events; failing to do so can lock you out of victory. Had anyone been paying attention, Infocom’s Ballyhoo had already pioneered a better way to advance the plot inside an adventure game, by tying events to the player’s progress rather than hard-coded timers. Like many such lessons, it would be learned only slowly by game designers, and largely by a process of reinventing the wheel at that. As it is, Maniac Mansion has some of the feel of the earlier Infocom mysteries, of needing to learn how to steer events just right over the course of multiple restores.
Shortly before the release of Labyrinth, Lucasfilm Games had severed their relationship with Epyx and moved on to Activision. It was thus under that company’s banner that Maniac Mansion made its public debut at the June 1987 Summer Consumer Electronics Show, host to so much of the last great wave of Commodore 64 software. Before Maniac Mansion could actually be released, however, Arnold made the huge decision to self-publish it under Lucasfilm’s own banner. Lucasfilm Games changed from being a mere developer to being an “affiliated publisher” of Activision, a status that gave them more independence and put their own name alone on their boxes but still gave them access to the larger company’s distribution network and other logistical support. Even with Activision’s support, publishing entailed engaging with entire facets of the software industry from which they’d always been happily insulated before. They learned a harsh lesson about the sensitivities of some Americans when Toys ‘R’ Us, one of the biggest Commodore 64 game retailers in the country, abruptly pulled the game off their shelves in response to a customer complaint. It seemed some old biddy had seen the tongue-in-cheek copy on the back of the box, which declared Maniac Mansion to be (amongst other things) a story of “love, lust, and power,” and had objected in no uncertain terms. Lucasfilm was forced to hurriedly redesign the box in order not to lose Toy ‘R’ Us forever.
Lucasfilm Games’s advertisements took aim at “most story game designers” who “seem to think people love to get clobbered.” I wonder which designers they’re talking about…
But it all worked out in the end. Coming out as it did with the Lucasfilm Games logo — and only the Lucasfilm Games logo — all over its box, Maniac Mansion proved a pivotal release for this little concern that, despite brilliant personnel and a name to die for, had struggled for years now to come up with a definitive commercial identity. One of the huge advantages of the SCUMM system was that it made porting games to new platforms relatively easy, just a matter of writing a new interpreter. Thus by 1988 Maniac Mansion could be bought in versions for the Amiga, Atari ST, Apple II, and MS-DOS in addition to the Commodore 64 original. In time it would even make its way to the Nintendo Entertainment System. (See Doug Crockford’s “The Expurgation of Maniac Mansion“ to learn of the hilarious lengths the Games Group had to go through to get it accepted by Nintendo’s censorious management regime, who made the Toys “R” Us lady look like a libertine.) While it never topped many sales charts, Maniac Mansion turned into a perennial back-catalog star, selling far more units when all was said and done than any game the Games Group had released before. Its continuing popularity was such that in 1990 it spawned a successful children’s television series, a claim to fame that very few games can boast. Such success enabled Lucasfilm Games at last to firmly plant their feet and adhere to Lucas’s dictum to “not lose any money” while they built upon the reputation it engendered for them. They were now known first and foremost as a maker of graphic adventure games, the yin to Sierra’s yang. They had traveled a long and winding road to get here, but it seemed they had finally found a calling.
Maniac Mansion’s intrinsic value as a game is often dismissed today in favor of its historical role as the urtext for the many much-loved SCUMM games that followed it. That, however, is a shame, for its charms as the best graphic adventure ever made for the Commodore 64 are real, varied, and considerable. Yes, it’s a bit of shaggy beast in contrast to those later Lucasfilm classics, but it’s also in many ways the most complex and interesting of any of them; no other SCUMM game boasts anything like its seven different playable characters, with all of the alternate storylines and solutions they bring with them.
Yet the most winning thing about Maniac Mansion is its personality, which is in turn a tribute to the personalities who created it. Gilbert and Winnick, one senses, want you to have a good time, want you to solve the game and then come back for more, trying on new combinations of characters for size. Thanks largely to the essential good faith and sense of fair play with which its authors approached it, Maniac Mansion is a game that’s hard to dislike, despite its occasional sins in the form of a puzzle or two that could have been clued slightly better and one really egregious example of hunt-the-hotspot (hint: check the library very carefully). Its puzzles are varied, usually logical in their wacky way, and always entertaining, and are given a wonderful added dimension by the need to coordinate two or sometimes even all three kids in far-flung corners of the mansion to solve some of the more intricate problems. (Interestingly, Level 9 in Britain was doing much the same thing during the same time period in the realm of text adventures.) One other thing that helps immeasurably is that the mansion is a relatively constrained environment, limiting the scope of possibility enough to keep things manageable. And of course it also helps that the game manages to evoke the sylvan atmosphere of a long teenage summer night so beautifully using the blunt instrument of 8-bit graphics and sound. Likeability, good faith, and good intentions will get you a long way, in games as in life, and talent doesn’t hurt one bit either. Thankfully, Ron Gilbert, Gary Winnick, and their colleagues were possessed of all of the above in spades.
(Sources: the book Droidmaker by Michael Rubin; The Transactor of July 1986; The LucasArts Adventurer of Spring 1991; Commodore Magazine of June 1987 and November 1988; Computer and Video Games of December 1986; Retro Gamer 94 and 116. Ron Gilbert has a wealth of material on his own history on his website and his “Making of Maniac Mansion” presentation was also invaluable.
In April of 1988, Brian Moriarty of Infocom flew from the East Coast to the West to attend the twelfth West Coast Computer Faire and the first ever Computer Game Developers Conference. Hard-pressed from below by the slowing sales of their text adventures and from above by parent company Activision’s ever more demanding management, Infocom didn’t have the money to pay for Moriarty’s trip. He therefore had to go on his own dime, a situation which left him, as he would later put it, very “grumpy” about the prospect of his ongoing employment by the very company at which he had worked so desperately to win a spot just a few years before.
The first West Coast Computer Faire back in 1977 had hosted the public unveiling of the Apple II and the Commodore PET, thus going down in hacker lore as the moment when the PC industry was truly born. By 1988, the Faire wasn’t the hugely important gathering it once had been, having been largely superseded on the industry’s calendar by glitzier events like the Consumer Electronics Show. Nevertheless, its schedule had its interesting entries, among them a lecture by Chris Crawford, the founder of the Computer Game Developers Conference which Moriarty would attend the next day. Moriarty recalls showing up a little late to Crawford’s lecture, scanning the room, and seeing just one chair free, oddly on the first row. He rushed over to take it, and soon struck up a conversation with the man sitting next to him, whom he had never met before that day. As fate would have it, his neighbor’s name was Noah Falstein, and he worked for Lucasfilm Games.
Attendees to the first ever Computer Game Developers Conference. Brian Moriarty is in the reddish tee-shirt at center rear, looking cool in his rock-star shades.
Falstein knew and admired Moriarty’s work for Infocom, and knew likewise, as did everyone in the industry, that things hadn’t been going so well back in Cambridge for some time now. His own Lucasfilm Games was in the opposite position. After having struggled since their founding back in 1982 to carve out an identity for themselves under the shadow of George Lucas’s Star Wars empire, by 1988 they finally had the feeling of a company on the rise. With Maniac Mansion, their big hit of the previous year, Falstein and his colleagues seemed to have found in point-and-click graphical adventures a niche that was both artistically satisfying and commercially rewarding. They were already hard at work on the follow-up to Maniac Mansion, and Lucasfilm Games’s management had given the go-ahead to look for an experienced adventure-game designer to help them make more games. As one of Infocom’s most respected designers, Brian Moriarty made an immediately appealing candidate, not least in that Lucasfilm Games liked to see themselves as the Infocom of graphical adventures, emphasizing craftsmanship and design as a way to set themselves apart from the more slapdash games being pumped out in much greater numbers by their arch-rivals Sierra.
Brian Moriarty on the job at Lucasfilm.
For his part, Moriarty was ripe to be convinced; it wasn’t hard to see the writing on the wall back at Infocom. When Falstein showed him some photographs of Lucasfilm Games’s offices at Skywalker Ranch in beautiful Marin County, California, and shared stories of rubbing elbows with movie stars and casually playing with real props from the Star Wars and Indiana Jones movies, the contrast with life inside Infocom’s increasingly empty, increasingly gloomy offices could hardly have been more striking. Then again, maybe it could have been: at his first interview with Lucasfilm Games’s head Steve Arnold, Moriarty was told that the division had just two mandates. One was “don’t lose money”; the other was “don’t embarrass George Lucas.” Anything else — like actually making money — was presumably gravy. Again, this was music to the ears of Moriarty, who like everyone at Infocom was now under constant pressure from Activision’s management to write games that would sell in huge numbers.
Brian Moriarty arrived at Skywalker Ranch for his first day of work on August 1, 1988. As Lucasfilm Games’s new star designer, he was given virtually complete freedom to make whatever game he wanted to make.
Noah Falstein in Skywalker Ranch’s conservatory. This is where the Games people typically ate their lunches, which were prepared for them by a gourmet chef. There were definitely worse places to work…
For all their enthusiasm for adventure games, the other designers at Lucasfilm were struggling a bit at the time to figure out how to build on the template of Maniac Mansion. Zak McKracken and the Alien Mindbenders, David Fox’s follow-up to Ron Gilbert’s masterstroke, had been published just the day before Moriarty arrived at Skywalker Ranch. It tried a little too obviously to capture the same campy charm, whilst, in typical games-industry fashion, trying to make it all better by making it bigger, expanding the scene of the action from a single night spent in a single mansion to locations scattered all around the globe and sometimes off it. The sense remained that Lucasfilm wanted to do things differently from Sierra, who are unnamed but ever-present — along with a sly dig at old-school rivals like Infocom still making text adventures — within a nascent manifesto of three paragraphs published in Zak McKracken‘s manual, entitled simply “Our Game Design Philosophy.”
We believe that you buy games to be entertained, not to be whacked over the head every time you make a mistake. So we don’t bring the game to a screeching halt when you poke your nose into a place you haven’t visited before. In fact, we make it downright difficult to get a character “killed.”
We think you’d prefer to solve the game’s mysteries by exploring and discovering. Not by dying a thousand deaths. We also think you like to spend your time involved in the story. Not typing in synonyms until you stumble upon the computer’s word for a certain object.
Unlike conventional computer adventures, Zak McKracken and the Alien Mindbenders doesn’t force you to save your progress every few minutes. Instead, you’re free to concentrate on the puzzles, characters, and outrageous good humor.
Worthy though these sentiments were, Lucasfilm seemed uncertain as yet how to turn them into practical rules for design. Ironically, Zak McKracken, the game with which they began publicly articulating their focus on progressive design, is the most Sierra-like Lucasfilm game ever made, with the sheer nonlinear sprawl of the thing spawning inevitable confusion and yielding far more potential dead ends than its designer would likely wish to admit. While successful enough in its day, it never garnered the love that’s still accorded to Maniac Mansion today.
Lucasfilm Games’s one adventure of 1989 was a similarly middling effort. A joint design by Gilbert, Falstein, and Fox, Indiana Jones and the Last Crusade: The Graphic Adventure — an Action Game was also made — marked the first time since Labyrinth that the games division had been entrusted with one of George Lucas’s cinematic properties. They don’t seem to have been all that excited at the prospect. The game dutifully walks you through the plot you’ve already watched unfold on the silver screen, without ever taking flight as a creative work in its own right. The Lucasfilm “Game Design Philosophy” appears once again in the manual in almost the exact same form as last time, but once again the actual game hews to this ideal imperfectly at best, with, perhaps unsurprisingly given the two-fisted action movie on which it’s based, lots of opportunities to get Indy killed and have to revert to one of those save files you supposedly don’t need to create.
So, the company was rather running to stand still as Brian Moriarty settled in. They were determined to evolve their adventure games in design terms to match the strides Sierra was making in technology, but were uncertain how to actually go about the task. Moriarty wanted to make his own first work for Lucasfilm a different, more somehow refined experience than even the likes of Maniac Mansion. But how to do so? In short, what should he do with his once-in-a-lifetime chance to make any game he wanted to make?
Flipping idly through a computer magazine one day, he was struck by an advertisement that prominently featured the word “loom.” He liked the sound of it; it reminded him of other portentous English words like “gloom”, “doom,” and “tomb.” And he liked the way it could serve as either a verb or a noun, each with a completely different meaning. In a fever of inspiration, he sat down and wrote out the basis of the adventure game he would soon design, about a Loom which binds together the fabric of reality, a Guild of Weavers which uses the Loom’s power to make magic out of sound, and Bobbin Threadbare, the “Loom Child” who must save the Loom — and thus the universe — from destruction before it’s too late. It would be a story and a game with the stark simplicity of fable.
Simplicity, however, wasn’t exactly trending in the computer-games industry of 1988. Since the premature end of the would-be Home Computer Revolution of the early 1980s, the audience for computer games had grown only very slowly. Publishers had continued to serve the same base of hardcore players, who lusted after ever more complex games to take advantage of the newest hardware. Simulations had collected ever more buttons and included ever more variables to keep track of, while strategy games had gotten ever larger and more time-consuming. Nor had adventure games been immune to the trend, as was attested by Moriarty’s own career to date. Each of his three games for Infocom had been bigger and more difficult than the previous, culminating in his adventure/CRPG hybrid Beyond Zork, the most baroque game Infocom had made to date, with more options for its onscreen display alone than some professional business applications. Certainly plenty of existing players loved all this complexity. But did all games really need to go this way? And, most interestingly, what about all those potential players who took one look at the likes of Beyond Zork and turned back to the television? Moriarty remembered a much-discussed data point that had emerged from the surveys Infocom used to send to their customers: the games people said were their favorites overlapped almost universally with those they said they had been able to finish. In keeping with this trend, Moriarty’s first game for Infocom, which had been designed as an introduction to interactive fiction for newcomers, had been by far his most successful. What, he now thought, if he used the newer hardware at his disposal in the way that Apple has historically done, in pursuit of simplicity rather than complexity?
The standard Lucasfilm interface of the late 1980s, shown here in Maniac Mansion.
Lucasfilm Games’s current point-and-click interface, while undoubtedly the most painless in the industry at the time, was nevertheless far too complicated for Moriarty’s taste, still to a large extent stuck in the mindset of being a graphical implementation of the traditional text-adventure interface rather than treating the graphical adventure as a new genre in its own right. Thus the player was expected to first select a verb from a list at the bottom of the screen and then an object to which to apply it. The interface had done the job well enough to date, but Moriarty felt that it would interfere with the seamless connection he wished to build between the player sitting there before the screen and the character of Bobbin Threadbare standing up there on the screen. He wanted something more immediate, more intuitive — preferably an interface that didn’t require words at all. He envisioned music as an important part of his game: the central puzzle-solving mechanic would involve the playing of “drafts,” little sequences of notes created with Bobbin’s distaff. But he wanted music to be more than a puzzle-solving mechanic. He wanted the player to be able to play the entire game like a musical instrument, wordlessly and beautifully. He was thus thrilled when he peeked under the hood of Lucasfilm’s SCUMM adventure-game engine and found that it was possible to strip the verb menu away entirely.
Some users of Apple’s revolutionary HyperCard system for the Macintosh were already experimenting with wordless interfaces. Within weeks of HyperCard’s debut, a little interactive storybook called Inigo Gets Out, “programmed” by a non-programmer named Amanda Goodenough, had begun making the rounds, causing a considerable stir among industry insiders. The story of a house cat’s brief escape to the outdoors, it filled the entire screen with its illustrations, responding intuitively to single clicks on the pictures. Just shortly before Moriarty started work at Lucasfilm Games, Rand and Robyn Miller had taken this experiment a step further with The Manhole, a richer take on the concept of an interactive children’s storybook. Still, neither of these HyperCard experiences quite qualified as a game, and Moriarty and Lucasfilm were in fact in the business of making adventure games. Loom could be simple, but it had to be more than a software toy. Moriarty’s challenge must be to find enough interactive possibility in a verb-less interface to meet that threshold.
In response to that challenge, Moriarty created an interface that stands out today as almost bizarrely ahead of its time; not until years later would its approach be adopted by graphic adventures in general as the default best way of doing things. Its central insight, which it shared with the aforementioned HyperCard storybooks, was the realization that the game didn’t always need the player to explicitly tell it what she wanted to do when she clicked a certain spot on the onscreen picture. Instead the game could divine the player’s intention for itself, based only on where she happened to be clicking. What was sacrificed in the disallowing of certain types of more complex puzzles was gained in the creation of a far more seamless, intuitive link between the player, the avatar she controlled, and the world shown on the screen.
The brief video snippet above shows Loom‘s user interface in its entirety. You make Bobbin walk around by clicking on the screen. Hovering the mouse over an object or character with which Bobbin can interact brings up an image of that object or character in the bottom right corner of the screen; double-clicking the same “hot spot” then causes Bobbin to engage, either by manipulating an object in some way or by talking to another character. Finally, Bobbin can cast “spells” in the form of drafts by clicking on the musical staff at the bottom of the screen. In the snippet above, the player learns the “open” draft by double-clicking on the egg, an action which in this case results in Bobbin simply listening to it. The player and Bobbin then immediately cast the same draft to reveal within the egg his old mentor, who has been transformed into a black swan.
Moriarty seemed determined to see how many of the accoutrements of traditional adventure games he could strip away and still have something that was identifiable as an adventure game. In addition to eliminating menus of verbs, he also excised the concept of an inventory; throughout the game, Bobbin carries around with him nothing more than the distaff he uses for weaving drafts. With no ability to use this object on that other object, the only puzzle-solving mechanic that’s left is the magic system. In the broad strokes, magic in Loom is very much in the spirit of Infocom’s Enchanter series, in which you collect spells for your spell book, then cast them to solve puzzles that, more often than not, reward you with yet more spells. In Loom the process is essentially the same, except that you’re collecting musical drafts to weave on your distaff rather than spells for your spell book. And yet this musical approach to spell weaving is as lovely as a game mechanic can be. Lucasfilm thoughtfully included a “Book of Patterns” with the game, listing the drafts and providing musical staffs on which you can denote their sequences of notes as you discover them while playing.
The audiovisual aspect of Loom was crucial to capturing the atmosphere of winsome melancholia Moriarty was striving for. Graphics and sound were brand new territory for him; his previous games had consisted of nothing but text. Fortunately, the team of artists that worked with him grasped right away what was needed. Each of the guilds of craftspeople which Bobbin visits over the course of the game is marked by its own color scheme: the striking emerald of the Guild of Glassmakers, the softer pastoral greens of the Guild of Shepherds, the Stygian reds of the Guild of Blacksmiths, and of course the lovely, saturated blues and purples of Bobbin’s own Guild of Weavers. This approach came in very handy for technical as well as thematic reasons, given that Loom was designed for EGA graphics of just 16 onscreen colors.
The overall look of Loom was hugely influenced by the 1959 Disney animated classic Sleeping Beauty, with many of the panoramic shots in the game dovetailing perfectly with scenes from the film. Like Sleeping Beauty, Loom was inspired and accompanied by the music of Pyotr Ilyich Tchaikovsky, whom Moriarty describes as his “constant companion throughout my life”; while Sleeping Beauty draws from Tchaikovsky’s ballet of the same name, Loom draws from another of his ballets, Swan Lake. Loom sounds particularly gorgeous when played through a Roland MT-32 synthesizer board — an experience that, given the $600 price tag of the Roland, far too few players got to enjoy back in the day. But regardless of how one hears it, it’s hard to imagine Loom without its classical soundtrack. Harking back to Hollywood epics like 2001: A Space Odyssey, the MT-32 version of Loom opens with a mood-establishing orchestral overture over a blank screen.
To provide the final touch of atmosphere, Moriarty walked to the other side of Skywalker Ranch, to the large brick building housing Skywalker Sound, and asked the sound engineers in that most advanced audio-production facility in the world if they could help him out. Working from a script written by Moriarty and with a cast of voice actors on loan from the BBC, the folks at Skywalker Sound produced a thirty-minute “audio drama” setting the scene for the opening of the game; it was included in the box on a cassette. Other game developers had occasionally experimented with the same thing as a way of avoiding having to cover all that ground in the game proper, but Loom‘s scene-setter stood out for its length and for the professional sheen of its production. Working for Lucasfilm did have more than a few advantages.
If there’s something to complain about when it comes to Loom the work of interactive art, it must be that its portentous aesthetics lead one to expect a thematic profundity which the story never quite attains. Over the course of the game, Bobbin duly journeys through Moriarty’s fairy-tale world and defeats the villain who threatens to rip asunder the fabric of reality. The ending, however, is more ambiguous than happy, with only half of the old world saved from the Chaos that has poured in through the rip in the fabric. I don’t object in principle to the idea of a less than happy ending (something for which Moriarty was becoming known). Still, and while the final image is, like everything else in the game, lovely in its own right, this particular ambiguous ending feels weirdly abrupt. The game has such a flavor of fable or allegory that one somehow wants a little more from it at the end, something to carry away back to real life. But then again, beauty, which Loom possesses in spades, has a value of its own, and it’s uncertain whether the sequels Moriarty originally planned to make — Loom had been envisioned as a trilogy — would have enriched the story of the first game or merely, as so many sequels do, trampled it under their weight.
From the practical standpoint of a prospective purchaser of Loom upon its initial release, on the other hand, there’s room for complaint beyond quibbling about the ending. We’ve had occasion before to observe how the only viable model of commercial game distribution in the 1980s and early 1990s, as $40 boxed products shipped to physical store shelves, had a huge effect on the content of those games. Consumers, reasonably enough, expected a certain amount of play time for their $40. Adventure makers thus learned that they needed to pad out their games with enough puzzles — too often bad but time-consuming ones — to get their play times up into the region of at least twenty hours or so. Moriarty, however, bucked this trend in Loom. Determined to stay true to the spirit of minimalism to the bitter end, he put into the game only what needed to be there. The end result stands out from its peers for its aesthetic maturity, but it’s also a game that will take even the most methodical player no more than four or five hours to play. Today, when digital distribution has made it possible for developers to make games only as long as their designs ask to be and adjust the price accordingly, Loom‘s willingness to do what it came to do and exit the stage without further adieu is another quality that gives it a strikingly modern feel. But in the context of the times of the game’s creation, it was a bit of a problem.
When Loom was released in March of 1990, many hardcore adventure gamers were left nonplussed not only by the game’s short length but also by its simple puzzles and minimalist aesthetic approach in general, so at odds with the aesthetic maximalism that has always defined the games industry as a whole. Computer Gaming World‘s Johnny Wilson, one of the more sophisticated game commentators of the time, did get what Loom was doing, praising its atmosphere of “hope and idealism tainted by realism.” Others, though, didn’t seem quite so sure what to make of an adventure game that so clearly wanted its players to complete it, to the point of including a “practice” mode that would essentially solve all the puzzles for them if they so wished. Likewise, many players just didn’t seem equipped to appreciate Loom‘s lighter, subtler aesthetic touch.Computer Gaming World‘s regular adventure-gaming columnist Scorpia, a traditionalist to the core, said the story “should have been given an epic treatment, not watered down” — a terrible idea if you ask me (if there’s one thing the world of gaming, then or now, doesn’t need, it’s more “epic” stories). “As an adventure game,” she concluded, “it is just too lightweight.” Ken St. Andre, creator of Tunnels & Trolls and co-creator of Wasteland, expressed his unhappiness with the ambiguous ending in Questbusters, the ultimate magazine for the adventuring hardcore:
The story, which begins darkly, ends darkly as well. That’s fine in literature or the movies, and lends a certain artistic integrity to such efforts. In a game, however, it’s neither fair nor right. If I had really been playing Bobbin, not just watching him, I would have done some things differently, which would have netted a different conclusion.
Echoing as they do a similar debate unleashed by the tragic ending of Infocom’s Infidel back in 1983, the persistence of such sentiments must have been depressing for Brian Moriarty and others trying to advance the art of interactive storytelling. St. Andre’s complaint that Loom wouldn’t allow him to “do things differently” — elsewhere in his review he claims that Loom “is not a game” at all — is one that’s been repeated for decades by folks who believe that anything labeled as an interactive story must allow the player complete freedom to approach the plot in her own way and to change its outcome. I belong to the other camp: the camp that believes that letting the player inhabit the role of a character in a relatively fixed overarching narrative can foster engagement and immersion, even in some cases new understanding, by making her feel she is truly walking in someone else’s shoes — something that’s difficult to accomplish in a non-interactive medium.
Responses like those of Scorpia and Ken St. Andre hadn’t gone unanticipated within Lucasfilm Games prior to Loom‘s release. On the contrary, there had been some concern about how Loom would be received. Moriarty had countered by noting that there were far, far more people out there who weren’t hardcore gamers like those two, who weren’t possessed of a set of fixed expectations about what an adventure game should be, and that many of these people might actually be better equipped to appreciate Loom‘s delicate aesthetics than the hardcore crowd. But the problem, the nut nobody would ever quite crack, would always be that of reaching this potential alternate customer base. Non-gamers didn’t read the gaming magazines where they might learn about something like Loom, and even Lucasfilm Games wasn’t in a position to launch a major assault on the alternative forms of media they did peruse.
In the end, Loom wasn’t a flop, and thus didn’t violate Steve Arnold’s dictum of “don’t lose money” — and certainly it didn’t fall afoul of the dictum of “don’t embarrass George.” But it wasn’t a big hit either, and the sequels Moriarty had anticipated for better or for worse never got made. Ron Gilbert’s The Secret of Monkey Island, Lucasfilm’s other adventure game of 1990, was in its own way as brilliant as Moriarty’s game, but was much more traditional in its design and aesthetics, and wound up rather stealing Loom‘s thunder. It would be Monkey Island rather than Loom that would become the template for Lucasfilm’s adventure games going forward. Lucasfilm would largely stick to comedy from here on out, and would never attempt anything quite so outré as Loom again. It would only be in later years that Moriarty’s game would come to be widely recognized as one of Lucasfilm Games’s finest achievements. Such are the frustrations of the creative life.
Having made Loom, Brian Moriarty now had four adventure games on his CV, three of which I consider to be unassailable classics — and, it should be noted, the fourth does have its fans as well. He seemed poised to remain a leading light in his creative field for a long, long time to come. It therefore feels like a minor tragedy that this, his first game for Lucasfilm, would mark the end of his career in adventure games rather than a new beginning; he would never again be credited as the designer of a completed adventure game. We’ll have occasion to dig a little more into the reasons why that should have been the case in a future article, but for now I’ll just note how much an industry full of so many blunt instruments could have used his continuing delicate touch. We can only console ourselves with the knowledge that, should Loom indeed prove to be the last we ever hear from him as an adventure-game designer, it was one hell of a swansong.
(Sources: the book Game Design Theory and Practice by Richard Rouse III; ACE of April 1990; Questbusters of June 1990 and July 1990; Computer Gaming World of April 1990 and July/August 1990. But the bulk of this article was drawn from Brian Moriarty’s own Loom postmortem for, appropriately enough, the 2015 Game Developers Conference, which was a far more elaborate affair than the 1988 edition.
Loom is available for purchase is from GOG.com. Sadly, however, this is the VGA/CD-ROM re-release — I actually prefer the starker appearance of the original EGA graphics — and lacks the scene-setting audio drama. It’s also afflicted with terrible voice acting which completely spoils the atmosphere, and the text is bowdlerized to boot. Motivated readers should be able to find both the original version and the audio drama elsewhere on the Internet without too many problems. I do recommend that you seek them out, perhaps after purchasing a legitimate copy to fulfill your ethical obligation, but I can’t take the risk of hosting them here.)