by Steven Levy
He wanted a girlfriend. The isolated times when he’d been out with members of this desirable yet elusive gender always seemed to end in some kind of disappointment. His housemates were often involved in romantic intrigue—they jokingly called the house "Peyton Place of Outer Space”—but John was rarely involved. There was one girl he saw for a couple of weeks, and had even made a New Year’s Eve date with. But she’d called him just before New Year’s. “I don’t know how to tell you this,” she said, “but I met a guy and I’m going to marry him.” That was typical.
So he kept hacking games. Just like the MIT hackers, or the Homebrewers, his reward was the satisfaction of doing it. He joined a local Atari users’ group and borrowed programs from their library to make them run faster and do neat things. He took, for instance, a version of the arcade game Missile Command and sped it up, jazzed up the explosions when one of your ICBMs stopped the enemy nuke from destroying your city. He’d show his work to others and they’d get a kick out of it. All his hacking automatically reverted to the public domain; ownership was a concept he never dealt with. When someone in the users’ group told John Harris that he had a little company that sold computer games and he’d like to market a game of John’s, Harris’ reaction was, “Sure, why not?” It was like giving a game away and getting money for it, too.
He gave the man a game called Battle Warp, which was remarkably like the old MIT Spacewar, a two-player game where ships “fly around and shoot at each other,” as John was later to describe it. Harris made around two hundred dollars from Battle Warp, but it was enough to get him thinking about having his stuff distributed more widely than through the users’ group network.
In March of 1981, Harris went to the Computer Faire in San Francisco, primarily to attend a seminar on programming the Atari given by one of Atari’s best programmers, Chris Crawford. John was extremely impressed with Crawford, a mousy fellow who bounced around when he talked and was skillful at explaining things. John Harris was on a high after that, wandering around the densely packed aisles of Brooks Hall, looking at all the hot new machines, and checking out the dozens of new software companies that had taken booths that year.
John had gotten the courage by then to ask a few companies whether they needed any programs on the Atari. They generally said no. Then he reached the booth rented by On-Line Systems. Someone introduced him to Ken Williams, who seemed nice, and John told Ken that he was an assembly-language business programmer, but he was kind of fed up with it.
Ken Williams at that time had been discovering that people who could write good assembly-language games were rare finds. He wanted to lure these assembly-language programmers to Coarsegold, California. On-Line Systems had seen explosive growth—at the last Computer Faire, Ken had been testing the waters for Mystery House, and one year later he was an established game publisher in need of products. He had placed an ad in Softalk headlined “Authors Wanted,” promising “highest royalties in the industry . . . No need to ever work anyone else’s hours again.” The ad mentioned another benefit: a chance to work with Apple guru Ken Williams, who would “be personally available at any time for technical discussions, helping to debug, brainstorming . . .” Ken was smart enough to realize that the programmers to create these products were not necessarily veteran computer workers. They might well be awkward teenagers. Like John Harris.
“Well,” Ken said to John Harris, not missing a beat, “how would you like to program amongst the trees?”
As appealing as that sounded, it meant working for On-Line Systems, which John Harris knew a little about. He knew they sold mostly Apple software. “I don’t know the Apple system,” he said, tactfully omitting that as far as he did know the Apple system he wanted to flush it down a toilet.
Ken said the magic words. “We want to expand to the Atari system. We just haven’t found anyone who can program it.”
John was almost speechless.
“Can you program it?” Ken asked.
Within a month, Ken Williams had bought John Harris a plane ticket to Fresno, where he was picked up at the airport and driven up Route 41 to Oakhurst. Ken promised Harris a place to live, and then they started talking salary. John had just gotten a raise at Gamma, so the one thousand dollars a month Ken offered him would actually have been a pay cut. John found the courage to say that he was getting more than that now. Did Ken think he could pay twelve hundred a month, and throw in the free place to live? Ken looked over at Roberta (at that time any employee in the tiny On-Line office could at any time look over at anyone else working there) and she said she didn’t think they could afford that.
Williams said, “I tell you what. How about if I put you on a thirty percent royalty basis and you won’t have to work with the company? You work out of your house and I’ll give you seven hundred dollars a month to live off of until you finish your first game, in two or three months. If you don’t have a game finished by that time, you won’t make it in this business anyway.”
John thought that sounded great. When he got home, though, his father told him he was being taken advantage of. Why not get a bigger salary and a lower percentage of royalties? What security did John have? John, who had been intimidated by the blustery Ken, did not want to jeopardize his chance to live in an atmosphere built around hacking games. He really wanted to get out of San Diego, hack games, and be happy. Even though it might mean less money, he’d hold on to the thirty percent royalty.
It was the most lucrative decision he ever made.
• • • • • • • •
Ken Williams had purchased several houses around Oakhurst for the benefit of his programmers. John Harris moved into the one called Hexagon House, named after the shape of the upper floor, which was the only part visible from the road: it jutted above the rest of the house like a large solid gazebo. From the front door, the living room and kitchen were visible; the bedrooms were downstairs. Living there with John was Ken’s twenty-year-old brother, John Williams, who ran On-Line’s advertising and marketing division. Though John Williams liked Harris, he considered him a nerd.
The first project that John Harris had mentioned to Ken was inspired by the arcade game Pac-Man. This was the hottest coin-operated game in 1981, and would soon be known as the most popular coin-op of all time. John Harris saw nothing wrong with going to the arcade, learning the game in and out, and writing his own version to run on the Atari 800. To a hacker, translating a useful or fun program from one machine to another was inherently good. The idea that someone could own Pac-Man, that clever little game where ghosts chase the dot-munching yellow Pac-Man, apparently was not a relevant consideration for John Harris. What was relevant was that the Pac-Man game seemed a natural fit to the Atari’s features. So even though he personally preferred games with space scenarios and lots of shooting, John suggested to Ken that he do an Atari 800 Pac-Man.
Ken had already been marketing a Pac-Man look-alike for the Apple under the name of Gobbler. The program had been written by a professional scientific programmer named Olaf Lubeck, who had sent Williams the game, unsolicited, after seeing the “Authors Wanted” ad. The program was selling around eight hundred copies a month, and Ken had arranged with Lubeck to duplicate it for the Atari home computer.
John Harris, though, was appalled at the Apple game. “It didn’t look spectacular, no animation,” he later explained. “The collision detection is very unforgiving.” Harris did not want Olaf to compound the error on his beloved Atari by translating the Apple game bit by bit on the 6502 chip, which the Apple and Atari shared. This would mean that none of what John considered the superior Atari features, most of which were housed on separate chips, would be utilized. The thought was horrifying.
John insisted that he could do a better-looking game within a month, and Ken Williams took Lubeck off the project. John Harris embarked on a period of intense hacking, often wrapping around till morning. John’s style was freewheeling. He improvised. “Whatever my mind is doing, I just let it flow with it .
. . things come out pretty creative,” he later explained. Sometimes John could be sensitive about this, particularly at times when a more traditional programmer, armed with flowcharts and ideas about standard structure and clear documentation, examined his code. When John left Gamma Scientific to move to Coarsegold, for example, he worried that his replacement would be someone like that, who would throwaway all his clever code, replacing it with something structured, concise . . . and worse. As it turned out, Gamma considered six programmers, five of whom “had degrees coming out of their ears,” John later said. The sixth was a hacker with no degrees; John begged his bosses to hire the hacker.
“But he wants as much money as the people who have degrees,” the boss told John.
John said, “He’s worth more.” His boss listened. When John broke this new employee in and explained his system, the new hacker became very emotional over John’s code. “You program like I do!” he said. “I didn’t think there was anyone in the world that does this!”
Working with large conceptual blocks and keeping focused, John had a Pac-Man-style game running on the Atari in a month. He was able to use some of the subroutines he had developed in earlier efforts. This was a fairly good example of the kind of growth that creative copying could encourage: a sort of subroutine reincarnation in which a programmer developed tools that far transcended derivative functions. One day, John’s subroutines would be modified and used in even more spectacular form. This was a natural, healthy outgrowth of the application of hacker principles. It was only too bad that this Third Generation of hackers had to write their own software tool kits, supplementing them only by haphazard additions from users’ groups and friends.
The Pac-Man game looked remarkably like the arcade version. It might well have been one of the best assembly-language programs written so far for the Atari Home Computer. But when Harris took his work to Ken Williams, there was a problem. Lately, some companies were insisting that the copyrights they owned on coin-operated games made unauthorized home computer translations illegal. One of the biggest owners of copyrights was Atari, and it had sent the following letter to small publishers like Brøderbund, Sirius, and On-Line:
ATARI SOFTWARE PIRACY THIS GAME IS OVER
Atari is a leader in the development of games such as Asteroids™ and MISSILE COMMAND™ . . . We appreciate the response we have received from videophiles of the world who have made our games so popular. Unfortunately, however, there are companies and individuals who have copied ATARI’s games in an attempt to reap undeserved profits from games they did not develop. ATARI must protect our investment so that we can continue to invest in the development of new and better games. Accordingly, ATARI gives warning to both the intentional pirates and to the individuals simply unaware of the copyright laws that ATARI registers the audiovisual works associated with its games with the Library of Congress and considers its games proprietary. ATARI will protect its rights by vigorously enforcing these copyrights and taking the appropriate action against unauthorized entities who reproduce or adapt substantial copies of ATARI games regardless of what computer or other apparatus is used in their performance . . .
Ken Williams knew that Atari had spent millions of dollars for the rights to Pac-Man. After looking at John Harris’ brightly colored, fast-moving, nonflickering duplication of the arcade game, he realized it was such a faithful copy that it was unmarketable. “It looks too much like Pac-Man,” he said. “You’ve wasted your time, John Harris.” He suggested that John alter the game. Harris took the game home and reprogrammed the graphics. This new version was virtually the same; the difference was that the ghosts, those goofy little shapes that chased the Pac-Man, were wearing tiny mustaches and sunglasses. Incognito ghosts! Perfect ironic commentary on the stupidity of the situation.
It wasn’t exactly what Ken Williams had in mind. For the next two weeks, John and Ken consulted with lawyers. How could they keep the essence of Pac-Man and still keep Atari at bay? The lawyers said that the only thing Atari really owned was the image of the character, what the game looked like.
So a new scenario was developed, with the unlikely theme of preventive dentistry. Ken’s brother John Williams suggested the ghosts be replaced with “happy faces.” They would spin and flip around. John Harris replaced the yellow Pac-Man with a set of clicking false teeth. Instead of dots, John drew “lifesavers,” and programmed a routine that would occur when the player cleared the dots—a toothbrush would appear and brush the teeth. None of this was difficult to program. John Harris simply drew the new images on shape tables and wrote them into his existing machine. One of the wonderful things about the computer was that you could change the world on impulse.
The lawyers assured Ken that this new Jawbreaker scenario presented no problem with Atari. They did not know Atari. It was a company owned by the Warner Entertainment Conglomerate; it was ruled by a former textile executive who saw little distinction between computer software and any other consumer item. Since engineers no longer ran Atari, the company had been characterized by a bureaucracy which stifled hacker impulses. Programmers at Atari were paid far less than the astronomical sales figures of their games would seem to call for, and convincing the marketing “experts” to release an innovative game was a formidable task. Atari would not include the name of the game programmer on the package; it even refused to give this artist credit when the press requested an author’s name. When some of the company’s top programmers complained, the textile alumnus who ran Atari reportedly called the hackers “towel designers”: Those hackers were among many who quit to form companies which would decimate Atari’s market share of game cartridges.
Atari did not seem to address this loss outright, but instead focused its creative efforts on litigation and high-rolling licensing of seemingly failure-proof properties from other media, from coin-operated games to movies. A prominent example was Pac-Man, for which Atari spent millions. The idea was to first convert the game to the VCS game machine, then to the Atari home computers, the 400 and 800. The two divisions were separate and competitive, but both shared the problem of disappearing programmers. So imagine the joy of the executives in Atari’s Home Computer Division when one day, out of the blue, some random person sent Atari a copy of a program that had been circulating around the users’ groups that summer of 1981. It was a brilliant version of Pac-Man which ran beautifully on the Atari 800.
It was the result of a classic John Harris real-world goof-up. When John had been working on the Jawbreaker revision, some people at a computer store in Fresno heard rumors of a brilliant Pac-Man hacked by the skinny, nervous kid who would often drop by and check out peripherals and software. They asked John Harris to show them the game. Without a thought to such nonhacker restrictions as corporate secrecy, John Harris drove down and proudly watched them play the version in progress, and saw nothing unusual about their request to borrow a copy of the disk. He left a copy there, went back to the Hexagon House, and continued writing his revision.
Copies of the game began circulating through users’ groups across America. When it reached Atari, people there called all the software companies they could think of to find its author. Eventually, they spoke to Ken Williams, who later recalled an Atari executive telling him that he was in possession of a Pac-Man game of obviously superior quality and was looking for its creator.
“Tell me about the game,” said Ken, and the Atari man described it as having happy faces. “That’s John Harris!” said Ken. The Atari man said he wanted to buy the program from John Harris. Ken had John Harris return the call to Atari’s head of acquisitions, Fred Thorlin—from Ken’s office. According to Ken, Thorlin was wild about John Harris’ game. He promised Harris a large royalty, mentioned a contest Atari was running for best software program, with a twenty-five-thousand-dollar prize, and said none of the entries so far had come close to Harris’ game.
But John Harris remembered how mean Atari had been when he had been trying to learn assembly language. He knew that it
had been Atari’s letter to On-Line that was forcing him to do all this revision in the first place. Atari had acted, John later said, like “a bunch of babies,” holding on to information like a selfish kid protecting a toy from his playmates. John Harris told Ken that he would not consider having his name on anything published by Atari (not that Atari had mentioned putting his name on the program), and that he would finish Jawbreaker for Ken.
Jawbreaker was an instant bestseller. Almost everyone who saw it considered it a landmark for the Atari Home Computer. Except Atari. The men who ran Atari thought John’s program infringed on their right, as owners of Pac-Man, to make as much money as they could from the game, by marketing it in any way they saw fit. If Ken Williams released a game that gave a player the feeling he was playing Pac-Man, especially if John Harris’ version was better than the one Atari’s programmer might come up with, that player would not be likely to buy Atari’s version of Pac-Man. And Atari felt that its purchase of the Pac-Man license entitled it to every penny to be earned from home computer games that played like Pac-Man.
It was a challenge to the Hacker Ethic. Why shouldn’t Atari be happy with a royalty paid by people who wanted to hack Pac-Man code and eventually improve the game? Did the public benefit from one company “owning” a piece of software and preventing others from making it more useful?