The Oracle of Delphi, Chapter 16: The Shame of Cleomenes and the Glory of Miltiades
The Shareware Scene, Part 3: The id Boys
On December 14, 1990, Scott Miller of Apogee Software uploaded the free first installment of his company’s latest episodic game. He knew as he did so that this release would be, if you’ll pardon the pun, a game changer for Apogee. To signal that this was truly a next-generation Apogee game, he doubled his standard paid-episode asking price from $7.50 to $15.
Rather than relying on the character graphics or blocky visual abstractions of Apogee’s previous games, Commander Keen 1: Marooned on Mars was an animated feast of bouncy color. Rather than looking like a typical boxed game of five to ten years earlier, it looked quite literally like nothing that had ever been seen on an MS-DOS-based computer before. In terms of presentation at least, it was nothing less than computer gaming’s answer to Super Mario Bros., the iconic franchise that had done so much to help Nintendo sell more than 30 million of their videogame consoles in the United States alone.
Yet even Miller, who has been so often and justly lauded for his vision in recognizing that many computer owners were craving something markedly different from what the big game publishers were offering them, could hardly have conceived of the full historical importance of this particular moment. For it introduced to the world a small group of scruffy misfits with bad attitudes and some serious technical chops, who were living and working together at the time in a rundown riverfront house in Shreveport, Louisiana. Within a few months, they would begin to call themselves id Software, and under that name they would remake the face of mainstream gaming during the 1990s.
I must admit that I find it a little strange to be writing about humble Shreveport for the second time in the course of two articles. It’s certainly not the first place one would look for a band of technological revolutionaries. The perpetually struggling city of 200,000 people has long been a microcosm of the problems dogging the whole of Louisiana, one of the poorest states in the nation. It’s a raggedly anonymous place of run-down strip malls and falling-down houses, with all of the crime and poverty of New Orleans but none of that city’s rich cultural stew to serve as compensation.
Life in Shreveport has always been defined by the Red River which flows through town. As its name would imply, the city was founded to serve as a port in the time before the nation’s rivers were superseded by its railroads and highways. When that time ended, Shreveport had to find other uses for its river: thanks to a quirk of Louisiana law that makes casinos legal on waterways but not on dry land, residents of northeastern Texas and southern Arkansas have long known it primarily as the most convenient place to go for legal gambling. The shabbily-dressed interstate gamblers who climb out of the casino-funded buses every day are anything but the high rollers of Vegas lore. They’re just ordinary working-class folks who really, really should find something more healthy to do with their time and money than sitting behind a one-armed bandit in a riverboat casino, dropping token after token into the slot and staring with glazed eyes at the wheels as they spin around and around. This image rather symbolizes the social and economic condition of Shreveport in general.
By 1989, Al Vekovius of Shreveport’s Softdisk Publications was starting to fear that the same image might stand in for the state of his business. After expanding so dramatically for much of the decade, Softdisk was now struggling just to hold onto its current base of subscribers, much less to grow their numbers. The original Softdisk and Loadstar, their two earliest disk magazines, catered to aged 8-bit computers that were now at the end of their run, while Big Blue Disk and Diskworld, for MS-DOS computers and the Apple Macintosh respectively, were failing to take up all of their slack. Everything seemed to be turning against Softdisk. In the summer of 1989, IBM, whose longstanding corporate nickname of “Big Blue” had been the source of the name Big Blue Disk, threatened a lawsuit if Softdisk continued to market a disk magazine under that name. Knowing better than to defy a company a thousand times their size, Softdisk felt compelled to rename Big Blue Disk to the less catchy On Disk Monthly.
While the loss of hard-won brand recognition always hurts, Softdisk’s real problems were much bigger and more potentially intractable than that of one corporate behemoth with an overgrown legal department. The fact was, the relationship which people had with the newer computers Softdisk was now catering to tended to be different from the one they had enjoyed with their friendly little Apple II or Commodore 64. Being a computer user in the era of Microsoft’s ascendancy was no longer a hobby for most of them, much less a lifestyle. People had less of a craving for the ramshackle but easily hackable utilities and coding samples which Softdisk’s magazines had traditionally published. People were no longer interested in rolling up their sleeves to work with software in order to make it work for them; they demanded more polished programs that Just Worked right off the disk. But this was a hard field for Softdisk to compete on. Programmers with really good software had little motivation to license their stuff to a disk magazine for a relative pittance when they could instead be talking to a boxed-software publisher or testing the exploding shareware market.
With high-quality submissions from outside drying up just as he needed them most, Vekovius hired more and more internal staff to create the software for On Disk. Yet even here he ran up against many of the same barriers. The programmers whom he could find locally or convince to move to a place like Shreveport at the salaries which Softdisk could afford to pay were generally not the first ones he might have chosen in an ideal world. For all that some of them would prove themselves to be unexpectedly brilliant, as we’ll see shortly, virtually every one of them had some flaw or collection thereof that prevented him from finding gainful employment elsewhere. And the demand that they churn out multiple programs every month in order to fill up the latest issue was, to say the least, rather inimical to the production of quality software. Vekovius was spinning his wheels in his little programming sweatshop with all the energy of those Shreveport riverboat gamblers, but it wasn’t at all clear that it was getting him any further than it was getting them.
Thus he was receptive on the day in early 1990 when one of his most productive if headstrong programmers, a strapping young metalhead named John Romero, suggested that Softdisk start a new MS-DOS disk magazine, dedicated solely to games — the one place where, what with Apogee’s success being still in its early stages, shareware had not yet clearly cut into Softdisk’s business model. After some back-and-forth, the two agreed to a bi-monthly publication known as Gamer’s Edge, featuring at least one — preferably two — original games in each issue. To make it happen, Romero would be allowed to gather together a few others who were willing to work a staggering number of hours cranking out games at an insane pace with no resources beyond themselves for very little money at all. Who could possibly refuse an offer like that?
The id boys: John Carmack, Kevin Cloud, Adrian Carmack, John Romero, Tom Hall, and Jay Wilbur.
The team that eventually coalesced around Romero included programmer Tom Hall, artist Adrian Carmack, and business manager and token adult-in-the-room Jay Wilbur. But their secret weapon, lured by Wilbur to Shreveport from Kansas City, Missouri, was a phenomenal young programmer named John Carmack. (In a proof that anyone who says things like “I don’t believe in coincidences” is full of it, John is actually unrelated to Adrian Carmack despite having the same not-hugely-common last name.) John Carmack would prove himself to be such a brilliant programmer that Romero and Hall, no slouches themselves in that department by most people’s standards, would learn to leave the heavy lifting to his genius, coding themselves only the less important parts of the games along with the utilities that they used to build them — and they would also design the games, for Carmack was in reality vastly more interested in the mathematical abstraction of code as an end unto itself than the games it enabled.
But all of these young men, whom I’ll call the id boys from here on out just because the name suited them so well even before they started id Software, will be more or less important to our story. So, we should briefly meet each of them.
Jay Wilbur was by far the most approachable, least intimidating member of the group. Having already reached the wise old age of 30, he brought with him a more varied set of life experiences that left him willing and able to talk to more varied sorts of people. Indeed, Wilbur’s schmoozing skills were rather legendary. While attending university in his home state of Rhode Island, he’d run the bar at his local TGI Friday’s, where his ability to mix drinks with acrobatic “flair” made him one of those selected to teach Tom Cruise the tricks of the trade for the movie Cocktail. But his love for the Apple II he’d purchased with an insurance settlement following a motorcycle accident finally overcame his love for the nightlife, and he accepted a job for a Rhode Island-based disk magazine called UpTime. When that company was bought out by Softdisk in 1988, he wound up in Shreveport, working as an editor there. The people skills he’d picked up tending bar would never desert him; certainly his new charges at Gamer’s Edge had sore need of them, for they were an abrasive collection of characters even by hacker standards.
These others loved heavy metal and action movies, and aimed a well-sharpened lance of contempt at anything outside their narrow range of cultural and technical interests. Their laser focus on their small collection of obsessions would prove one of their greatest strengths, if perhaps problematic for gaming writ large in the long run, in the way that it diminished the scope of what games could do and be.
Yet even this band of four, the ones who actually made the games for Gamer’s Edge under Wilbur’s benevolent stewardship, was not a monolith. Once one begins to look at them as individuals, the shades of difference quickly emerge.
Like Wilbur, the 25-year-old Wisconsinite Tom Hall was a middle-class kid with a university degree, but he had none of his friend and colleague’s casual bonhomie with the masses. He lived in a fantasy world drawn from the Star Wars movies, the first of which he’d seen in theaters 33 times, and the Hitchhiker’s Guide to the Galaxy novels, which he could all but recite from heart. At Softdisk, to which he’d come after deciding that he couldn’t stand the idea of a job in corporate data processing, he ran around talking in a cutsey made-up alien language: “Bleh! Bleh! Bleh!” He was the kind of guy you either found hilarious or were irritated out of your mind by.
The 21-year-old Adrian Carmack also lived in a world of fantasy, but his fantasies had a darker hue. Growing up right there in Shreveport, he had spent many hours at arcades, attracted not so much by the games themselves as by the lurid art on their cabinets. He worked for a time as an aide at a hospital, then went home to sketch gunshot wounds, severed limbs, and festering bedsores with meticulous accuracy. Instead of a cat or a dog, he chose a scorpion as a pet. He’d come to Softdisk on a university internship after telling his advisor he wanted to work in “fine art” someday.
Still, and with all due respect to these others, the id boys would come to be defined most of all by their two Johns. The 22-year-old John Romero was pure id, a kettle of addled energy that was perpetually spilling over, sending F-bombs spewing every which way; David Kushner, author of the seminal history Masters of Doom, memorably describes him as “a human exclamation point.” The not-quite-20-year-old John Carmack was as quiet and affectless as Romero was raucous, often disturbingly so; Sandy Petersen, a game designer who will come to work with him later in our story, remembers musing to himself after first meeting Carmack that “he doesn’t know anything about how humans think or feel.”
Yet for all their surface differences, the two Johns had much in common. Both were brought up in broken homes: Romero was physically abused by his stepfather while growing up in the Sacramento area, while Carmack suffered under the corporeal and psychological rigors of a strict private Catholic school in Kansas. Both rebelled by committing petty crimes among other things; Carmack was sentenced to a year in a boys’ detention center at age 14 after breaking into his school using a homemade bomb. (The case notes of the police officer who interviewed him echo the later impressions of Sandy Petersen: “Boy behaves like a walking brain… no empathy for other human beings.”)
Both found escape from their circumstances through digital means: first via videogames at the local arcades, then via the Apple II computers they acquired by hook or by crook. (Carmack’s first computer was a stolen one, bought off the proverbial back of a truck.) They soon taught themselves to program well enough to put professionals to shame.
Romero got his games published regularly by print magazines as type-in listings, then parlayed that into a job with the disk magazine UpTime, where he became friends with Jay Wilbur. After that, he got a job as a game porter for Origin Systems of Ultima fame. Meanwhile Wilbur moved on to Softdisk while Romero was at Origin. When Romero found himself bored by the life of a porter, he came to Shreveport as well to join his friend.
John Carmack, being more than two years younger than Romero and much more socially challenged, brought a shorter résumé with him to Shreveport when he became the only id boy to be hired specifically to work on Gamer’s Edge rather than being transferred there from another part of Softdisk. He had mostly sold his games for $1000 apiece to a little mom-and-pop company near his home called Nite Owl Productions, who had made them a sideline to their main business of supplying replacement batteries for Apple II motherboards. But he had also sold one or two games to Jay Wilbur at Softdisk. Finding these to be very impressive, the id boys asked Wilbur to deploy his considerable charm to recruit the new kid for Gamer’s Edge. After a concerted effort, he succeeded.
Gamer’s Edge was far more than just a new job or a workplace transfer for the young men involved. It was a calling; they spent virtually all day every day in one another’s company. Pooling all of their meager salaries, Wilbur rented them a rambling old four-bedroom house on the Red River, complete with a Jacuzzi and a swimming pool and a boat deck which he soon complemented with a battered motorboat. It was an Animal House lifestyle of barbecuing, skiing, and beer drinking in between marathon hacking sessions, fueled by pizza and soda. Wilbur — in many ways the unsung hero of this story — acted as their doting den mother, keeping the lights on, the basement beer keg filled, the refrigerator stocked with soda and junk food, and the pizza deliveries coming at all hours of the day and night.
Inside the riverfront house in Shreveport. John Carmack sits near center frame, while John Romero is to his left, mostly hidden behind a pillar.
For the first issue of Gamer’s Edge, the two Johns agreed to each port one of their old Apple II games to MS-DOS. Romero chose a platformer called Dangerous Dave, while Carmack chose a top-down action-adventure called Catacomb. They raced one another to see who could finish first; it was after losing rather definitively that Romero realized he couldn’t hope to compete with Carmack as a pure programmer, and should probably leave the most complicated, math-intensive aspects of coding to his friend while he concentrated on all the other things that make a good game. For the second issue, the two Johns pooled their talents with that of the others to make a completely original shoot-em-up called Slordax: The Unknown Enemy. So far, so good.
And then came John Carmack’s first great technical miracle — the first of many that would be continually upending everything the id boys were working on in the best possible way. To fully explain this first miracle, a bit of background is necessary.
Although they were making games for MS-DOS, the id boys had little use for the high-concept themes of most other games that were being made for that platform in 1990; neither complicated simulations nor elaborate interactive movies did anything for them. They preferred games that were simple and visceral, fast-paced and above all action-packed. Tellingly, most of the games they preferred to play these days lived on the Nintendo Entertainment System rather than a personal computer.
Much of the difference between the two platforms’ design aesthetics was cultural, but there was also more to it than that. As I’ve often taken pains to point out in these articles, the nature of games on any given platform is always strongly guided by that platform’s technical strengths and weaknesses.
When first looking at the NES and an MS-DOS personal computer of 1990 vintage, one might assume that the latter so thoroughly outclasses the former as to make further comparison pointless. The NES was built around a version of the MOS 6502, an 8-bit CPU dating back to the 1970s, running at a clock speed of less than 2 MHz; a state-of-the-art PC had a 32-bit CPU running at 25 MHz or more. The NES had just 2 K of writable general-purpose memory; the PC might have 4 MB or more, plus a big hard drive. The NES could display up to 25 colors from a palette of 48, at a resolution of 256 X 240; a PC with a VGA graphics card could display up to 256 colors from a palette of over 262,000, at a resolution of 320 X 200. Surely the PC could effortlessly do anything the NES could do. Right?
Well, no, actually. The VGA graphics standard for PCs had been created by IBM in 1987 with an eye to presenting crisp general-purpose displays rather than games. In the hands of a talented team of pixel artists, it could present mouth-watering static illustrations, as adventure-game studios like Sierra, LucasArts, and Legend were proving. But it included absolutely no aids for fast animation, no form of graphical acceleration whatsoever. It just gave the programmer a big chunk of memory to work with, whose bytes represented the pixels on the screen. When she wanted to change said pixels, she had to sling all those bytes around by main force, using nothing but the brute power of the CPU. All animation on a PC was essentially page-flipping animation, requiring the CPU to redraw every pixel of every frame in memory, at the 20 or 30 frames per second that were necessary to create an impression of relatively fluid motion, and all while also finding cycles for all of the other aspects of the game.
The graphics system of the NES, on the other hand, had been designed for the sole purpose of presenting videogames — and in electrical engineering, specialization almost always breeds efficiency. Rather than storing the contents of the screen in memory as a linear array of pixels, it operated on the level of tiles, each of which was 8 X 8 or 8 X 16 pixels in size. After defining the look of each of a set of tiles, the programmer could mix and match them on the screen as she wished, at a fairly blazing speed thanks to the console’s custom display circuitry; this enabled the smooth scrolling of the Super Mario Bros. games among many others. She also had up to 64 sprites to work with; these were little 8 X 8 or 8 X 16 images that were overlaid on the tiled background by the display hardware, and could be moved about almost instantaneously, just by changing a couple of numbers in a couple of registers. They were, in other words, perfect for showing Super Mario bouncing around on a scrolling background, at almost no cost in CPU cycles. Freed from the heavy lifting of managing the display, the little 6502 could concentrate almost entirely on the game logic.
The conventional wisdom of 1990 held that the PC, despite all its advantages in raw horsepower, simply couldn’t do a game like Super Mario Bros. The problem rankled John Carmack and his friends particularly, given how much more in tune their design aesthetic was with the NES than with the current crop of computer games. And so Carmack turned the full force of his giant brain on the problem, and soon devised a solution.
As so often happens in programming, said solution turned out to be deceptively simple. It hinged on the fact that one could define a virtual screen in memory that was wider and/or taller than the physical screen. In this case, Carmack made his virtual screen just eight pixels wider than the physical screen. This meant that he could scroll the background with silky smoothness through eight “frames” by changing just two registers on the computer — the ones telling the display hardware where the top left corner of the screen started in the computer’s memory. And this in turn meant that he only had to draw the display anew from scratch every eighth frame, which was a manageable task. Once he had the scrolling background working, he added some highly optimized code to draw and erase in software alone bouncing sprites to represent his pseudo-Mario and enemies. And that was that. His technique didn’t even demand VGA graphics; it could present a dead ringer for the NES Super Mario Bros. 3 — the latest installment in the franchise — using the older MS-DOS graphics standard of EGA.
I should note at this point that the scrolling technique which John Carmack “invented” was by no means entirely new in the abstract; programmers on computers like the Commodore 64 and Commodore Amiga had in fact been using it for years. (I point readers to my article on the techniques used by the Commodore 64 sports games of Epyx and particularly to my book-length study of the Amiga for more detailed explanations of it than the one I’ve provided here.) A big part of the reason that no one had ever done it before on an MS-DOS computer was that no one had ever been hugely motivated to try, in light of the types of games that were generally accepted as “appropriate” for that platform; technological determinism is a potent force in game development, but it’s never the only force. And I should also note a certain irony that clings to all this. As we’ll see, John Carmack would soon toll the death knell for the era of bouncing sprites superimposed over scrolling 2D backgrounds. How odd that his first great eureka moment should have come in imitation of just that classic videogame style.
Carmack first showed his innovation to Tom Hall, the biggest Super Mario fan of all among the id boys, late in the afternoon of September 20, 1990. Hall recognized its significance immediately, and suggested that he and Carmack recreate some of the first level of Super Mario Bros. 3 right then and there as a proof of concept. They finally stumbled off to bed at 5:30 the following morning.
A few hours later, John Romero woke up to find a floppy disk sitting on his keyboard. He popped it into the drive, and his jaw hit the floor when he saw a Nintendo game playing there on his computer monitor. He went off to find Jay Wilbur and Adrian Carmack. They all agreed that this was big — way too big for the likes of Softdisk.
In one 72-hour marathon, the id boys recreated all of the first level of Super Mario Bros. 3, along with bits and pieces of those that followed. Then Wilbur typed up a letter to Nintendo of America and dropped it in the mail along with the disk; it said that the id boys were ready and willing to license their PC port of Super Mario Bros. 3 back to the Nintendo mother ship. This was a profoundly naïve thing to do; virtually anyone in the industry could have told them that Nintendo never let any of their intellectual property escape from the walled garden of their own console. And sure enough, the id boys would eventually receive a politely worded response saying no thank you. Given Nintendo’s infamous ruthlessness when it came to matters of intellectual property, they were probably lucky that a rejection letter was all they received, rather than a lawsuit.
At any rate, the id boys weren’t noted for their patience. Long before Nintendo’s response arrived, they would be on to the next thing: an original game using John Carmack’s scrolling technique.
For some time now, John Romero had been receiving fawning fan mail care of Softdisk, not a usual phenomenon at all. His gratification was lessened somewhat, however, by the fact that the letters all came from the same address near Dallas, Texas, all asked him to call the fan in question at the same phone number, and were all signed with suspiciously similar names: “Byron Muller,” “Scott Mulliere,” etc.
It was in fact our old friend Scott Miller. His attention had been captured by Romero’s games for On Disk and Gamer’s Edge; they would be perfect for Apogee, he thought. But how to get in touch? The only contact information he had was that of Softdisk’s main office. He could hardly write them a letter asking if he could poach one their programmers. His solution was this barrage of seemingly innocent fan mail. Maybe, just maybe, Romero really would call him…
Romero didn’t call, but he did write back, and included his own phone number. Miller rang it up immediately. “Fuck those letters!” he said when Romero started to ask what kind of prank he thought he was pulling. “We can make a ton of money together selling your games as shareware.”
“Dude, those old games are garbage compared to the stuff we can make now,” said Romero, with John Carmack’s new scrolling technique firmly in mind. They struck a deal: Miller would send the id boys an advance of $2000, and they would send him a brand-new three-part game as soon as possible.
The Gamer’s Edge magazine, which just six months ago had seemed like the perfect job, now fell to the back burner in light of the riches Miller was promising them. Since they were making a Nintendo-like game in terms of action, it seemed logical to copy Nintendo’s bright and cheerful approach in the new game’s graphics and fiction as well. This was Tom Hall’s moment to shine; he already seemed to live every day in just such a primary-colored cartoon fantasy. Now, he created an outline for Commander Keen, blending Nintendo with The Hitchhiker’s Guide to the Galaxy and old science-fiction serials — the last being perfect for an episodic game.
Billy Blaze, eight-year-old genius, working diligently in his backyard clubhouse, has created an interstellar spaceship from old soup cans, rubber cement, and plastic tubing. While his folks are out on the town and the babysitter is asleep, Billy sneaks out to his backyard workshop, dons his brother’s football helmet, and transforms into… Commander Keen, Defender of Justice! In his ship, the Bean with Bacon Megarocket, Keen dispenses justice with an iron hand!
In this episode, aliens from the planet Vorticon VI find out about the eight-year-old genius and plan his destruction. While Keen is out exploring the mountains of Mars, the Vorticons steal his ship and leave pieces around the galaxy! Can Keen recover all the pieces of his ship and repel the Vorticon invasion? Will he make it back before his parents get home? Stay tuned!
Commander Keen
When Miller received the first Commander Keen trilogy in the post barely two months later, he was thrilled beyond his wildest dreams. He had known that the id boys were talented, but this… he had never imagined this. This wasn’t a throwback to the boxed games of yore, wasn’t even on a par with the boxed games of current times. It was something entirely different, something never seen on an MS-DOS computer at all before, as visually striking and technically innovative within its chosen sphere as any of the latest boxed games were within theirs. Just like that, shareware games had come of age.
All of Apogee’s games together had been earning about $7000 per month. Commander Keen alone made $20,000 in the first month of its availability. It caused such a stir online that the established industry took a casual notice for the first time of this new entity called Apogee with this odd new way of selling games. Computer Gaming World magazine even deigned to give Commander Keen a blurb in the new-releases section. It was “of true commercial quality,” they noted, only slightly condescendingly.
Despite their success in shareware and the big checks that started coming in the mail from Apogee as a result, the id boys continued to make games for Gamer’s Edge throughout 1991. Betwixt and between, they provided Miller with a second Commander Keen trilogy, which did every bit as well as the first. No one could ever accuse them of being lazy.
But making a metaphorical name for themselves outside of Softdisk meant that they needed a literal name for the world to know them by. When they had sent their Super Mario Bros. 3 clone to Nintendo, they had called themselves “Ideas from the Deep.” Deciding that was too long-winded, they became “ID” when they started releasing games with Apogee — short for “In Demand.” The only one of their number who cottoned onto the Freudian implications of the acronym was Jay Wilbur; none of the other id boys knew Sigmund Freud from Siegmund the Norse hero. But when Wilbur explained to them how Freud’s id was the seat of a person’s most basic, impulsive desires, they were delighted. By this happenstance, then, id Software got a name which a thousand branding experts could never have bettered. It encapsulated perfectly their mission to deconstruct computer gaming, to break it down into a raw essence of action and reaction. The only ingredient still missing from the eventual id Software formula was copious violence.
And that too was already in the offing: Tom Hall’s cheerful cartoon aesthetic had started to wear thin with John Romero and Adrian Carmack long before they sent the first Commander Keen games to Scott Miller. Playing around one day with some graphics for the latest Gamer’s Edge production, Adrian drew a zombie clawing out the eyes of the player’s avatar, sending blood and gore flying everywhere. Romero loved it: “Blood! In a game! How fucking awesome is that?”
Adrian’s reply was weirdly pensive. “Maybe one day,” he said in a dreamy voice, “we’ll be able to put in as much blood as we want.”
In September of 1991, the id boys’ lease on their riverside frat house expired, and they decided that it was time to leave the depressing environs of Shreveport, with its crime, its poverty, and its homeless population who clustered disturbingly around the Softdisk offices. Their contract stipulated that they still owed Gamer’s Edge a few more games, but Al Vekovius had long since given up on trying to control them. The id boys decamped for Madison, Wisconsin, at the suggestion of Tom Hall, who had attended university there. He promised them with all of his usual enthusiasm that it was the best place ever. Instead they found the Wisconsin winter to be miserable. Cooped up inside their individual apartments, missing keenly their big old communal house and their motorboat, they threw themselves more completely than ever into making games. Everyone, with the exception only of Tom Hall, was now heartily sick and tired of Commander Keen. It was time for something new.
Whilst working at Origin Systems in the late 1980s, John Romero had met Paul Neurath, who had since gone on to start his own studio known as Blue Sky Productions. During their occasional phone calls, Neurath kept dropping hints to his friend about the game his people were working on: an immersive first-person CRPG, rendered using texture-mapped 3D graphics. When Romero mentioned it to John Carmack, his reply was short, as so many of them tended to be: “Yeah, I can do that.”
Real-time 3D graphics in general were hardly a new development. Academic research in the field stretched back to well before the era of the microchip. Bruce Artwick had employed them in the original Radio Shack TRS-80 Flight Simulator in 1980, and Ian Bell and David Braben had used them in Elite in 1984; both games were among the best sellers of their decade. Indeed, the genre of vehicular simulations, one of the most popular of them all by the late 1980s, relied on 3D graphics almost exclusively. All of which is to say that you didn’t have to look very hard in your local software store to find a 3D game of some stripe.
And yet, according at least to the conventional wisdom, the limitations of 3D graphics made them unsuitable for the sort of visceral, ultra-fast-paced experience which the id boys preferred. All of the extra affordances built into gaming-oriented platforms like the NES to enable 2D sprite-based graphics were useless for 3D graphics. 3D required radical compromises in speed or appearance, or both: those early versions of Flight Simulator were so slow that it could take the program a full second or two to respond to your inputs, which made flying their virtual airplanes perversely more difficult than flying the real thing; Elite managed to be more responsive, but only by drawing its 3D world using wire-frame outlines instead of filled surfaces. The games-industry consensus was that 3D graphics had a lot of potential for many types of games beyond those they were currently being used for, but that computer hardware was probably five to ten years away from being able to realize most of it.
John Carmack wasn’t that patient. If he couldn’t make true 3D graphics run at an acceptable speed in the here and now, he believed that he could fake it in a fairly convincing way. He devised a technique of presenting a fundamentally 2D world from a first-person perspective. Said world was a weirdly circumscribed place to inhabit: all angles had to be right angles; all walls had to stretch uniformly from floor to ceiling; all floors and ceilings had to be colored in the same uniform gray. Only interior scenes were possible, and no stairways, no jumping, no height differences of any kind were allowed; in this egalitarian world, everything and everyone had to stay permanently on the same level. You weren’t even allowed to look up or down. But, limited though it was, it was like nothing anyone had ever seen.
“You know,” said John Romero one day when they were all sitting around discussing what to do with the new technology, “it’d be really fucking cool if we made a remake of Castle Wolfenstein and did it in 3D.” With those words, id’s next game was born, one that would make all the success of Commander Keen look like nothing.
The original Castle Wolfenstein.
Written by Silas Warner, one of the Apple II scene’s early superstar programmers, and published by the long-defunct Muse Software, Castle Wolfenstein was an established classic from 1981, a top-down action-adventure that cast you as a prisoner of the Nazis who must escape, preferably taking his captors’ secret war plans with him. It remains historically notable today for incorporating a significant stealth component; ammunition was scarce and your enemies tough, which often made avoidance a better strategy than confrontation.
But avoidance wasn’t the id boys’ style. Very early on, they jettisoned everything beyond the core theme of the original Castle Wolfenstein. Wolfenstein 3D was to be, as Romero put it, “a totally shocking game. There should be blood, lots of blood, blood like you never see in games. When the player gets really low in health, at like 10 percent, he could run over the bloody guts of a dead Nazi soldier and suck those up for extra energy. It’s like human giblets. You can eat up their gibs!” In other words, Tom Hall’s aesthetic vision was out; John Romero and Adrian Carmack’s was in. “Hey, you know what we should have in here? Pissing! We should make it so you can fucking stop and piss on the Nazi after you mow him down! That would be fucking awesome!”
In early 1992, the id boys came face to face with the gaming establishment for the first time thanks to Wolfenstein 3D. They sent an early demo of the game to Sierra, and that company’s founder and CEO Ken Williams invited them to fly out to California and have a chat. Sierra was one of the three biggest computer-game publishers in the world, and was at the forefront of the interactive-movie trend which the id boys loathed. King’s Quest VI, the upcoming new installment in Sierra’s flagship series, would be so weighted down with multimedia that most reviewers, hopelessly dazzled, could spare only a few sentences for the rather rote little adventure game underneath it all. Williams himself was widely recognized as one of the foremost visionaries of the new era, proclaiming that by the end of the decade much or most of the Hollywood machine would have embraced interactivity. A meeting between two more disparate visions of gaming than his and that of the id boys can scarcely be imagined.
And yet the meeting was a cordial one on the whole. Williams had been quick to recognize when he saw Wolfenstein 3D that id had some remarkable technology, while the id boys remembered the older Apple II games of Sierra fondly. Williams took them on a tour of the offices where many of those games had come from, and then, after lunch, offered to buy id Software for $2.5 million in Sierra stock. The boys discussed it for a bit, then asked for an additional $100,000 in cash. Williams refused; he was willing to move stock around to pay for the Wolfenstein 3D technology, but he wasn’t willing to put his cash on the table. So, the negotiation ended. Instead Williams bought Bright Star Technologies, a specialist in educational software, for $1 million in cash later that year — for educational software, he believed, would soon be bigger than games. Time would prove him to be as wrong about that as he was about the future of Hollywood.
Not long after the Sierra meeting, the id boys left frigid Wisconsin in favor of Dallas, Texas, home of Scott Miller, who had been telling them about the warm weather, huge lakes, splendid barbecue, and nonexistent state income tax of the place for more than eighteen months now. One Kevin Cloud, who had held the oft-thankless role of being the id boys’ liaison with Softdisk but also happened to be a talented artist, joined them in Dallas as a sixth member of their little collective, thereby to relieve some of the burden on Adrian Carmack.
After making the move, they broke the news to Softdisk that they wouldn’t be doing Gamer’s Edge anymore. Al Vekovius was disappointed but not devastated. Oddly given how popular Commander Keen had become, the gaming disk magazine had never really taken off; it still only had about 3000 subscribers.
And so Softdisk Publications of Shreveport, Louisiana, that unlikely tech success story in that most unlikely of locales, finally exits our story permanently at this point. Nothing if not a survivor, Vekovius would keep the company alive through the 1990s and beyond by transitioning into the next big thing in computing: he turned it into an Internet service provider. He was bought out circa 2005 by a larger regional provider.
Wolfenstein 3D
This screenshot of the Wolfenstein 3D map editor illustrates why the game’s name is a misnomer: the environment is really a 2D maze much like that of the original game, albeit shown from a first-person perspective. At bottom, the engine understands just two dimensions rather than three.
If the id boys were worried about how Scott Miller would react to the ultra-violence of Wolfenstein 3D, they needn’t have been. Apogee had already been moving in this direction with considerable success; their only game to rival Commander Keen in sales during 1991 had been Duke Nukem by Todd Replogle, whose titular protagonist was a cigar-chomping Arnold Schwarzenegger facsimile with a machine gun almost as big around as his biceps. When Miller saw Wolfenstein 3D for the first time, he loved the violence as much as he did John Carmack’s pseudo-3D graphics engine. He knew what his customers craved, and he knew that they would swoon over this. He convinced the id boys to make enough levels to release a free episode followed by five paid ones rather than the usual two. On May 5, 1992 — the very same day on which the boys had handed the final version to Miller — the free installment appeared on Software Creations, Apogee’s new online service.
As it happened, Paul Neurath’s Blue Sky Productions had released their own immersive first-person 3D game, which had spent roughly five times as long in production as Wolfenstein 3D, just two months before. It was called Ultima Underworld, and was published as a boxed product by Origin Systems. It boasted a far more complete implementation of a 3D world than did id’s creation. You could look up, down, and all around; could jump and climb ledges; could sneak around corners and hide in shadows; could swim in rivers or fly through the air by means of a levitation spell. But Ultima Underworld was cerebral, old school — dull, as the id boys and many of their fan base saw it. Combat was only a part of its challenge. You also had to spend your time piecing together clues, collecting spells, solving puzzles, annotating maps, leveling up and assigning statistics and skills to your character. Even the combat happened at a speed most kindly described as “stately” if you didn’t have a cutting-edge computer.
Wolfenstein 3D, by contrast, ran like greased lightning on just about any computer, thanks to John Carmack’s willingness to excise any element from his graphics engine that he couldn’t render quickly. After all, the id boys really only wanted to watch the blood spurt as they mowed down Nazis; “just run over everything and destroy” was their stated design philosophy. And many others, it seemed, agreed with their point of view.
For, while Ultima Underworld became a substantial hit, Wolfenstein 3D became a phenomenon. It made $200,000 in the first month, then kept selling at that pace for the next eighteen months. It was, as Scott Miller would later put it, a “paradigm shift” in shareware games. Whatever that elusive “it” was that so many gamers found to be missing in the big boxed offerings — immediacy? simplicity? violence? id in the Freudian sense? all of the above? — Wolfenstein 3D had it in spades.
The shareware barbarians were truly at the gates now; they could no longer be ignored by the complacent organs of the establishment. This time out, id got a feature review in Computer Gaming World to go along with the full-page color advertisements which Apogee was now able to pay for. “I can’t remember a game making such effective use of perspective and sound and thereby evoking such intense physiological responses from its player,” the review concluded. “I recommend gamers take a look at this one, if only for a cheap peek at part of interactive entertainment’s potential for a sensory-immersed ‘virtual’ future.”
Yet, as that “if only” qualifier intimates, the same magazine was clearly bothered by all of the gleefully gory violence of the game. An editorial by editor-in-chief Johnny Wilson, the former pastor who had built Computer Gaming World into the most thoughtful and mature journal in the industry, drove the point home: “What are we saying when we depict lifelike carnage in a game where the design is geared for you to kill nearly everyone you encounter?”
If Wilson thought id’s first 3D shooter was disturbing, he hadn’t seen anything yet. Their next game would up the ante on the violence and gore even as their first competitors jumped into the act, starting a contest to see who could be most extreme. Everyone working in games or playing them would soon have to reckon with the changes — distributional, technical, and cultural — which a burgeoning new genre, born on the streets instead of in the halls of power, was wreaking.
Crashing the halls of power: Tom Hall, Jay Wilbur, and John Romero in black tie for the Shareware Industry Awards of 1992.
(Sources: the books Masters of Doom by David Kushner, Game Engine Black Book: Wolfenstein 3D by Fabien Sanglard, Principles of Three-Dimensional Computer Animation by Michael O’Rourke, Sophistication & Simplicity: The Life and Times of the Apple II Computer by Steven Weyhrich, and I Am Error by Nathan Altice; PC Magazine of September 12 1989; InfoWorld of June 12 1989; Retro Gamer 75; Game Developer premiere issue and issues of June 1994 and February/March 1995; Computer Gaming World of August 1991, January 1992, August 1992, and September 1992; The Computist 88; inCider of November 1989. Online sources include “Apogee: Where Wolfenstein Got Its Start” by Chris Plante at Polygon, “Rocket Jump: Quake and the Golden Era of First-Person Shooters” by David L. Craddock at Shack News, Samuel Stoddard’s Apogee FAQ, Benj Edwards’s interview with Scott Miller for Gamasutra, Jeremy Peels’s interview with John Romero for PC Games N, Lode Vandevenne’s explanation of the Wolfenstein 3D rendering engine, and Jay Wilbur’s old Usenet posts, which can now be accessed via Google Groups.
The company once known as Apogee, which is now known as 3D Realms, has released many of their old shareware games for free on their website, including Commander Keen. All of the Wolfenstein 3D installments are available as digital purchases at GOG.com.)
The Shareware Scene, Part 2: The Question of Games
In one of the last interviews he gave before his death, shareware pioneer Jim Button said that he “had written off the idea of shareware games” prior to the beginning of the 1990s. At the time, it seemed a reasonable position to take, one based on quite a bit of evidence. While any number of people had tried to sell their games this way, there had been no shareware success stories in games to rival those of Andrew Fluegelman, Jim Button, or Bob Wallace.
Naturally, many pondered why this should be so. The answers they came up with were often shot through with the prejudices of the period, which held that programming or playing frivolous games was a less upstanding endeavor than that of making or using stolid business software. Still, even the prejudiced answers often had a ring of truth. You had a long-term relationship with your telecommunications program, database, or word processor, such that sending its author a check in order to join the mailing list, acquire a printed manual, and be assured of access to updates felt as much like a wise investment as merely “the honest thing to do.” But you had a more transient relationship with games; you played a game only until you beat it or got tired of it, then moved on to the next one. Updates and other forms of long-term support just weren’t a factor at all. No one could seem to figure out how to untangle this knot of motivation and contingency and make shareware work for games.
Luckily, there was an alternative to the shareware model for those game programmers who lacked the right combination of connections, ambitions, and talents to go the traditional commercial route — an alternative that offered a better prospect than shareware during the 1980s of getting paid at least a little something for one’s efforts. It was the odd little ghetto of the disk magazines, and so it’s there that we must start our story today.
The core idea behind the disk magazines is almost as old as personal computing itself. In February of 1978, Ralph McElroy of Goleta, California, published the first issue of CLOAD, a monthly collection of software for the Radio Shack TRS-80, the first pre-assembled microcomputer to rack up really impressive sales numbers. “To join the somewhat elite club of computer users,” wrote McElroy in his introductory editorial, “one [previously] had to learn the mysterious art of speaking in a rather obscure tongue” — i.e., one had to learn to program. Before any commercial software industry to speak of existed, CLOAD proposed to change that by offering “vast quantities of software to be shared.” It was actually distributed on cassette tape rather than floppy disk — a disk drive was still a very exotic piece of hardware in 1978 — but otherwise it put all the pieces into place.
By 1981, the TRS-80’s early momentum was beginning to flag and the more capable Apple II was coming on strong. Jim Mangham, a programmer at the Louisiana State University Medical Center in Shreveport, decided that the market was ready for a CLOAD equivalent for the Apple II — albeit published not on cassettes but on floppy disks, which were now steadily gaining traction. He recruited a buddy named Al Vekovius to join him in the venture, and the two prepared the first issue of something they called The Harbinger. They called up Softalk magazine, the journal of record for early Apple II users, to discuss placing an advertisement, whereupon said magazine’s founder and editor Al Tommervik got so excited by their project that he asked to become an investor and official marketing partner. Thus The Harbinger acquired the rather less highfalutin name of Softdisk to connote its link with the print magazine.
Starting with just 50 subscribers, Mangham and Vekovius built Softdisk into a real force in Apple II computing. Well aware that they couldn’t possibly write enough software themselves to fill a disk every single month, they worked hard from the beginning to foster a symbiotic relationship with their readership; most of the programs they published came from the readers themselves. In the early days, the spirit of reciprocity extended to the point of expecting readers to mail their disks back each month; this both allowed Mangham and Vekovius to save money on media and provided a handy way for readers to send in their programs and comments. Even after this practice was abandoned in the wake of falling disk prices, Softdisk subscribers felt themselves to be part of a real digital community, long before the rise of modern social media made such things par for the course. At a time when telecommunications was a slow, difficult, complicated endeavor, Softdisk provided an alternative way of feeling connected with a larger community of people who were as passionate as oneself about a hobby which one’s physical neighbors might still regard as hopelessly esoteric.
Thus Mangham and Vekovius’s little company Softdisk Publishing slowly turned into a veritable disk-magazine empire. In time, Mangham stepped back from day-to-day operations, becoming a nearly silent partner to Vekovius, always the more business-focused of the pair. He expanded Softdisk to two disks per issue in August of 1983; started reaching retail stores by January of 1984; launched a companion disk magazine called Loadstar for the Commodore 64 in June of 1984. Softdisk survived the great home-computer bust of the second half of 1984, which took down Softalk among many other pioneering contemporaries, then got right back to expanding. In November of 1986, Vekovius launched a third disk magazine by the name of Big Blue Disk, for MS-DOS-based computers; it soon had a monthly circulation of 15,000, comparable to that of Softdisk and Loadstar. A fourth disk magazine, for the Apple Macintosh this time, followed in 1988. At least a dozen competitors sprang up at one time or another with their own disk magazines, but none seriously challenged the cross-platform supremacy of the Softdisk lineup.
In order to encourage software submissions, all of the Softdisk magazines ran a periodic programming competition called CodeQuest. Readers were encouraged to send in programs of any type, competing for prizes of $1000 for the top submission, $500 for second place, and $250 for third place, on top of the money Softdisk would pay upon eventually publishing the winning software. Big Blue Disk‘s second incarnation of the contest ended on January 31, 1988, yielding two winners that were fairly typical disk-magazine fare: the gold-winning The Compleat Filer was a file-management program to replace the notoriously unfriendly MS-DOS command line, while the bronze-winning Western was a sort of rudimentary text-based CRPG set in, you guessed it, the Old West. But it was the silver winner — a game called Kingdom of Kroz, submitted by one Scott Miller from a suburb of Dallas, Texas — that interests us today.
At the time of the contest, Miller didn’t seem to be going much of anywhere in life. In his late twenties, he was still attending junior college in a rather desultory fashion whilst working dead-end gigs at the lower end of the data-processing totem pole, such as babysitting his college’s computer lab. His acquaintances hardly expected him to ever move out of his parents’ house, much less change an industry. Yet this seeming slacker had reserves of ambition, persistence, marketing acumen, and sheer dogged self-belief that would in the end prove a stick in the eye to every one of his doubters. Scott Miller, you see, wanted to make money from videogames — make a lot of money. And by God, he was going to find a way to do it.
The young Scott Miller.
Before entering the CodeQuest contest, he’d written a column on games for the local newspaper, written a book on how to beat popular arcade games, and, last but not least, tested the early shareware market for games: he’d written and distributed a couple of shareware text adventures under the name of Apogee Software — a name which would later become very, very famous among a certain segment of gamers. But on this occasion he was disappointed by the response, just like everyone else making shareware games at the time. Unlike most of those others, though, Miller didn’t give up. If shareware text adventures wouldn’t do the trick, he’d just try something else.
Put crudely, Kingdom of Kroz was a mash-up of the old mainframe classic Rogue and the arcade game Gauntlet — or, if you like, a version of Rogue that played in real time and had handcrafted levels instead of procedurally-generated ones. It wasn’t much to look at — like classic Rogue, it was rendered entirely in ASCII graphics — but many people found it surprisingly addictive once they got into it. It went over very well indeed with Big Blue Disk‘s subscribers when it appeared in the issue dated June 1988 — went over so well that Miller provided two sequels, called Dungeons of Kroz and Caverns of Kroz, almost immediately, although the magazine wouldn’t find an opening for them in its editorial calendar until the issues dated March and September of 1989.
While he waited on Big Blue Disk to release those sequels, Miller started to explore a new idea for marketing games outside the traditional publishing framework. In fact, this latest idea would eventually prove his greatest single stroke of marketing genius, even if its full importance would take some time yet to crystallize. He would later sum up his insight in an interview: “People aren’t willing to pay for something they’ve already got in their hands, but they are willing to pay if it gets them something new.” Call it a cynical notion if you must, but, in the context of games at least, it would prove the only way to make shareware pay on a scale commensurate with Scott Miller’s ambitions.
Miller and George Broussard, his longtime best friend and occasional partner in the treacherous world of shareware, made an engine for multiple-choice trivia games — not exactly a daunting programming challenge after the likes of Kroz. They compiled sets of questions dealing with different topics: general trivia, vocabulary, the original Star Trek and Star Trek: The Next Generation. They created “volumes” in each category consisting of 100 questions. Then they released the first volume of each category online, accompanied by an advertisement for additional volumes for the low, low price of $4 each.
Alas, the scheme proved not to be a surefire means of selling trivia games; the economics of getting just 100 questions for $4 were perhaps a bit dodgy even in the late 1980s, when just about everything involving computers cost exponentially more than it does today. But a seed had been planted; the next time Miller tried something similar, he would finally hit pay dirt.
The next time in question came in the second half of 1989, just after Big Blue Disk published the last Kroz game. The magazine’s contract terms were far more generous than those of any traditional software publisher: Miller had retained the Kroz copyright throughout, and the magazine’s license to it became non-exclusive as soon as it published the third and last game of the trilogy. Miller, in other words, could now do whatever he wished with his three Kroz games, while still benefiting from the buzz their appearance in Big Blue Disk had caused in some quarters.
Kingdom of Kroz
So, he decided to try the same scheme he had used with his trivia games: release the first part of the trilogy for free, but ask people to send him $7.50 each for the second and third parts. A tactic that had prompted an underwhelming response the first time around worked out much better this time. Unlike those earlier exercises in multiple choice, the Kroz trilogy was made up of real games — or, perhaps better said, was actually one real game artificially divided into three. After you’d played the first part of said game, you wanted to see the rest of it through.
In short, Scott Miller — and shareware gaming in general — finally got their equivalent to that day when Jim Button returned home from a Hawaiian vacation to find his basement drowning in paid registrations. Suddenly Miller as well was drowning in mail, making thousands of dollars every month. He’d done it; his dogged persistence had paid off. He’d found a way around the machinations of the big publishers, found a way to sell games on his own terms, cracked the code of shareware gaming. His sense of vindication after so many years of struggle must defy description.
From here, things happened very, very quickly. Miller whipped up a second trilogy of Kroz games to sell under the same model — first part free, second and third must be paid for — and was rewarded with more checks in the mail. Most people at this point would have been content to continue writing lone-wolf games and reaping huge rewards — but Miller was, as I’ve already noted, a man of unusual ambition. At heart, he was more passionate about marketing games than programming them; in fact, he would never program another game at all after the second Kroz trilogy.
Already before 1989 was over, he had reached out to a Silicon Valley youth named Todd Replogle, who had created and uploaded to various bulletin-board systems a little action-adventure called Caves of Thor that was similar in spirit to the Kroz games. Miller convinced Replogle to re-release his free game under the Apogee imprint, and to make two paid sequels to accompany it. Replogle followed that trilogy up with a tetralogy called Monuments of Mars. Meanwhile George Broussard returned on the scene to make two more four-volume series, called Pharaoh’s Tomb and Arctic Adventure.
By 1991, Apogee was off and running as a real business. Miller quit his dead-end day jobs, moved out of his parents’ house, convinced Broussard to join him as a full-time partner, found an accountant, leased himself an office, and started hiring helpline attendants and clerical help to deal with a workload that was mushrooming for all the right reasons. His life had undergone a head-spinning transformation in the span of less than two years.
At this point, then, we might want to ask ourselves in a more holistic way just why Apogee became so successful so quickly. Undoubtedly, a huge part of the equation is indeed the much-vaunted “Apogee model” of selling shareware: hook them with a free game, then reel them in with the paid sequels. Yet that wasn’t a silver bullet in and of itself, as Miller’s own early lack of success with his trivia games illustrates. It had to be executed just right — which tells us that Miller got it just right the second time around. The price of $7.50 was enough to make the games extremely profitable for Apogee in relation to the negligible amounts of money it took to create and market them, but cheap enough that customers could take the plunge without feeling guilty about it or needing to justify it to a significant other. Likewise, each game was perfectly calibrated to be just long enough for the customer not to feel cheated, but not so long that she spent hours playing it which she could have sunk into another Apogee game.
If all of this sounds a bit mercenary, so be it; Miller was as hard-nosed as capitalists come, and he certainly wasn’t running Apogee as a charity. Yet it’s seldom good business, at least in the long run, to sell junk, and this too Miller understood. Apogee maintained a level of quality control that was often lacking even from the big publishers, who often felt compelled to release a game before its time to meet the Christmas market or to pump up the quarterly numbers. Apogee games, on the other hand, seldom appeared under a Christmas tree, and Miller had no shareholders other than his best friend to placate. “Our philosophy is never to let an arbitrary date dictate when we release a game,” said Miller in an interview. As a result, their games were small but also tight: bug-free, stable, consistent. They evinced a sense of care, felt like creations worth paying a little something for. Soon enough, people learned that they could trust Apogee. If none of Apogee’s early games were revolutionary advances within the medium, there were few to no complete turkeys among them either.
I’ll be the first to admit that the Apogee style of game does little for me. Still, my personal tastes in no way blind me to the reality that these unprepossessing but well-crafted little games filled a space in the market of the early 1990s that the big publishers were missing entirely as they rushed to cement a grand merger of Silicon Valley and Hollywood and begin the era of the “interactive movie.” While the boxed-games industry went more and more high-concept, with prices and system requirements to match, Apogee kept things simple and fun, as befit their slogan: “Apogee means action!” Apogee games were quick to play, quick to get in and out of; they had some of the same appeal that the earliest arcade games had, albeit implemented in a more user-friendly way, with the addictive addition of a sense of progression through their levels. The traditional industry regarded this sort of thing as hopelessly passé on a personal computer, suitable only for videogame consoles like the Nintendo Entertainment System. But, as the extraordinary success of Nintendo and the only slightly less extraordinary success of Apogee both demonstrated, people still wanted these sorts of games. Their near-complete absence from the boxed-computer-game market left a massive hole which Scott Miller was happy to fill. Younger people with limited disposable income found Apogee particularly appealing; they could buy six or seven Apogee games for the price of one boxed production that would probably just bore them anyhow.
But of course a business model as profitable as Miller’s must soon attract rivals who hope to execute it even better. Already in 1992, a company called Epic MegaGames appeared to challenge Apogee for the title of King of Shareware; they as well employed Scott Miller’s episodic approach, and also echoed Apogee’s proven action-first design aesthetic. Shareware gaming was becoming a thriving shadow industry of its own, right under the noses of the big boys who were still chasing after their grand cinematic fantasias. They would have gotten the shock of their lives if they had ever bothered to compare their slim profit margins to the fat ones of Apogee and Epic. As it was, though, they felt nary an inkling in their ivory towers that a proletarian revolution in ludic aesthetics was in the offing out there on the streets. But even they wouldn’t be able to ignore it for much longer.
This shareware sales chart from July of 1993 shows how dominant Apogee was at that time. Seven out of the top ten games are theirs, with a further two going to Epic MegaGames, their only remotely close competitor. Although the fast-and-simple design aesthetic in which those companies specialized ruled the charts, they pulled with them a long tail of many other types of shareware games, as we’ll see in the next part of this article. The very fact that there existed a sales chart like this one at all says much about how quickly shareware had exploded in a very short time.
Many of you doubtless have an inkling already of where this series of articles must go from here — of how not only the story of Apogee Software but also that of Softdisk Publications will feed directly into that of the most transformative computer game in history. And never fear, I’ll get to all of that — but in my next article rather than this one.
For in addition to that other story which threatens to suck all the oxygen out of the room, there are a thousand other, smaller ones of individual creators being inspired to program all kinds of games and sell them as shareware in the wake of Apogee’s success. Exactly none of them made as much money from their endeavors as did Scott Miller, but some became popular enough to still be remembered today. Indeed, many of us who were around back then still have our obscure little hobby horses from the shareware era that we like to take out and ride from time to time. My personal favorite of the breed might just be Pyro II, a thunderously non-politically-correct puzzle game in which you play a pyromaniac who must burn down famous buildings all over the world. Truly, though, the list of old shareware games that come up in any given discussion is guaranteed to be almost as long as the list of old-timers reminiscing about them. The shareware gaming scene in the aggregate which took off after Apogee’s success touched a lot of people’s lives, regardless of how much money this or that individual game might have earned.
Like the Apogee games, many other shareware titles identified holes in the market which the big publishers, who all seemed to be rushing hell-bent in the exact same direction, were failing to fill. In many cases, these were genres from which the traditional industry had actually done very well in the past, but which it had now judged to no longer be worth its while. For example, the years between the collapse of Infocom in 1989 and the beginning of the Internet-based Interactive Fiction Renaissance circa 1995 were marked by quite a number of shareware text adventures. Likewise, as boxed CRPGs got ever more plot- and multimedia-heavy at the expense of the older spirit of free-form exploration, other shareware programmers rushed to fill that gap. Still others mimicked the look and feel of the old ICOM Simulations graphic adventures, while lots more catered to the eternal need just to blow some stuff up after a long, hard day. There were shareware card games, board games, strategy games, fighting games, action puzzlers, proto-first-person shooters of various stripes, and even ballistics simulators.
In terms of presentation, most of these shareware games were dead ringers for the games that had been sold on store shelves five to ten years earlier. And by the same token, the people who made them in the 1990s were really not all that different from the bedroom programmers who had built the boxed-games industry in the 1980s. Just as many creators of non-game shareware were uncomfortable with time-limited or otherwise crippled software, not all creators of shareware games embraced the Apogee model — not even after it had so undeniably demonstrated its efficacy. Even then, some idealistic souls were still willing to place their faith in people sending in checks simply because it was the right thing to do. All of which is to say that shareware gaming encompassed a vast swath of motivations, styles, and approaches. Apogee, Epic, and that other company which we’ll get to in my next article tend to garner all the press when the early 1990s shareware scene is remembered today, but they were by no means the sum total of its personality.
By way of illustration, I’d like to conclude this article with a short case study of a shareware partnership that didn’t make its principals rich, that didn’t even allow them to quit their day jobs. In fact, neither partner ever really even tried to achieve either of those things. They just made games in two unfashionable styles which they still happened to love, and said games made some other people with the same tastes very happy. And that was more than enough for Daniel Berke and Matthew Engle.
Excelsior Phase I: Lysandia
Matthew remembers his best childhood Christmas ever as the one in 1983, when he was twelve years old and his family got an Apple IIe computer. A sheet of Apple-logo stickers came in the box that housed the computer, and Matthew stuck one of them on his notebook. Soon Daniel, another student at his Los Angeles-area school, noticed the sticker and came over to chat. “I’ve got an Apple II also!” he said. Just like that, a lifelong friendship was born.
The two joined an informal community of fellow travelers, the likes of which could be found in school cafeterias and playgrounds all over the country, swapping tips and exploits and most of all games. Their favorites of the games they traded were the text adventures of Infocom and the Ultima CRPGs of Origin Systems; if the pair’s friendship was born over the Apple II, it was cemented during the many hours they spent plumbing the depths of Zork together. Matthew and Daniel eventually joined the minority of kids like them who took the next step beyond playing and trading games: they started to experiment with making them. Their roles broke down into a classic game-development partnership: the analytical Daniel took to programming like a duck takes to water, while the more artistically-minded Matthew was adept at drawing and storytelling.
So many things in life are a question of timing — not least the careers of game developers. One story which Matthew Engle shared with me when I interviewed him in preparation for this article makes that point disarmingly explicit. In 1986, Daniel, Matthew, and another friend created a BASIC text adventure called Zapracker, which they attempted to sell through their local software stores. Matthew:
We made our own boxes and packaged the game with the floppy disk and the manual, just like Richard Garriott did back in the day. Our box was designed to hang on a peg in a software store. We got on a bus with 25 or so copies and visited a few different stores. We’d say, “Hey, would you like to sell this on consignment? You get half the money and we get half.” A few stores took us up on it, and we sold a few copies.
Zapracker (A Lost Classic?)
This tale is indeed almost eerily similar of that of Richard Garriott selling a Ziploc-bagged Akalabeth through his local Computerland just six years earlier; if anything, our heroes in 1986 would appear to have put more effort into their packaging, and perhaps into their game as well, than Garriott did into his. But in the short span of barely half a decade, the possibility of parlaying a homemade game hanging on a rack in a local computer store into an iconic franchise had evaporated. Instead Daniel and Matthew would have to go another route.
Their game-making efforts were growing steadily more sophisticated, as evinced by Daniel’s choice of programming languages: after starting off in Apple II BASIC, he moved on to an MS-DOS C compiler. Adopting unknowingly the approach that had already been used by everyone from Scott Adams to Infocom, from Telarium to Polarware to Magnetic Scrolls, Daniel wrote an interpreter in C which could present a text adventure written in a domain-specific language of his own devising. Matthew then wrote most of the text for what became Skyland’s Star, a science-fiction scenario.
During the pair’s last year in high school, the Los Angeles school district and the manufacturing conglomerate Rockwell International co-sponsored a contest for interesting student projects in computer science. Once Daniel and Matthew decided to enter it, it gave them a thing which many creators find invaluable: a deadline. They finished up their game, and submitted it alongside the technological framework that enabled it. They were soon informed that their project was among the finalists, and were invited to a dinner and awards ceremony at a fancy hotel. Matthew:
All of the finalists were there, demonstrating their entries. We did a couple of interviews for a local TV station. Then the dinner started. They started running down the list of winners, and before we knew it, it was down to two finalists: my and Dan’s project and another one. Then they announced the other one as second place; we had won. It was quite a night!
Matthew Engle and Daniel Berke win the contest with Skyland’s Star in 1989. That’s Daniel’s Apple II GS running the game; he wrote it on that machine in MS-DOS via a PC Transporter emulator card.
Daniel and Matthew gave little initial thought to monetizing their big win. After finishing high school in 1989, they went their separate ways, at least in terms of physical location: Daniel moved to New York to study computer science, while Matthew stayed in Los Angeles to study film. But they kept in touch, and soon started talking about making another game, this time in the spirit of their other favorite type from the 1980s: an old-school Ultima.
It was 1991 by now, and, fed by the meteoric success of Apogee, shareware games of many different stripes were appearing. Daniel and Matthew as well finally caught the fever. They belatedly released Skyland’s Star as shareware for $15, using it as a sort of test case for the eventual marketing of their Ultima-alike. They were among those noble or naïve souls who eschewed the Apogee model in favor of releasing their whole game at once. Instead of offering the rest of the game as an enticement, Daniel and Matthew offered a printed instruction manual, hint book, and map — nice things to have, to be sure, but perhaps not things that played on the psychological compulsions of gamers so powerfully as the literal rest of a game which they dearly wanted to finish. Daniel and Matthew weren’t overwhelmed with registrations.
Progress on the Ultima-like game, which was to be called Excelsior Phase I: Lysandia, was inevitably slowed by their respective university studies; the biggest chunk of the work got done in the summers of 1991, 1992, and 1993, when Daniel came back to Los Angeles and they both had more free time. Then they would sit for hours many days at their favorite pizza restaurant, sketching out their plans. Matthew did most of the scenario design, graphics, and writing, while Daniel did all of the programming.
Calling themselves by now 11th Dimension Entertainment, they finished and released Excelsior in 1993 as shareware, with a registration price of $20. Once again, they relied on a manual, a hint book, and a map alongside players’ consciences to convince them to register. Although it certainly didn’t become an Apogee-sized success story, Excelsior did garner more attention and registrations than had Skyland’s Star. It was helped not only by its being in a (marginally) more commercially viable genre, but also by its coming into a world that was just on the cusp of the Internet Revolution, with the additional distribution possibilities which that massive change to the way that everyday people used their computers brought with it.
As they were finishing Excelsior, Daniel and Matthew had also been finishing their degree programs. Daniel got a programming job at Electronic Arts after a few false starts, while Matthew started a career in Hollywood that would put him, ironically given the retro nature of Excelsior, on teams making cutting-edge CD-ROM-enabled multimedia products at companies like Disney Interactive. Despite their busy lives, they were both still excited enough by independent game development, and gratified enough by the response to Excelsior I, that they embarked on a sequel in 1994. Whereas Excelsior I had aimed for a point somewhere between Ultima IV and Ultima V, Excelsior II took Ultima VI as its model, with all of the increased graphics sophistication that would imply. For this reason not least, the partners wound up spending fully five years making it, communicating almost entirely electronically.
The sheer quantity of labor which Matthew in particular put into this retro-game with limited commercial prospects could have been motivated only by love. Matthew:
We went all out. I ultimately made about 3800 16 X 16-pixel tiles. It was an exhausting process. For every tile, I had to specify whether you could walk on it or it would block you. There was also transparency; we had layers of tiles, overlaid upon one another. There might be a grass tile, then the player-character tile. Then, if you’re walking through a doorway, for example, the arch at the top of the doorway.
Then, after that exhausting process, began the arduous process of putting the tiles down to create the map, which was 500 X 500 tiles if I’m not mistaken — so, 250,000 tiles to place. Plus all of the town and castle and dungeon maps had to be created.
By the time they released Excelsior Phase II: Errondor in 1999, software distribution had changed dramatically from what it had been six years before. It was now feasible to accept credit-card registrations online, and to offer registrants the instant satisfaction of downloadable PDF documents and the like. The motivating ethic of the original shareware movement was alive and well in its way, but, just as with other types of software, the phrase “shareware games” was soon to fall out of use. The more tactile, personal side of the shareware experience, entailing mailed checks, documents, and disks, had already mostly faded into history. Excelsior II did reasonably well for a niche product in this brave new world, but even before its release Daniel and Matthew knew that it would be their last game together. “We realized we just didn’t have it in us to do an Excelsior III,” says Matthew.
In the end, the two of them sold roughly 500 copies each of Excelsior I and II — “small potatoes” by any standard, as Matthew freely admits. He believes that they made perhaps $5000 to $10,000 in all on their games, after the cost of postage and all those printed manuals was subtracted.
I must confess that I personally have some reservations about the 11th Dimension games. It seems to me that Skyland’s Star‘s scenario isn’t quite compelling enough to overcome the engine’s limited parser and lack of player conveniences, and that the Excelsior games, while certainly expansive and carefully put-together, rely a bit too much on needle-in-the-haystack hunting over their enormous maps. Then again, though, I have the exact same complaints about the Ultima games which Excelsior emulates, which would seem to indicate that Daniel and Matthew actually achieved their goal of bringing old-school Ultima back to life. If you happen to like those Ultima games a little more than I do, in other words, you’ll probably be able to say the same about the Excelsior games. One thing that cannot be denied is that all of the 11th Dimension games reflect the belief on the part of their makers that anything worth doing at all is worth doing well.
Shareware gave a place for games like those of Daniel and Matthew to live and breathe when the only other viable mode of distribution was through the boxed publishers, who interested themselves only in a fairly small subset of the things that games can do and be. Long before the likes of Steam, the shareware scene was the indie-games scene of its time, demonstrating all of the quirky spirit which that phrase has come to imply. While the big boys were all gazing fixedly at the same few points in the middle distance, shareware makers dared to look in other directions — even, as in the case of Daniel and Matthew, to look behind them. In the face of a mainstream industry which seemed hell-bent on forgetting its history, that was perhaps the most radically indie notion of them all.
(Sources: the books Masters of Doom by David Kushner, Rocket Jump: Quake and the Golden Age of First-Person Shooters by David L. Craddock, and Sophistication & Simplicity: The Life and Times of the Apple II Computer by Steven Weyhrich; Computer Gaming World of December 1992, January 1993, March 1993, May 1993, June 1993, July 1993, September 1993, January 1994, February 1994, and June 1994; Game Developer of January/February 1995; PC Powerplay of May 1996; Questbusters of November 1991; Los Angeles Times of February 6 1987; the tape magazine CLOAD of February 1978; the disk magazine Big Blue Disk of January 1988, May 1988, June 1988, March 1989, April 1989, September 1989, and August 1990. Online sources include the archives on the old 3D Realms site, the M & R Technologies interview with Jim Knopf, Samuel Stoddard’s Apogee FAQ, Al Vekovius’s old faculty page at Louisiana State University Shreveport, Stephen Vekovius’s appearance on All Y’all podcast, “Apogee: Where Wolfenstein Got Its Start” at Polygon, Benj Edwards’s interview with Scott Miller for Gamasutra, and Matt Barton’s interview with Scott Miller. Most of all, I owe a warm thank you to Matthew Engle for giving me free registered copies of the 11th Dimension games and talking to me at length about his experiences in shareware games.
In the interest of full disclosure as well as a full listing of sources, I have to note that a small part of this article is drawn from lived personal experience. I actually knew Scott Miller and George Broussard in the late 1980s and early 1990s, albeit only in a very attenuated, second-hand sort of way: Scott dated my sister for several years. Scott and George came by my room from time to time to see the latest Amiga games when I was still in high school. Had I known that my sister’s lovelife had provided me with a front-row seat to gaming history, and that I would later become a gaming historian among other things, I would doubtless have taken more interest in them. As it was, though, they were just a couple of older guys with uncool MS-DOS computers wanting to see what an Amiga could do.
A year and a half to two years after finishing high school, I interviewed for a job at Apogee, which was by then flying high. Again, had I known what my future held I would have paid more attention to my surroundings; I retain only the vaguest impression of a chaotic but otherwise unremarkable-looking office. Scott and George were perceptive enough to realize that I would never have fit in with them, and didn’t hire me. For this I bear them no ill will whatsoever, given that their choice not to do so was the best one for all of us; I would have been miserable there. I believe that the day of that interview in 1992 was the last time I ever saw Scott and George; Scott and my sister broke up permanently shortly thereafter if not before.
The company once known as Apogee, which is now known as 3D Realms, has released all of their old shareware games for free on their website. Daniel Berke and Matthew Engle continue to maintain their old games in updated versions that work with modern incarnations of Windows; you can download them and purchase registrations on the 11th Dimension Entertainment home page.)
The Shareware Scene, Part 1: The Pioneers
The digital society which we’ve created over the last few decades has upended many of our traditional notions about commerce. Everyday teenagers now stress over their ratings and advertising revenues on YouTube; gamers in “free” games pay staggering sums for the privilege of advancing through them a little faster (wasn’t the actual playing supposed to be the point of a game?); “clicks” and “likes” have become commodities that are traded in the same way that soybean futures are in the “real” world; consumers have become speculators in their own future entertainment on crowd-funding platforms like Kickstarter; a writer like me can ask for support from readers like you to allow me to make content that I then give away for free. (Thank you for that!) And, in the most direct parallel to our main topic for today, even some of the biggest corporations on the planet have learned to give away their products for free, then ask us to pay for them later.
Some of these new modes of commerce reflect the best in us, some perhaps the very worst. They all share in common, however, the quality of being markedly different from the old model wherein you paid someone an upfront amount of money and got some concrete good or service in exchange. As those of you with elderly parents or grandparents may well have learned, our modern digital economies have departed so far from that model in some areas that just explaining how they work to someone still wedded to the old ways can be a daunting task indeed. (I know that my 86-year-old father has literally no idea what I do all day or how I can possibly be earning money from it…) Maybe we too should ask the question that so many of our elders are already asking themselves every day: exactly how did we get from there to here so quickly?
It’s a bigger question than any one article can possibly answer. Still, it does turn out that we can trace at least one point of origin of our strange new ways of commerce to a trio of American pioneers who, all within a year of one another, embraced a new model for selling software — a model which has, one might say, taken over the world.
Andrew Fluegelman
The first of our pioneers is one Andrew Fluegelman. Born in 1943, Fluegelman within his first 35 years of life finished law school, passed the Bar exam, took up and then gave up corporate law, and settled into a whole new career as the owner, editor, and sole employee of the Headlands Press, a boutique book publisher in Marin County, California. He worked from time to time with the techno-utopian visionary Stewart Brand on The Whole Earth Catalog, and even the books he edited and published on his own had much the same counter-cultural DIY flavor: The New Games Book (a selection of friendly outdoor sporting activities for groups of adults), How to Make and Sell Your Own Record, Worksteads: Living and Working in the Same Place. Yet for all their hippie bona fides, Headlands books went out under the larger imprint of the international publishing titan Doubleday. The ability to speak the language of both the idealistic dreamer and the everyday businessperson proved a vital asset for Fluegelman throughout his life.
Like Brand and so many others of a similar bent, Fluegelman saw great potential in the personal computer as a force for social liberation. Therefore in 1981, before ever actually purchasing a computer of his own, he signed a contract with Doubleday to embark on a new book project, this time with himself in the role of coauthor rather than just editor. It was to be an exploration of the role of computers in the writing process, in terms of both current practicalities and future potential. He would of course need to buy himself a computer to complete the project. Just as he was about to pull the trigger on an Apple II, the IBM PC was announced. “I took one look at it and just had this gut feeling,” he said in a later interview. “This is what I want.”
While he waited for the machine he had ordered to arrive, Fluegelman, who had never touched a computer before in his life, started teaching himself BASIC from books. Even after the computer came in, learning to word-process on it remained on the back burner for a time while he continued to pursue his new passion for programming. His bible was that touchstone of a generation of amateur programmers, David Ahl’s million-selling book BASIC Computer Games. Fluegelman:
I got Ahl’s [book], and I said, “This is just what I want to do.” I typed [one of the games] in. It took me a day to get the bugs out and get the thing to run. And as soon as I saw the program running, I immediately started thinking, “Well, gee, I’d really like to add up the scores, and say this, and make a little noise…” I’d look through the book, and I’d say, “Oh, there’s something I could use. What happens if I stick it in there?”
I’m a real believer in the Berlitz method of programming. Which is: you learn how to say, “Please pass the salt,” [then] you look in the dictionary and look up the word for “pepper,” stick it in there, and, by God, someone gives you the pepper. And you know you’re making progress. Purely trial and error.
I liked it a lot. I abandoned all bodily functions for about a month.
Programmers are born as much as made. You either feel the intrinsic joy of making a machine carry out your carefully stipulated will or you don’t; the rest is just details. Clearly Fluegelman felt the joy.
Still, the book project wouldn’t wait forever. Fluegelman and Jeremy Joan Hewes, his coauthor, had the idea that they would indeed write the book together, but with each working on his or her own machine from his or her own office. They would share their files electronically; it would be one more way of practicing what they intended to preach in the book proper, about the new methods of working that were unlocked by the computer. But Hewes had an older CP/M computer rather than a flashy new IBM PC, and this stopped them in their tracks — for the only telecommunications package currently available for the latter came from IBM themselves, and could only swap files using IBM’s proprietary protocols. Fluegelman thus found himself in the ironic position of being able to trade files with an IBM mainframe, but not with most of his peers in the world of personal computing. He could see only one solution:
[I] started out to write a communications program. I said, “Gee, I’d really like to do this, and I’d like to do that, and we should have a dialing directory, and we should have some macros…” And I just kept adding to it for my own use.
We eventually typeset the book using the program I wrote. In the process, I gave it to a lot of my friends, and they started using it. At the time it was the only program that let you do these things on the IBM PC; this was the early spring of 1982. And inevitably one of my friends said, “You know, you really ought to publish that.”
If I hadn’t been in the publishing business for eight years, I would have gone the traditional route — find a publisher, royalties — but I’d been through all that, and I’d seen the pitfalls and all the ways things can get derailed. And this was kind of a new medium, and I was still very exhilarated by it. And I said, having had all this fun, I just can’t go the same publishing route that I’ve gone before.
Throughout his life, Fluegelman had a special relationship with San Francisco’s Golden Gate Bridge. “I think it’s a power point,” he said once only semi-facetiously. “I have more inspirations driving across the Golden Gate Bridge…” One day shortly after finishing his program, he was driving across while thinking back to the pledge drive he had seen the night before on the local PBS television station.
My American readers will doubtless recognize the acronym, but, for the benefit of those of you in other places: PBS stands for “Public Broadcasting System.” It’s a network of over-the-air television stations which show children’s programs (most famously Sesame Street) as well as documentaries, news, and high-culture content such as symphony concerts and dramatizations of classic literature. Although the stations are free to watch, they are unlike other free stations in that they don’t sustain themselves with advertising. Instead they rely on a limited degree of taxpayer funding, but most of all on donations, in any amount and frequency, from viewers who appreciate their content and consider it worth supporting. In some ways, then, PBS can be called the great forefather of the many non-coercive digital-funding models of today. And indeed, the tale of Andrew Fluegelman makes the otherwise tangential thread that runs from PBS to so many modern Internet economies much more direct.
For, driving across his favorite bridge that day, Fluegelman had a PBS-inspired epiphany. He would market his little telecommunications package under the name of PC-Talk, using a method no one had ever dreamed of before.
I said, I’ll just set it out there, encourage people to use it. If they like it, I’ll ask them to send me some money. [He set the initial “suggested” donation at $25.]
So, I sent out the first version of the program that way. I put some notices on The Source and CompuServe: I’ve got this program, I wrote it, it’ll do this and this. It’s available for free, but if you like it, send me the money. And even if you don’t like it, still make copies for your friends because maybe they’ll like it and send some money.
The response was really overwhelming. I was getting money! I remember on the first day I got a check in the mail, and I just couldn’t believe it. I almost got driven out of business filling orders. At the time I was still producing books, and software programming was my own late-night thing. And suddenly I was standing there all day filling orders and licking stamps and sending things out, and I had to hire someone to start doing that. I was totally unprepared for it.
While I had written the program to work very well in my own situation, once you start sending software out into the world you start hearing about people with all sorts of crazy circumstances that you haven’t anticipated at all. I think if I had tried to publish this first version of the program [conventionally], people would have reacted very negatively. But they didn’t because I’d sent it out in this unrestricted way. So people would write back and say, “This is great, but why don’t you add this? Why don’t you try this?” In many cases people even helped me re-program to deal with their situations. And I ended up calling that “freeback” instead of “feedback” because it was really getting free support back from the community.
The usually savvy Fluegelman did make a couple of puzzling decisions during these early days. The first was to name his revolutionary scheme for software distribution “Freeware.” If you twist your synapses around just right, you can almost arrive at the sense he was trying to convey, but under any more straightforward reading the name becomes dangerously counter-intuitive. Thousands upon thousands of developers who came after Fluegelman would work desperately, but only partially successfully, to make people understand that their software wasn’t in fact “free” in the sense that using it regularly placed no ethical demand upon the user to financially compensate the creator.
Then, having coming up with such a flawed name, the lawyer in Fluegelman came to the fore: he went out and trademarked it. He imagined creating a proprietary “Freeware catalog,” collecting a lot of software that was marketed on the same model. Accordingly, he also included in his program’s liner notes a request for other programmers with useful software of their own to contact him, thereby to join him in a “unique marketing experiment.”
In the meanwhile, PC-Talk’s success was such that it quickly caught the attention of the business-computing mainstream. Already in August of 1982, the widely read InfoWorld magazine published an article on the subject, under the heading “CA man likens ‘Freeware’ to user-supported TV.” Fluegelman noted sensibly therein that, rather than fighting against the natural desire people had to make copies of their software and share them with their friends, Freeware leveraged it. He estimated that five copies of PC-Talk were made for every one that was downloaded directly from one of the commercial online services or sent out on disk by himself in response to a mailed request — and, unlike a conventional software publisher, he thought this ratio was just great.
Jim Knopf/Button
Our second pioneer was a far more experienced programmer than Fluegelman. Seattle-area resident Jim Knopf was only one year older than our first pioneer, but had already worked for IBM for many years as a systems analyst by the dawn of the microcomputer era. He built his first personal computer himself in 1978, then sold it to partially finance an Apple II. Among other things, he used that machine to keep track of the names and addresses of his church’s congregation. Knopf later wrote that “I liked what I produced so much [that] the program itself became a hobby — something I continued to work on and improve in my spare time.”
When the IBM PC was released in 1981, Knopf sold his Apple II and bought one of those instead. His first project on his new computer was to write a new version of his database program. As soon as said program was far enough along, Knopf started sharing it with his colleagues at IBM. They in turn shared it with their friends, and soon the database, which he called Easy File, went beyond his office, beyond Seattle, beyond Washington State. People encouraged him to upload it to the early online services; this he obligingly did, and it spread still faster.
Knopf was gratified by its popularity, but also bothered by it in a certain way. His database was still under active development; he was improving it virtually every week. But how to get these updates out to users? He included a note in the program asking users to “register” themselves so he could keep in touch with them; he maintained the resulting mailing list in Easy File itself. Yet keeping everyone up to date was prohibitively complicated and expensive in a world where most software was still passed around on floppy disks — a world where the idea of a program as a changing, improving entity rather than a static tool that just was what it was barely existed in the minds of most people. “How could I identify which of the users were serious ones – those that desired and required enhancements?” Knopf later wrote about his mindset at the time. “How could I afford to send mailings to notify them of the availability of improvements?”
So, in September of 1982, Knopf made a few moves which would define his future. First, he changed his own name for purposes of business. Worried that his Germanic surname would be too difficult for potential customers to pronounce and remember, he quite literally translated it into English. “Knopf,” you see, is the German word for the English “button” — and so Jim Knopf became Jim Button. (I’ll refer to him by the latter name from now on. Coincidentally, “Jim Knopf” is also the name of a character from a popular series of children’s books in Germany.) Next, he registered a company that referenced his new nom de plume: Buttonware. And, last but by no means least, he added a new note to his program. “I would ask those who received it to voluntarily send a modest donation to help defray my costs,” remembered Button later. “The message encouraged users to continue to use and share the program with others, and to send a $10 donation only if they wanted to be included in my mailing list.”
The very first person to contact Button in response told him that his approach was just the same as the one used by another program called PC-Talk. Button found himself a copy of PC-Talk, read its pitch to other programmers interested in joining the ranks of Freeware, and sent his own Easy File to Andrew Fluegelman. Fluegelman phoned Button excitedly on the same day that he received the package in the mail. The two of them hit it off right away.
While they waited for Fluegelman to find enough other quality software to make up his Freeware Catalog, the two agreed to form a preliminary marketing partnership. Button would rename his Easy File to PC-File and raise its price to $25 to create a kinship between the two products, and each program would promote the other, along with the Freeware trademark, in its liner notes. Button:
My wife said I was “a foolish old man” if I thought even one person would voluntarily send me money for the program. I was more optimistic. I suspected that enough voluntary payments would come to help pay for expansions to my personal-computer hobby – perhaps several hundred dollars. Maybe even a thousand dollars (in my wildest dreams!).
As it happened, he would have to learn to dream bigger. Like PC-Talk, PC-File turned into a roaring success.
The founding staff of PC World magazine. Andrew Fluegelman stands in the very back, slightly right of center.
Both programs owed much of their early success to the extracurricular efforts of the indefatigable Andrew Fluegelman. Shortly after releasing PC-Talk to such gratifying interest, Fluegelman had given the final manuscript of his word-processing book to Doubleday, who would soon publish it under the title Writing in the Computer Age. Still as smitten as ever by the potential of personal computing, he now embarked on his third career: he became a full-time computer journalist. He initially wrote and edited articles for PC Magazine, the first periodical dedicated to the IBM PC, but got his big break when he was asked to join the staff of a new rival known as PC World. Within a few issues, Fluegelman became editor-in-chief there.
Not coincidentally, the magazine lavished glowing coverage upon PC-Talk and PC-File. The latest version of Button’s program, for example, got a six-page feature review — as much space as might be devoted to a major business-software release from the likes of Microsoft or VisiCorp — in PC World‘s September 1983 issue. “What was previously a very desirable program is now just about mandatory for much of the PC population,” the review concluded. “If you use PC-File and don’t send Jim Button a check, the guilt will kill you. And it should.”
Button and his family were vacationing in Hawaii when the review appeared. Button:
The response was overwhelming. Our house sitter had to cart the mail home daily in grocery sacks.
When we arrived home, the grocery sacks were strewn all over the basement floor. We had to step over and around them just to get into our basement office. My son, John, worked days, evenings, and weekends just catching up on the mail. Life would never be the same for any of us!
Button would later date the beginning of Buttonware as a real business to these events. Nine months later, he quit his job with IBM, by which time he was making ten times as much from his “moonlighting” gig as from his day job.
Ironically, though, Button had already parted ways to some extent with Fluegelman by the time that life-changing review appeared. Fluegelman was finding it difficult to focus on his idea of starting a Freeware catalog, given that he was already spending his days running one of the biggest magazines in the computer industry and his evenings improving and supporting PC-Talk. Button:
Andrew got questions about my program and I got questions and requests about his. Checks were sent to the wrong place. The work required to correct all this grew exponentially. We had to make the separation.
Button came up with his own moniker for the distribution model he and Fluegelman had pioneered: “user-supported software.” That name was perhaps less actively misleading than “Freeware,” but still didn’t really get to the heart of the matter. Other names that were tried, such as “quasi-public domain,” were even worse. Luckily, the perfect moniker — one that would strike exactly the right note, and do it in just two syllables at that — was about to arrive along with Bob Wallace, the third principal in our little drama.
In this iconic picture of the early Microsoft, Bob Wallace is in the middle of the back row.
Like Jim Button, Bob Wallace was based in Seattle, and was a veteran of the kit era of personal computing. In fact, his experience with microcomputers stretched back even further than that of his counterpart: he had been the founder in 1976 of the Northwest Computer Society, one of the first hobbyist user groups in the country. Shortly thereafter, he was recruited from the computer store where he worked by Paul Allen, whereupon he became Microsoft’s ninth employee. In time, he became the leading force behind Microsoft’s implementation of the Pascal programming language. But, as an unreformed hippie whose social idealism paralleled his taste for psychedelic drugs, he found both Microsoft’s growing bureaucracy and its founders’ notoriously sharp-elbowed approach to business increasingly uncongenial as time went on. In March of 1983, he was for the first time refused permission to barge into Bill Gates’s office unannounced to argue some technical point or other, as had always been his wont. It was the last straw; he quit in a huff.
Taking note of Fluegelman and Button’s success, he wrote a word processor using his own Pascal implementation, and released it as PC-Write under the same payment model. To encourage its distribution, he added an extra incentive. He sent to any user who mailed in the suggested donation of $75 a special registration code, which she was then expected to enter into her copy of the program. When she gave this copy to others, it was thus tagged with its source. If any users of those copies sent in the fee, Wallace would send $25 to the user whose tag it bore; he later claimed that at least one person made $500 in these commissions. In its roundabout way, the scheme pioneered the idea of not just asking users for a donation out of the goodness of their hearts, but marking and altering the functionality of the software for those who sent in the payment, all through the use of the soon-to-be ubiquitous mechanism of the registration code.
But Wallace’s biggest contribution of all came in the form of a name. And therein lies a tale in itself.
Back in July of 1982, an InfoWorld magazine editor named Jay Lucas had started a column on “freeware” without being aware of Fluegelman’s counter-intuitive use of that term; Lucas took the word to mean any and all freely distributed software, whether the author asked for an eventual payment in return or not. The following spring, Fluegelman contacted the magazine to inform them of his trademark and ask them to cease and desist from violating it. So, Lucas launched a contest among his readers to come up with a new name. He reported in the InfoWorld dated May 30, 1983, that “at least a dozen” readers had sent in the same suggestion: “shareware.” He announced that he would be using this name henceforth. At the time, he still made no distinction between “free” software that came with financial strings attached and software that didn’t. He was, in other words, effectively using “shareware” as a synonym for all types of freely distributed software.
But when Bob Wallace saw the name, he knew that it was perfect for his distribution model: pithy, catchy, with all the right intimations. He contacted Lucas, who told him that he was free to use it; InfoWorld made no legal claim on the name. So, when PC-Write went out later that year, it described itself as “shareware.”
In early 1984, Softalk IBM, a brief-lived spinoff of a much-loved Apple II magazine, hired one Nelson Ford to write a regular column about “public-domain software.” Unsure what he should call the distribution model being used by each of Fluegelman, Button, and Wallace under a different name, he started off by employing the manifestly inadequate placeholder “quasi-public domain.” But in his May 1984 column, he announced a contest of his own: “A free disk of software and widespread publicity for the person sending in the best name for quasi-PD, contribution-suggested software. Since Andy won’t let anyone use ‘freeware,’ we’ll have to come up with another catchy name.”
He received such dubious suggestions as “conscience-wear” — “the longer you use the software, the more it wears on your conscience if you do not pay” — and “tryware.” But, just as Lucas had over at InfoWorld, Ford kept getting most of all the suggestion of “shareware.” Unaware of the name’s origin at InfoWorld, but well aware of its use by Wallace, he suspected that “shareware” would be as impossible for him to appropriate as “freeware.” Nevertheless, he inquired with Wallace — and was pleasantly surprised to be told that he was more than welcome to it. Ford announced the new name in the August 1984 issue of Softalk IBM.
It’s questionable whether the actual column in which he made the announcement was all that influential in the end, given that the issue in which it appeared was also the last one that Softalk IBM ever published. Still, Ford himself was a prominent figure online and in user-group circles. His use of the name going forward in those other contexts, combined with that of Jay Lucas in InfoWorld, probably had a real impact. Yet one has to suspect that it was PC-Write itself which truly spread the name hither and yon.
For, perhaps because a word processor, unlike a telecommunications program or a database, was a piece of software which absolutely every computer owner seemed to need, Wallace was even more successful with his first piece of shareware than the two peers who had beaten him onto the scene had been with theirs. The company he founded, which he called QuickSoft, would peak with annual sales of more than $2 million and more than 30 employees, while PC-Write itself would garner more than 45,000 registered users. Staying true to his ideals, Wallace would always refuse to turn it into a boxed commercial product with a price tag in the hundreds of dollars, something many conventional software publishers were soon pressuring him to do. “I’m out to make a living, not a killing,” he said.
Jim Button was less inclined to vocalize his ideals, but one senses that much the same sentiment guided him. Regardless, he too did very well for himself. Already by 1984, he was getting approximately $1000 worth of checks in the mail every day. While PC-File itself never garnered quite the popularity of PC-Write — about 7000 users registered their copies in the end — Button soon branched out well beyond that first effort. Buttonware would peak with annual sales of $4.5 million and 35 employees.
Those who jumped on the shareware bandwagon afterward would find it very difficult to overtake these two pioneers in terms of either income or market impact. As late as 1988, Compute! magazine judged that the two most impressive shareware products on the market were still PC-File and PC-Write, two of the first three ever released. But PC-Talk would have a shorter lifespan — and, much more tragically, so would its creator.
The founding staff of Macworld magazine. Andrew Fluegelman can just be seen at the very back, slightly left of center.
The PC World issue with the landmark review of PC-File was still on newsstands when Andrew Fluegelman had his next life-changing encounter with a computer: he was one of a select few invited to Apple for an early unveiling of the new Macintosh. He was so smitten by this whole new way of operating a computer that he immediately began lobbying for a companion magazine to PC World, to be named, naturally enough, Macworld. Its first issue appeared in time to greet the first Macintosh buyers early in 1984. Fluegelman held down the editor-in-chief job there even as he continued to fill the same role at PC World.
He was utterly unfazed to thus be straddling two encampments between which Apple was trying to foment a holy war. He spoke about the differences between the two aesthetics of computing in an interview that, like so much of what he said back then, rings disarmingly prescient today:
People [say the Macintosh is] more of a right-brain machine and all that. I think there is some truth to that. I think there is something to dealing with a graphical interface and a more kinetic interface; you’re really moving information around, you’re seeing it move as though it had substance. And you don’t see that on [an IBM] PC. The PC is very much a conceptual machine; you move information around the way you move formulas, elements on either side of an equation. I think there’s a difference.
I think the most important thing is to realize that computers are tools, that unless you want to become an expert programmer, the main thing that a computer provides you is the ability to express yourself. And if it’s letting you do that, if you now have hands on those tools, then you can be a force for good out in the world, doing the things that you used to do, that you’re still doing — representing your own ideas, not changing your persona to suddenly become a “computer person.”
And I think that may be the advantage of the Macintosh.
At bottom, Fluegelman himself wasn’t really a “computer person” in the sense of Button and Wallace, both of whom had been programming since the 1960s. And then, running not one but two of the biggest computer magazines in the country could hardly leave him with much free time. Thus PC-Talk was somewhat neglected, and other telecommunications software — some of it released under the burgeoning shareware model — took its place. Fluegelman accepted this with equanimity; he was never inclined to stay in one place for very long anyway. In an interview conducted at the very first Macworld Expo in January of 1985, he spoke of his excitement about the future — both his personal future and the world’s technological future:
I think this is just the next adventure for a lot of us to get into. I know the intellectual excitement the [computer] has caused for me. It’s really been a rejuvenation, and anything that gets you that pumped up has got to be something that you can use in a good way.
I also think that people who do get excited about computers and involved in all this are almost uniformly intelligent, interesting people. I never have been as socially involved, as interconnected with as many different kinds of people, as when I started getting involved with computers. I think that the easier it is for people to express themselves, and to share their views with others, that’s got to be a good democratic force.
It’s great to go along for 40 years and still find your life changing and new things happening. It makes you look forward to what’s going to happen when you’re 60, what’s going to happen when you’re 80.
Quotes like these are hard to square with what happened to Andrew Fluegelman just six months later.
On July 6, 1985, Fluegelman left his office as usual at the end of a working day, but never arrived at his home; he simply disappeared. A week later, police discovered his Mazda hatchback parked near the toll plaza at the entrance to the Golden Gate Bridge. They found a note addressed to his wife and family inside, but its contents have never been published. Nevertheless, we can piece some things together. It seems that his health hadn’t been good; he’d been suffering from colitis, for which he’d begun taking strong medication that was known to significantly impact many patients’ psychology — and, indeed, friends and colleagues in the aftermath mentioned that he’d been acting erratically in the final few days before his disappearance. There are reports as well that he may have recently received a cancer diagnosis. At any rate, the implications seem clear: the 41-year-old Andrew Fluegelman went back to one of his favorite places in the world — the bridge where he had invented the revolutionary concept of shareware if not the name — and jumped 220 feet into the water below. His body was never recovered.
The legacy of those brief four years between his discovery of the joys of BASIC and his death by suicide encompasses not only the shareware model but also PC World and especially Macworld. It went on to become arguably the most literate, thoughtful computer magazine ever, one of the vanishingly few to evince a genuine commitment to good writing in the abstract. In doing so, it merely held to the founding vision of its first editor-in-chief. One can’t help but wonder what else this force of nature might have done, had he lived.
At shareware’s peak in the early and mid-1990s, at least one glossy newsstand magazine was devoted exclusively to the subject in quite a number of countries.
By that fateful day in 1985, shareware was already becoming an unstoppable force, with more and more programmers throwing their hats into the ring. To be sure, most of them didn’t build seven-figure businesses out of it, as Jim Button and Bob Wallace did. Inevitably for a distribution model that placed all of its quality control on the back end, much of the shareware that was released wasn’t very good at all. Yet even many of those who didn’t get to give up their day jobs did receive the satisfaction and capitalistic validation of being paid real money, at least every once in a while, for something they had created. In time, this loose-knit band of fellow travelers began to take on the trappings of a movement.
To wit: in February of 1987, a “Meeting of Shareware Authors” assembled in Houston to chat and kibitz about their efforts. Out of that meeting grew the Association of Shareware Professionals six months later, with founding chairmen Jim Button and Bob Wallace. In the years that followed, the ASP published countless shareware catalogs and pamphlets; they even published a 780-page book in 1993 called The Shareware Compendium, which represented the last attempt anyone ever made to list in one place all of the staggering quantity of shareware that was available by that point. But perhaps even more importantly, the ASP acted as a social outlet for the shareware authors themselves, a way of sharing hints and tips, highs and lows, dos and don’ts with one another.
There arose more big success stories out of all this ferment. For example, one Phil Katz was responsible for what remains today the most tangible single software artifact of the early shareware scene. In 1986, he started a little company called PKWare to distribute a reverse-engineered shareware clone of ARC, the most popular general-purpose compression program of the time. When the owners of ARC came after him with legal threats, he switched gears and in 1989 released PKZIP, which used an alternative, much more efficient compression format of his own design. Although he sold PKZIP as shareware — $25 donation requested, $47 for a printed manual — he also scrupulously documented the compression format it used and left the door open for other implementations of it. He was rewarded with sweet revenge: ZIP quickly superseded ARC all across the digital world. Striking a fine balance between efficiency and ease of implementation, not to mention being unentangled by patents, it has remained the world’s most common compression format to this day, a de facto standard that is now built right into many operating systems.
Another success story is less earthshaking and more esoteric, but instructive nonetheless as an illustration of just how far the shareware model could be stretched. In a time when desktop publishing was one of the biggest buzzwords in computing, a veteran of print publishing named Gary Elfring took a hard look at the current state of digital fonts, and noted how expensive those offered by major foundries like Adobe tended to be. He started Elfring Soft Fonts to distribute shareware typefaces, and made a lot of money from them in the late 1980s and early 1990s, before the established vendors of word processors and operating systems got their acts together in that department.
I could go on and on with such stories, but suffice to say that many people did very, very well from shareware during its heyday.
Like any movement, shareware also came complete with internecine disputes. One constant source of tension were the many third parties who collected shareware which they didn’t own on physical media for distribution. As early as 1984, the librarian of the Silicon Valley Computer Society users group caused an uproar when he started selling floppy disks filled with shareware for $6 apiece, a figure somewhat above the cost of blank disks and postage alone. “It’s not legal,” said Andrew Fluegelman flatly at the time. “I’m opposed to it because when somebody spends even $6 for a disk, they feel they’ve paid for it and see little reason to pay again for it. I’m concerned about somebody building a product around my product.” But, in a rare break with Fluegelman, Jim Button had a different point of view: “With that [price], all he’s doing is helping me distribute sample copies.” He continued in later years to believe that “distribution is one of the cornerstones of sales. All other factors being equal, if you can double your distribution you will double your sales.”
In the end, Button’s point of view carried the day. Shareware authors were never entirely comfortable with the “parasites” who profited off their software in this way, and Fluegelman’s worry that many users would fail to distinguish between paying a cataloger and paying the actual creator of the software was undoubtedly well-founded. Yet the reality was that the vast majority of computer owners would not go online until the World Wide Web struck in the mid-1990s. In the meantime, floppy disks — and eventually CD-ROMs — were the only realistic mechanism for reaching all of these otherwise isolated users. The catalogers and the authors had to learn to live with one another in an uneasy symbiotic relationship.
Another, even more bitter dispute within the ranks of shareware was touched off near the end of the 1980s, when some authors started opting to “encourage” registration by releasing crippled versions of their software — programs that only functioned for a limited time, or that blocked access to important features — that could only have their full potential unlocked via the input of a valid registration code. Although Bob Wallace had ironically pioneered the idea of a registration code that was input directly into a program, he and most of the other early shareware pioneers hated to see the codes used in this way. For the socially conscious Wallace, it was a moral issue; his vision for shareware had always been to collect payment from those who could pay, but not to deprive those who couldn’t of quality software. Button as well preferred to rely upon the honor system: “Don’t get off on the wrong foot with your users with things like crippled programs, time-limited programs, and other negative incentives to register your software. If you can’t trust your users to pay for truly good software, then you should stay out of the shareware business.” Under the influence of these two founding chairmen, the ASP refused for a time to admit shareware authors who freely distributed only crippled versions of their software.
In the end, though, the ASP would be forced to relax their stance, and “crippleware” would become nearly synonymous with shareware in many circles, for better or for worse. In 1989, Nelson Ford, the earlier popularizer of the name “shareware,” set up a service for authors which let people register their software over the telephone using their credit cards instead of having to mail checks or cash through the post. The ease of passing out registration codes this way, without having to send out disks and/or documentation or do any additional work at all, probably led many more authors to go the crippleware route. In fairness to those who decided to implement such schemes, it should be noted that they didn’t have the advantages that went along with being first on the scene, and were often marketing to less committed computer users with a less nuanced sense of the ethics of intellectual property and the sheer amount of work that goes into making good software of any stripe.
In a strange sort of way, Windows 10 is actually a shareware product.
The buzz around shareware gradually faded in the second half of the 1990s, and by soon after the turn of the millennium the term was starting to seem like an antiquated relic of computing’s past. Even the Association of Shareware Professionals eventually changed their name to the Association of Software Professionals, before doddering off entirely. (A website still exists for the organization today, but it doesn’t appear to have been updated in some years.)
Yet it would be profoundly inaccurate to say that shareware died as anything but a name. On the contrary: it conquered the world to such an extent that it became the accepted means of distributing much or most software, and as such is no longer in need of any particular name. Just about everyone is selling shareware today — not only the sometimes honest, sometimes dodgy small vendors of “try before you buy” utilities of many types, but also some of the biggest corporations in the world. Microsoft, for example, now distributes Windows using what is essentially the shareware model: users download a copy for free, enjoy a limited trial period, and then need to purchase a registration code if they wish to go on using it. Many other software developers have stuck to their idealistic guns and put their creations out there uncrippled, asking for a donation only from those who can afford it. And, as I mentioned to open this piece, the overarching spirit of shareware, if you will, has infected countless digital economies that don’t involve downloads or registration keys at all.
Jim Button and Bob Wallace got to see some of these later developments, but they weren’t active participants in most of them. Wallace gradually divested himself from Quicksoft after 1990. Ever the hippie, he devoted his time to the study and promotion of psychedelic drugs and other “mind-expanding technologies” via publications and foundations. He died in 2002 at age 53 from a sudden attack of pneumonia that may or may not have been related to his quest for chemical transcendence.
Jim Button (né Knopf) very nearly died even younger. At the age of 49 in 1992, he had a major heart attack. He survived, but wasn’t sure that he could continue to cope with the stress of running his shareware business. At the time, big players like Microsoft were pouring enormous resources into their own productivity software, and the likes of little Buttonware had no real hope of competing with them anymore. This combination of factors prompted Button to slowly wind his company down; after all, his decade in shareware had already left him with enough money to enjoy a comfortable early retirement. He died in 2013, a few weeks shy of his 71st birthday. He continued until the end to downplay his role in the evolution of software distribution and digital culture. “I’m not a visionary man,” he said. “I never saw the future, but I was lucky enough to be in the right place at the right time, with the right ideas and a proper amount of energy.”
Some might say that the “right ideas” are synonymous with vision, but no matter; we’ll let him keep his modesty. What he and his fellow pioneers wrought speaks for itself. All you have to do is look around this place we call the Internet.
(Sources: the books The New Games Book by the New Games Foundation, Writing in the Computer Age by Andrew Fluegelman and Jeremy Joan Hewes, and Gates by Stephen Manes and Paul Andrews; Softalk IBM of May 1984, June 1984, July 1984, and August 1984; Byte of June 1976, June 1983, July 1984, March 1985, and September 1987; 80 Computing of May 1987; Ahoy! of February 1984; CompuServe Magazine of December 1990 and March 1992; Family Computing of March 1984; InfoWorld of July 5 1982, August 23 1982, December 20 1982, March 7 1983, May 30 1983, June 27 1983, July 30 1984, September 17 1984, October 22 1984, July 29 1985, December 23 1985, August 25 1986, and December 7 1987; MicroTimes of May 1985 and August 1985; Games Machine of October 1987; Compute! of February 1985 and June 1988; PC World of September 1983; Macworld premiere issue. Online sources include The Association of Software Professional’s website, Michael E. Callahan’s “History of Shareware” on Paul’s Picks, The Charley Project‘s entry on Andrew Fluegelman’s disappearance, the Shareware Junkies interview with Jim “Button” Knopf, “Jim Button: Where is He Now?” at Dr. Dobb’s, the M & R Technologies interview with Jim Knopf, the Brown Alumni Monthly obituary of Bob Wallace, and a 1989 online discussion of the newly released PKZip archived by Jason Scott. My thanks to Matthew Engle for giving me the picture of Shareware Magazine included in this article.)
The (7th) Guest’s New Clothes
Once upon a time, two wizards decided to remake the face of computer gaming with the help of a new form of magic known as CD-ROM. They labored for years on their task, while the people waited anxiously, pouncing upon the merest hint the wizards let drop of what the final product would look like.
At long last — well after the two wizards themselves had hoped — the day of revelation came. Everyone, including both the everyday people and the enlightened scribes who kept them informed on the latest games, rushed to play this one, which they had been promised would be the best one ever. And at first, all went as the wizards had confidently expected. The scribes wrote rapturously about the game, and hordes of people bought it, making the wizards very rich.
But then one day a middle-aged woman, taking a break from reckoning household accounts by playing the wizards’ game, said to her husband, “You know, honey, this game is really kind of slow and boring.” And in time, a murmur of discontent spread through many ranks of the people, gaining strength all the while. The cry was amplified by a disheveled young man with a demon of some sort on his tee-shirt and a fevered look in his eyes: “That’s what I’ve been saying all along! The wizards’ game sucks! Play this one instead!” And he hunched back down over his computer to continue playing his very different sort of game, muttering something about “gibs” and “frags” as he did so.
The two wizards were disturbed by this growing discontent, but resolved to win the people over with a new game that would be just like their old one, except even more beautiful. They worked on it too for years to make it as amazing as possible. Yet when they offered it to the people, exponentially fewer of them bought it than had bought their first game, and their critics grew still louder and more strident. They tried yet one more new game of the same type, yet more beautiful, but by now the people had lost interest entirely; few could even be bothered to criticize it. The wizards started bickering with each other, each blaming the other for their failures.
One of the wizards, convinced he could do better by himself, went away to make still more games of the same type, but the people remained stubbornly uninterested; he finally gave up and found another occupation. From time to time, he tries again to see if the people want another game like the one they seemed to love so much on that one occasion long ago, but he is invariably disappointed.
The other wizard — perhaps the wiser of the two — said, “If you can’t beat ’em, join ’em.” He joined the guild that included the violent adolescent with the demon on his shirt, and enjoyed a return to fortune if not fame.
Such is the story of Trilobyte Games in a nutshell. Today, we remember 1993 as the year that Cyan Productions and id Software came to the fore with Myst and Doom, those two radically different would-be blueprints for gaming’s future. But we tend to forget that the most hyped company and game of the year were in fact neither of those pairings: they were rather Trilobyte and their game The 7th Guest. Echoing the conventional wisdom of the time, Bill Gates called The 7th Guest “the future of multimedia,” and some even compared Graeme Devine and Rob Landeros, the two “wizards” who had founded Trilobyte together, to John Lennon and Paul McCartney. Sadly for the wizards, however, The 7th Guest had none of the timeless qualities of the Beatles’ music; it was as of its own time as hula hoops, love beads, or polyester leisure suits were of theirs.
Rob Landeros and Graeme Devine
Unlike their alter egos in the Beatles, Graeme Devine and Rob Landeros grew up in vastly different environments, separated not only by an ocean but by the equally enormous gulf of seventeen years.
Born in Glasgow, Scotland, in 1966, Devine was one of the army of teenage bedroom coders who built the British games industry from behind the keyboards of their Sinclair Spectrums. His first published work was actually a programming utility rather than a game, released as part of a more complete Speccy programmer’s toolkit by a company known as Softek in the spring of 1983. But it was followed by his shoot-em-up Firebirds just a few months later. That game’s smooth scrolling and slick presentation won him a reputation. Thus one day the following year the phone rang at his family’s home; a representative from Atari was on the line, asking if he would be free to port their standup-arcade and console hit Pole Position to the Spectrum.
Over the next several years, Devine continued to port games from American publishers to the Europe-centric Spectrum, while also making more original games of his own: Xcel (1985), Attack of the Killer Tomatoes (1986), Metropolis (1987). His originals tended to be a bit half-baked once you really dove in, but their technical innovations were usually enough to sustain them, considering that most of them only cost a few quid. Metropolis, the first game Devine programmed for MS-DOS machines, provides a prime example of both his technical flair and complete lack of detail orientation. A sort of interactive murder mystery taking place in a city of robots, sharing only a certain visual sensibility with the Fritz Lang film classic of the same name, it includes almost-decipherable “voice acting” for its characters, implemented without the luxury of a sound card, being played entirely through the early IBM PC beeper. The game itself, on the other hand, is literally unfinished; it breaks halfway through its advertised ten cases. Perhaps Devine decided that, given that he included no system for saving his rather time-consuming game, no one would ever get that far anyway.
Metropolis
Metropolis was published through the British budget label Mastertronic, whose founder Martin Alper was a force of nature, famous as a cultivator of erratic young talent like Devine. Alper sold Mastertronic to Richard Branson’s Virgin Media empire just after Metropolis was released, and soon after that absconded to Southern California to oversee the newly formed American branch of Virgin Games. On a routine visit back to the Virgin mother ship in London in 1988, he dropped in on Devine, only to find him mired in a dark depression; it seemed his first serious girlfriend had just left him. “England obviously isn’t treating you well,” said Alper. “Why don’t you come with me to California?” Just like that, the 22-year-old Devine became the head of Virgin Games’s American research and development. It was in that role that he met Rob Landeros the following year.
Landeros’s origin story was about as different from Devine’s as could be imagined. Born in 1949 in Redlands, California, he had lived the life of an itinerant bohemian artist. After drifting through art school, he spent much of the 1970s in hippie communes, earning his keep by drawing underground comic books and engraving tourist trinkets. By the early 1980s, he had gotten married and settled down somewhat, and found himself fascinated by the burgeoning potential of the personal computer. He bought himself a Commodore 64, learned how to program it in BASIC, and even contributed a simple card game to the magazine Compute!’s Gazette in the form of a type-in listing.
But he remained a computer hobbyist only until the day in early 1986 that an artist friend of his by the name of Jim Sachs showed him his new Commodore Amiga. Immediately struck by the artistic possibilities inherent in the world’s first true multimedia personal computer, Landeros worked under Sachs to help illustrate Defender of the Crown, the first Amiga game from a new company called Cinemaware. After that project, Sachs elected not to stay on with Cinemaware, but instead recommended Landeros for the role of the company’s art director. Landeros filled that post for the next few years, illustrating more high-concept “interactive movies” which could hardly have been more different on the surface from Devine’s quick-and-dirty budget games — but which nevertheless tended to evince some of the same problems when it came to the question of their actual gameplay.
Whatever its flaws in that department, Martin Alper over at Virgin was convinced that the Cinemaware catalog was an early proof of concept for gaming’s future. As Cinemaware founder Bob Jacob and many others inside and outside his company well recognized, their efforts were hobbled by the need to rely on cramped, slow floppy disks to store all of their audiovisual assets and stream them into memory during play. But with CD-ROM on the horizon for MS-DOS computers, along with new graphics and sound cards that would make the platform even more audiovisually capable than the Amiga, that could soon be a restriction of the past. Alper asked Devine to interview Landeros for the role of Virgin’s art director.
Landeros was feeling “underappreciated and underpaid” at Cinemaware, as he puts it, so he was very receptive to such an offer. When he called Devine back after hearing the message the latter had left on his answering machine, he found the younger man in an ebullient mood. He had just gotten engaged to be married, Devine explained, to a real California girl — surely every cloistered British programmer’s wildest fantasy. Charmed by the lad’s energy and enthusiasm, Landeros let himself be talked into a job. And indeed, Devine and Landeros quickly found that they got on like a house on fire.
Tall and skinny and bespectacled, with unkempt long hair flying everywhere, Devine alternated the euphoria with which he had first greeted Landeros with bouts of depression such as the one Martin Alper had once found him mired in. Landeros was calmer, more grounded, as befit his age, but still had a subversive edge of his own. When you first met him, he had almost a patrician air — but when he turned around for the first time, you noticed a small ponytail snaking down his back. While Devine was, like so many hackers, used to coding for days or weeks on end, sometimes to the detriment of his health and psychological well-being, Landeros needed a very good reason indeed to give up his weekend motorcycle tours. Devine was hugely impressed by Landeros’s tales of his free-spirited life, as he was by the piles of self-inked comic books lying about his home; Landeros was repeatedly amazed simply at the things Devine could make computers do. The two men complemented each other — perhaps were even personally good for one another in some way that transcends job and career.
Their work at Virgin, however, wasn’t always the most exciting. The CD-ROM revolution proved late in arriving; in the meantime, the business of making games continued pretty much as usual. In between his other duties, Devine made Spot, an abstract strategy game which betrayed a large debt to the ancient Japanese board game of Go whilst also serving as an advertisement for the soft drink 7 Up; if not quite a classic, it did show more focus than his earlier efforts. Meanwhile Landeros did the art for a very Cinemaware-like cross-genre concoction called Spirit of Excalibur. In his spare time, he also helped his friend and fellow Cinemaware alumnus Peter Oliphant with a unique word-puzzle/game-show hybrid called Lexi-Cross. (Rejected by Alper because “game shows need a license in order to sell,” it was finally accepted by Interplay after that company’s head Brian Fargo brought a copy home to his wife and she couldn’t stop playing it. Nonetheless, it sold hardly at all, just as Alper had predicted.)
Devine and Landeros were itching to work with CD-ROM, but everywhere they went they were told that the market just wasn’t there yet. As they saw it, no one was buying CD-ROM drives because no one was making compelling enough software products for the new medium. It was a self-fulfilling prophecy, a marketplace Gordian knot which someone had to break. Accordingly, they decided to put together their own proposal for a showpiece CD-ROM game. Both were entranced by Twin Peaks, the darkly quirky murder-mystery television series by David Lynch, which had premiered in the spring of 1990 and promptly become an unlikely mass-media sensation. Sitting in the airport together one day, they overheard the people around them debating the question of the year: who killed Laura Palmer?
Imagine a game that can fascinate in the same way, mused Devine. And so they started to brainstorm. They pictured a game, perhaps a bit like the board game Clue — tellingly, the details of the gameplay were vague in their minds right from the start — that might make use of a Twin Peaks license if such a thing was possible, but would go for that sort of vibe regardless. Most importantly, it would pull out all the stops to show what CD-ROM — and only CD-ROM — could do; there would be no floppy version. Indeed, the project would be thoroughly uncompromising in all of its hardware requirements, freeing it from the draconian restrictions that came with catering to the lowest common denominator. It would require one of a new generation of so-called “Super” VGA graphics cards, which would let it push past the grainy resolution of 320 X 200, still the almost universal standard in games, to a much sharper 640 X 480.
To keep the development complications from spiraling completely out of control, it could take place in a haunted house that had a group of people trapped inside, being killed one by one. Sure, Agatha Christie had done it before, but this would be different. Creepier. Darker. A ghost story as well as a mystery, all served up with a strong twist of David Lynch. “Who killed Laura Palmer? Who killed Laura Palmer? We wanted to create that sort of intrigue,” remembers Landeros.
When they broached the possibility of a Twin Peaks game with Alper, he was definitive on one point: there wasn’t enough room in his budget to acquire a license to one of the hottest media properties in the country. They should therefore focus their thinking on a Twin Peaks-like game, not the real thing. Otherwise, he was noncommittal. “Give me a detailed written proposal, and we’ll see,” he said.
At this point in our story, it would behoove us to know something more of Martin Alper the man, a towering figure whose shadow loomed large over all of Virgin Games. A painter and sculptor of some talent during his free time, Alper was also an insatiable culture vulture, reading very nearly a novel per day and seeing several films per week. His prodigious consumption left no space for games. “I’ve never played any game,” he liked to boast. “What interests me is the cultural progress that games can generate. I’m looking to make a difference in society.” He liked to think of himself as a 1990s incarnation of Orson Welles, nudging his own group of Mercury Players into whole new fields of creative expression. When Devine and Landeros’s detailed proposal landed on his desk in November of 1990, full of ambition to harness the current zeitgeist in the service of a new medium, it hit him right where he lived. Even the proposed budget of $300,000 — two to three times that of the typical Virgin game — put him off not at all.
So, he invited Devine and Landeros to a lunch which has since gone down in gaming lore. After the niceties had been dispensed with, he told the two bluntly that they had “no future at Virgin Games.” He enjoyed their shock for a while — a certain flair for drama was also among his character traits — then elaborated. “Your idea is too big to be developed here. If you stayed here, you’d quickly overrun our offices. I can’t afford to let you do that. Other games have to be made here as well.”
“What do you suggest?” ventured Devine.
And so Alper laid out his grand plan. They should start their own studio, which Virgin Games would finance. They could work where they liked and hire whomever they liked, as long as the cost didn’t become too outrageous and as long as they stayed within 90 minutes of Virgin’s headquarters, so that Alper and David Bishop, the producer he planned to assign to them, could keep tabs on their progress. And they would have to plan for the eventuality of a floppy-disk release as well, if, as seemed likely, CD-ROM hadn’t yet caught on to a sufficient degree with consumers by the following Christmas, the game’s proposed release date. They were simple requirements, not to mention generous beyond Devine and Landeros’s wildest dreams. Nevertheless, they would fail to meet them rather comprehensively.
In the course of his hippie wanderings, Landeros had fallen in love with the southern part of Oregon. After the meeting with Alper, he suggested to Devine that they consider setting up shop there, where the biking and motorcycling were tremendous, the scenery was beautiful, the people were mellow, and the cost of living was low. When Devine protested that one certainly couldn’t drive there from Virgin’s offices within 90 minutes, Landeros just winked back. Alper hadn’t actually specified a mode of transportation, he noted. And one could just about fly there in an hour and a half.
On December 5, 1990, the pair came for the first time to Jacksonville, Oregon, a town of just 2000 inhabitants. It so happened that the lighting of the town Christmas tree was taking place that day. All of the people had come out for the occasion, dressed in Santa suits and Victorian costumes, caroling and roasting chestnuts. Just at sunset, snow started to fall. Devine, the British city boy far from home, looked around with shining eyes at this latest evolution of his American dream. Oregon it must be.
So, during that same visit, they signed a lease on a small office above a tavern in an 1884-vintage building — wood floors, a chandelier on the ceiling, even a fireplace. They hired Diane Moses, a waitress from the tavern below, to serve as their office manager. Then they went back south to face the music.
The 7th Guest was created in this 1884-vintage building in Jacksonville, Oregon, above a tavern which is now known as Boomtown Saloon.
Alper was less than pleased at first that they had so blatantly ignored his instructions, but they played up the cheap cost of living and complete lack of distractions in the area until he grudgingly acquiesced. The men’s wives were an even tougher sell, especially when they all returned to Jacksonville together in January and found a very different scene: a bitter cold snap had caused pipes to burst all over town, flooding the streets with water that had now turned to treacherous ice, making a veritable deathtrap of the sidewalk leading up to their new office’s entrance. But the die was now cast, for better or for worse.
The studio which Devine and Landeros had chosen to name Trilobyte officially opened for business on February 1, 1991. The friends found that working above a tavern had its attractions after a long day — and sometimes even in the middle of one. “It’s fun to watch the fights spill out onto the street,” said Devine to a curious local newspaper reporter.
The first pressing order of business was to secure a script for a game that was still in reality little more than a vague aspiration. Landeros had already made contact over the GEnie online service with Matthew Costello, a horror novelist, gaming journalist, and sometime tabletop-game designer. He provided Trilobyte with a 100-page script for something he called simply Guest. Graeme Devine:
We presented the basic story to Matt, and he made it into a larger story, built the characters and the script. He created it out of what was really just a sketch. We were anxious that the [setting] be very, very closed. One that would work as a computer environment. That’s what he gave us.
The script took place within a single deserted mansion, and did all of its storytelling through ghostly visions which the player would bump into from time to time, and which could be easily conveyed through conveniently non-interactive video snippets. Like so many computer games, in other words, Guest would be more backstory than story.
Said backstory takes place in 1935, and hinges on a mysterious toy maker named Henry Stauf — the anagram of Faust is intentional — who makes and sells a series of dolls which cause all of the children who play with them to sicken and die. When the people of his town figure out the common thread that connects their dead children, they come for him with blood in their eyes. He barricades himself in his mansion to escape their wrath — but sometime shortly thereafter he lures six guests into spending a night in the mansion, with a promise of riches for those who survive. Falling victim either to Stauf’s evil influence or their own paranoia, or both, the six guests all manage to kill one another, Agatha Christie-style, over the course of the night, all without ever meeting Stauf himself in the flesh. But there is also a seventh, uninvited guest, a street kid named Tad who sneaks in and witnesses all of the horror, only to have his own soul trapped inside the mansion. It becomes clear only very slowly over the course of the game that the player is Tad’s spirit, obsessively recapitulating the events of that night of long ago, looking for an escape from his psychic prison in the long-deserted mansion.
The backstory of how Stauf came to take up residence in his mansion is shown in the form of narrated storybook right after the opening credits.
The only thing missing from Costello’s script was any clear indication of what the player would be expected to do in the course of it all. Trilobyte planned to gate progress with “challenges to the player’s intellect and curiosity. Our list of things to avoid includes: impossible riddles, text parsers, inventories, character attribute points, sword fights, trolls, etc. All actions are accomplished via mouse only. Game rules will either be self-explanatory or simple enough to discover with minimal experimentation.” It sounded good in the abstract, but it certainly wasn’t very specific. Trilobyte wouldn’t seriously turn to the game part of their game for a long, long time to come.
The question of Guest‘s technical implementation was almost as unsettled, but much more pressing. Devine and Landeros first imagined showing digitized photographs of a real environment. Accordingly, they negotiated access to Jacksonville’s Nunan House, a palatial three-story, sixteen-room example of the Queen Anne style, built by a local mining magnate in 1892. But, while the house was fine, the technology just wouldn’t come together. Devine had his heart set on an immersive environment where you could see yourself actually moving through the house. Despite all his technical wizardry, he couldn’t figure out how to create such an effect from a collection of still photographs.
The Mansion
The Nunan House in Jacksonville, Oregon, whose exterior served as the model for the Stauf Mansion. The interior of the latter was, however, completely different, with the exception only of a prominent central staircase.
A breakthrough arrived when Devine and Landeros shared their woes with a former colleague from Virgin, an artist named Robert Stein. Stein had been playing for several months with 3D Studio, a new software package from a company known as Autodesk which let one build and render 3D scenes and animations. It was still an awkward tool in many ways, lagging behind similar packages for the Commodore Amiga and Apple Macintosh. Nonetheless, a sufficiently talented artist could do remarkable things with it, and it had the advantage of running on the MS-DOS computers on which Trilobyte was developing Guest. Devine and Landeros were convinced when Stein whipped up a spooky living room for them, complete with a ghostly chair that flew around of its own accord. Stein soon came to join them in Jacksonville, becoming the fourth and last inhabitant of their cozy little office.
3D Studio
The 7th Guest was the first major game to make extensive use of Autodesk’s 3D Studio, a tool that would soon become ubiquitous in the industry. Here we see the first stage of the modeling process: the Shaper, in which an object is created as a two-dimensional geometric drawing, stored in the form of points and vectors.
In the Lofter, an object’s two dimensions are extruded into three, as the X- and Y-coordinates of its points are joined to Z-coordinates.
The Materials Editor is used to apply textured surfaces to what were previously wire-frame objects.
The 3D Editor is used to build a scene by hanging objects together in a virtual space and defining the position, color, and intensity of light sources.
The Keyframer is used to create animation. The artist arranges the world in a set of these so-called key frames, then tells the computer to extrapolate all of the frames in between. The process was an extremely time-consuming one on early-1990s computer hardware; each frame of a complex animation could easily take half an hour to render.
Even using 3D Studio, Guest must fall well short of the ideal of an immersive free-scrolling environment. At the time, only a few studios — most notably Looking Glass Technologies and, to a much more limited extent, id Software of eventual Doom fame — were even experimenting with such things. The reality was that making interactive free-scrolling 3D work at all on the computer hardware of the era required drastic compromises in terms of quality — compromises which Trilobyte wasn’t willing to make. Instead they settled for a different sort of compromise, in the form of a node-based approach to movement. The player is able to stand only at certain pre-defined locations, or nodes, in the mansion. When she clicks to move to another node, a pre-rendered animation plays, showing her moving through the mansion.
Just streaming these snippets off CD fast enough to play as they should taxed Devine’s considerable programming talents to the utmost. He would later muse that he learned two principal things from the whole project: “First, CD-ROM is bloody slow. Second, CD-ROM is bloody slow.” When he could stretch his compression routines no further, he found other tricks to employ. For example, he got Landeros to agree to present the environment in a “letter-boxed” widescreen format. Doing so would give it a sense of cinematic grandeur, even as the black bars at the top and bottom of the monitor dramatically reduced the number of pixels Devine’s routines had to move around. A win win.
With the interior of the mansion slowly coming into being, the time was nigh to think about the ghostly video clips which would convey the story. Trilobyte recruited local community-theater thespians to play all the parts; with only $35,000 to spend on filming, including the camera equipment, they needed actors willing to work for almost nothing. The two-day shoot took place in a rented loft in Medford, Oregon, on a “stage” covered with green butcher paper. The starring role of Stauf went to Robert Hirschboeck, a fixture of the annual Oregon Shakespeare Festival, which was (and is) held in nearby Ashland. Diane Moses, Trilobyte’s faithful office manager, also got a part.
Robert Hirschboeck, the semi-professional Shakespearean actor who played the role of Stauf in The 7th Guest and its sequel. He was bemused by the brief fame the role won him: “I’ll be walking down the street and meet someone with all the CD-ROM gear, and they’ll say, ‘Ah, man, I’ve been looking at your ugly mug for 60 hours this week.'”
Trilobyte believed, with some justification, that their game’s premise would allow them to avoid some of the visual dissonance that normally resulted from overlaying filmed actors onto computer-generated backgrounds: their particular actors represented ghosts, which meant it was acceptable for them to seem not quite of the world around them. To enhance the impression, Trilobyte added flickering effects and blurry phosphorescent trails which followed the actors’ movements.
The Chroma-Key Process
A technique known as chroma-keying was used by The 7th Guest and most other games of the full-motion-video era to blend filmed actors with computer-generated backgrounds. The actor is filmed in front of a uniform green background. After digitization, all pixels of this color are rendered transparent. (This means that green clothing is right out for the actors…)
Meanwhile a background — the “stage” for the scene — has been created on the computer.
Finally, the filmed footage is overlaid onto the background.
While Trilobyte built their 3D mansion and filmed their actors, the project slipped further and further behind schedule. Already by May of 1991, they had to break the news to Alper that there was no possibility of a Christmas 1991 release; Christmas 1992 might be a more realistic target. Luckily, Alper believed in what they were doing. And the delay wasn’t all bad at that; it would give consumers more time to acquire the SVGA cards and CD-ROM drives they would need to run Guest — for by now it was painfully clear that a floppy-disk version of the game just wasn’t going to happen.
In January of 1992, Devine, Landeros, and Stein flew to Chicago for the Winter Consumer Electronics Show. They intended to keep a low profile; their plan was simply to check out the competition and to show their latest progress to Alper and his colleagues. But when he saw what they had, Alper broke out in goosebumps. Cinema connoisseur that he was, he compared it to Snow White and the Seven Dwarfs, Walt Disney’s first feature film, which forever changed the way people thought about cartoon animation. What Snow White had done for film, Alper said, Guest could do for games. He decided on the spot that it needed to be seen, right there and then. So, he found a computer on the show floor that was currently demonstrating a rather yawn-inducing computerized version of Scrabble and repurposed it to show off Guest. To make up for the fact that Trilobyte’s work had no music as of yet, he put on a CD of suitably portentous Danny Elfman soundtrack extracts to accompany it.
Thanks to this ad hoc demonstration, Guest turned into one of the most talked-about games of the show. Its stunning visuals were catnip to an industry craving killer apps that could nudge reluctant consumers onto the CD-ROM bandwagon. Bill Gates hung around the demo machine like a dog close to feeding time. Virgin’s competitor Origin Systems, of Wing Commander and Ultima fame, also sat up and took notice. They highlighted Guest as the game to watch in their internal newsletter:
Here’s a tip: keep an eye out for Guest, a made-for-CD-ROM title from Oregon developer Trilobyte for Virgin Games. In it, you explore a 22-room haunted mansion, complete with elaborate staircases, elegant dining rooms, a gloomy laboratory, and see-through ghosts. The version we saw is in a very primitive stage; there’s no real story line yet and many of the rooms are only rendered in black and white. But the flowing movement and brilliant detail in a few scenes which are fleshed-out are nothing less than spectacular. Ask anybody who saw it.
None of the press or public seemed to even notice that it was far from obvious what the player was supposed to do amidst all the graphical splendor, beyond the vague notion of “exploring.” The Trilobyte trio flew back to Oregon thoroughly gratified, surer than ever that all of their instincts had been right.
Still, with publicity came expectations, and also cynicism; Bill Gates’s enthusiasm notwithstanding, a group of multimedia experts at Microsoft said publicly that what Trilobyte was proposing to do was simply impossible. Some believed the entire CES demo had been a fake.
Trilobyte remained a tiny operation: there were still only Devine, Landeros, Stein, and Moses in their digs above the tavern. Other artists, as well as famed game-soundtrack composer George “The Fat Man” Sanger, worked remotely. But Devine, who had always been a lone-wolf coder, refused to delegate any of his duties now, even when they seemed about to kill him. “I’ve never seen someone work so hard on a project,” remembers one Virgin executive. The Fat Man says that “Graeme wanted to prove everyone else a liar. He knew he was going to be able to do it.” This refusal to delegate began to cause tension with Alper and others at Virgin, especially as it gradually became clear that Trilobyte was going to miss their second Christmas deadline as well. Virgin had now sunk twice the planned $300,000 into the project, and the price tag was still climbing. Incredibly, Trilobyte’s ambitions had managed to exceed the 650 MB of storage space on a single CD, a figure that had heretofore seemed inconceivably enormous to an industry accustomed to floppy disks storing barely 1 MB each; Guest was now to ship on two CDs. Devine and Landeros agreed to work without salary to appease their increasingly impatient handlers.
Only in these last months did an already exhausted Devine and Landeros turn their full attention to the puzzles that were to turn their multimedia extravaganza into a game. Trilobyte was guided here by a simple question: “What would Mom play?” They found to their disappointment that many of the set-piece puzzles and board and card games they wanted to include were still under copyright. Their cutting-edge game would have to be full of hoary puzzles plundered from Victorian-era texts.
But at least Trilobyte could now see the light at the end of the tunnel. In January of 1993, they made a triumphant return to CES, this time with far more pomp and circumstance, to unveil the game they were now calling The 7th Guest. Alper sprang for a haunted-house mock-up in the basement of the convention hall, to which only a handpicked group of VIPs were admitted for a “private screening.” Bill Gates was once again among those who attended; he emerged a committed 7th Guest evangelist, talking it up in the press every chance he got. And why not? It blew Sherlock Holmes Consulting Detective, the current poster child for CD-ROM gaming, right out of the water. Sherlock‘s herky-jerky video clips, playing at a resolution of just 160 X 100, paled next to The 7th Guest‘s 3D-rendered SVGA glory.
When it was finally released in April of 1993, the reaction to The 7th Guest exceeded Virgin and Trilobyte’s fondest hopes. Virgin began with a production run of 60,000, of which they would need to sell 40,000 copies to break even on a final development budget of a little over $700,000. They were all gone within days; Virgin scrambled to make more, but would struggle for months to keep up with demand. “Believe it or not, The 7th Guest really does live up to all the hype,” wrote Video Games and Computer Entertainment magazine. “It takes computer entertainment to the next level and sets new standards for graphics and sound.” What more could anyone want?
Well, in the long run anyway, a lot more. The 7th Guest would age more like raw salmon than fine wine. Already just two and a half years after its release to glowing reviews like the one just quoted, the multimedia trade magazine InterAction was offering a much more tepid assessment:
As a first-generation CD-ROM-based experience, The 7th Guest broke new ground. It also broke a lot of rules – of course, this was before anyone knew there were any rules. The music drowns out the dialog; the audio is not mixable. The video clips, once triggered, can’t be interrupted, which in a house of puzzles and constant searching leads to frustration. How many times can you watch a ghost float down a hallway before you get bored?
Everywhere The 7th Guest evinces the telltale signs of a game that no one ever bothered to play before its release — a game the playing of which was practically irrelevant to its real goals of demonstrating the audiovisual potential of the latest personal computers. Right from the moment you boot it up, when it subjects you to a cheesy several-seconds-long sound clip you can’t click past, it tries your patience. The Ouija Board used to save and restore your session seems clever for about half a minute; after that’s it’s simply excruciating. Ditto the stately animations that sweep you through the mansion like a dancing circus elephant on Quaaludes; the video clips that bring everything to a crashing halt for a minute or more at a time; the audio clips of Stauf taunting you which are constantly freezing the puzzles you’re trying to solve. The dominant impression the game leaves you with is one of slowness: the slowness of cold molasses coming out of the jar, of a glacier creeping over the land, of the universe winding down toward its heat death. I get fidgety just thinking about it.
One of the game’s few concessions to player convenience is this in-game map. Yet it’s made so annoying to use that you hardly want to. First, you have to click through a menu screen which forces you to watch it tediously fading in and out, like every screen in the game. And then you have to watch the game fill in the map with colors square by exasperating square to indicate where you’ve solved the puzzles and where you still have puzzles remaining. This game would make an excellent trial of patience for a Zen school, if such institutions exist.
The puzzles that are scattered through the rooms of the mansion gate your progress, but not for any reason that is discernable within the environment. When you solve certain puzzles, the game simply starts letting you go places you couldn’t go before. In practice, this means that you’re constantly toing and froing through the mansion, looking for whatever arbitrary new place the game has now decided to let you into. And, as already noted, moving around takes forever.
The puzzles themselves were already tired in 1993. Landeros has been cheeky enough to compare The 7th Guest to The Fool’s Errand, Cliff Johnson’s classic Macintosh puzzler, but the former’s puzzles haven’t a trace of the latter’s depth, grace, wit, or originality. Playing The 7th Guest exposes a pair of creators who were, despite being unquestionably talented in other ways, peculiarly out of their depth when it came to the most basic elements of good game design.
For example, one of the puzzles, inevitably, is an extended maze, which the vast majority of players solve, assuming they do so at all, only through laborious trial and error. “The solution to the maze was on a rug in one of the bedrooms,” notes Devine. “We thought people would copy that down.” A more experienced design team would have grasped that good game design requires consistency: all of the other puzzles in the game are completely self-contained, a fact which has trained the player long before she encounters the maze not to look for clues like this one in the environment. Alternately, testers could have told the designers the same thing. The 7th Guest provides yet one more illustration of my maxim that the difference between a bad and a good one is the same as that between a game that wasn’t played before its release and one that was. “Our beta testing was, well, just us,” admits Devine.
Another infamous lowlight — easily the worst puzzle in the game in purely abstract design terms — is a shelf of lettered soup cans which you must rearrange to spell out a message. The problem is that the sentence you’re looking for makes sense only under a mustily archaic Scottish diction that vanishingly few players are likely to be familiar with.
But the worst puzzle in practical terms is actually Devine’s old abstract strategy game Spot, imported wholesale, albeit with the intelligence of your computer opponent cranked up to literally superhuman levels. It’s so difficult that even the official strategy guide throws up its hands, offering only the following clarification: “It is not necessary to beat this game to advance through The 7th Guest, and you will not be missing anything if you can’t beat it. To our knowledge, nobody has a consistent strategy to beat this game, not even Graeme!” The most serious problem here, even beyond the sheer lunacy of including a mini-game that even the programmer doesn’t know how to beat, is that the player doesn’t know that the puzzle is unnecessary. Thus she’s likely to waste hours or days on an insurmountable task, thinking all the while that it must gate access to a critical part of the plot, just like all the other puzzles. (What did I say about consistency?) Its presence is unforgivably cruel, especially in a game that advertised itself as being suitable for casual players.
None of the other puzzles are quite as bad as these, but they are samey — three of the 22 are chess puzzles, doubtless all drawn from the same Victorian book — at wild variance with one another in difficulty, and just generally dull, in addition to being implemented in ways calculated to maximize their tedium. Playing the game recently to prepare for this article, I never once felt that rush that accompanies the solution of a really clever puzzle. Working through these ones does indeed feel like work, made all the more taxing by the obstinately form-over-function interface. The best thing to be said about the puzzles is that they can all be bypassed by consulting an in-game hint book in the mansion’s library, albeit at the cost of missing the video clips that accompany their successful solutions and thus missing out on that part of the plot.
Still, one might want to argue that there is, paradoxical though it might sound, more to games than gameplay. Aesthetics have a value of their own, as does story; certainly The 7th Guest is far from the first adventure game with a story divorced from its puzzles. In all of these areas as well, however, it’s long since curdled. The graphics, no longer able to dazzle the jaded modern eye with their technical qualities, stand revealed as having nothing else to offer. There’s just nothing really striking in the game’s visual design — no compelling aesthetic vision. The script as well manages only to demonstrate that Matthew Costello is no David Lynch. It turns out that subversive surrealistic horror is harder to pull off than it looks.
As for the actors… I hesitate to heap too much scorn on them, given that they were innocent amateurs doing their best with a dodgy script in what had to feel like a thoroughly strange performing situation. Suffice to say, then, that the acting is about as good as that description would suggest. On the other hand, it does seem that they had some fun at least some of the time by hamming it up.
Indeed, the only claim to aesthetic or dramatic merit which The 7th Guest can still make is that of camp. Even Devine acknowledges today that the game is more silly than scary. He now admits that the story is “a bit goofy” and calls the game “Scooby Doo spooky” rather than drawing comparisons to The Shining and The Haunting, as he did back in the day. Which is progress, I suppose — but then, camp is such a lazy crutch, one that far too many games try to lean upon.
“The 7th Guest just kept selling and selling,” says its producer David Bishop of the months after its release. “We’d look at the sales charts and it had incredible legs. Sales were picking up, not slowing down.” By the end of 1996, the game would sell well over 2 million copies. Trilobyte was suddenly flush with cash; they earned $5 million in royalties in the first year alone. Nintendo gave them a cool $1 million upfront for the console rights; Paul Allen came along with another $5 million in investment capital. Trilobyte moved out of their little office above the tavern into a picturesque old schoolhouse, and started hiring the staff that had been so conspicuously missing while they made their first game. Then they moved out of the schoolhouse into a 29,000-square-foot monstrosity, formerly a major bank’s data center.
The story of Trilobyte after The 7th Guest becomes that of two merely smart men who started believing that they really were the infallible geniuses they were being hyped as. “Trilobyte thought they could pick up any project and it would turn to gold,” says one former Virgin staffer. “They had huge egos and wanted to grow,” says another. Even writer Matthew Costello says that he “could see the impact the attention from The 7th Guest had on [Devine and Landeros’s] perceptions of themselves.”
Despite the pair’s heaping level of confidence and ambition, or perhaps because of it, Trilobyte never came close to matching the success of The 7th Guest. The sequel, called The 11th Hour, shipped fully two and a half years later, but nonetheless proved to be just more of the same: more dull puzzles, more terrible acting, more technically impressive but aesthetically flaccid graphics. The zeitgeist instant for this sort of thing had already passed; after a brief flurry of early sales, The 11th Hour disappeared. Other projects came and went; Trilobyte spent $800,000 on Dog Eat Dog, a “workplace-politics simulator,” before cancelling it. Meanwhile Clandestiny, another expensive game in the mold of The 7th Guest, sold less than 20,000 copies to players who had now well and truly seen that the guest had no clothes.
Dog Eat Dog, Trilobyte’s never-released “workplace-politics simulator.”
Rob Landeros gradually revealed himself to be a frustrated filmmaker, always a dangerous thing to have around a game-development studio. Worse, he was determined to push Trilobyte into “edgy” content, rife with adult themes and nudity, which he lacked sufficient artistic nuance to bring to life in ways that didn’t feel crass and exploitative. When Devine proved understandably uncomfortable with his direction, the two fast friends began to feud.
The two founders were soon pulling in radically different directions, with Landeros still chasing the interactive-movie unicorn as if Doom had never happened, while Devine pushed for a move into real-time 3D games like the ones everyone else was making. New Media magazine memorably described Landeros’s Tender Loving Care as “a soft-porn film with a weak plot and rancid acting” after getting a sneak preview; the very name of Devine’s Extreme Warfare sounded like a caricature of bro-gamer culture. The former project was eventually taken by an embittered Landeros to a new company he founded just to publish it, whereupon it predictably flopped; the latter never got released at all. Trilobyte was officially wound up in January of 1999. “In the end, I never outran the shadow of The 7th Guest,” wrote Devine in a final email to his staff. “Mean old Stauf casts his long and bony shadow across this valley, and Trilobyte will always be remembered for those games and none other.”
In the aftermath, Devine continued his career in the games industry as an employee rather than an entrepreneur, working on popular blockbusters like Quake III, Doom 3, and Age of Empires III. (Good things, it seems, come to him in threes.) Landeros intermittently tried to get more of his quixotic interactive movies off the ground, whilst working as a graphic designer for the Web and other mediums. He’s become the keeper of the 7th Guest flame, for whatever that is still worth. In 2019, he launched a remastered 25th anniversary edition of the game, but it was greeted with lukewarm reviews and little enthusiasm from players. It seems that even nostalgia struggles to overcome the game’s manifest deficiencies.
The temptation to compare The 7th Guest to Myst, its more long-lived successor in the role of CD-ROM showcase for the masses, is all but irresistible. One might say that The 7th Guest really was all the things that Myst was so often accused of being: shallow, unfair, a tech demo masquerading as a game. Likewise, a comparison of the two games’ respective creators does Devine and Landeros no favors. The Miller brothers of Cyan Productions, the makers of Myst, took their fame and fortune with level-headed humility. Combined with their more serious attitude toward game design as a craft, this allowed them to weather the vicissitudes of fortune — albeit not without a few bumps along the way, to be sure! — and emerge with their signature franchise still intact. Devine and Landeros, alas, cannot make the same claim.
And yet I do want to be careful about using Myst as a cudgel with which to beat The 7th Guest. Unlike so many bad games, it wasn’t made for cynical reasons. On the contrary: all indications are that Devine and Landeros made it for all the right reasons, driven by a real, earnest passion to do something important, something groundbreaking. If the results largely serve today as an illustration of why static video clips strung together, whether they were created in a 3D modeler or filmed in front of live actors, are an unstable foundation on which to build a compelling game, the fact remains that we need examples of what doesn’t work as well as what does. And if the results look appallingly amateurish today on strictly aesthetic terms, they shouldn’t obscure the importance of The 7th Guest in the history of gaming. As gaming historians Magnus Anderson and Rebecca Levene put it, “The 7th Guest wasn’t anywhere near the league of professional film-making, but it moved games into the same sphere — a non-gamer could look at The 7th Guest and understand it, even if they were barely impressed.”
A year before Myst took the Wintel world by storm, The 7th Guest drove the first substantial wave of CD-ROM uptake, doing more than any other single product to turn 1993 into the long-awaited Year of CD-ROM. It’s been claimed that sales of CD-ROM drives jumped by 300 percent within weeks of its release. Indeed, The 7th Guest and CD-ROM in general became virtually synonymous for a time in the minds of consumers. And the game drove sales of SVGA cards to an equal degree; The 7th Guest was in fact the very first prominent game to demand more than everyday VGA graphics. Likewise, it undoubtedly prompted many a soul to take the plunge on a whole new 80486- or Pentium-based wundercomputer. And it also prompted the sale of countless CD-quality 16-bit sound cards. Thanks to The 7th Guest‘s immense success, game designers after 1993 had a far broader technological canvas on which to paint than they had before that year. And some of the things they painted there were beautiful and rich and immersive in all the ways that The 7th Guest tried to be, but couldn’t quite manage. While I heartily and unapologetically hate it as a game, I do love the new worlds of possibility it opened.
(Sources: the books La Saga des Jeux Vidéo by Daniel Ichbiah, Grand Thieves and Tomb Raiders: How British Video Games Conquered the World by Magnus Anderson and Rebecca Levene, and The 7th Guest: The Official Strategy Guide by Rusel DeMaria; Computer Gaming World of December 1990, May 1991, November 1992, October 1994, November 1994, June 1995, November 1998, December 1999, and July 2004; Electronic Entertainment of June 1994 and August 1995; Game Players PC Entertainment Vol. 5 No. 5; InterActivity of February 1996; Retro Gamer 85, 108, 122, and 123; Video Games and Computer Entertainment of August 1993; Zero of May 1992; Run 1986 Special Issue; Compute!’s Gazette of April 1985 and September 1986; ZX Computing of April 1986; Home Computing Weekly of July 19 1983; Popular Computing Weekly of May 26 1983; Crash of January 1985; Computer Gamer of December 1985 and February 1986; Origin Systems’s internal newslatter Point of Origin dated January 17 1992. Online sources include Geoff Keighly’s lengthy history of Trilobyte for GameSpot, John-Gabriel Adkins’s “Two Histories of Myst,” and “Jeremiah Nunan – An Irish Success Story” at the Jacksonville Review.
The 25th anniversary edition of The 7th Guest is available for purchase at GOG.com, as is the sequel The 11th Hour.)
