Bill and Andy's Excellent Adventure II
Bill Bill was on the project first, when General Magic was an Apple project called Pocket Crystal. Crystal was the putative answer to the question, "Where do we go after the personal computer?" It was hatched in 1979 by Marc Porat, an MBA refugee from the Aspen Institute who had barely arrived at the company. A non-techie, Porat could perform no magic with code, but he was a wizard with a business plan: The Pocket Crystal strategy, thick as a phone book, was a clearly outlined and seductively convincing document justifying the pursuit of what he called the "militantly humanistic" paradigm outlined in the mission statement above. The plan, combined with Porat's astonishing persuasive powers (The Wall Street Journal referred to him as a "Silver tongued devil"), won John Sculley's enthusiastic go-ahead. But Porat's most amazing feat was probably convincing Bill Atkinson and Andy Hertzfeld to link forces with him and Pocket Crystal. Bill Atkinson was no easy mark. Though in recent years he has finally achieved hard-earned credit and fame as the author of HyperCard, the 42-year-old software designer can recall times when things were different. As a neurosciences student at the University of Washington, Bill did some virtuoso work in computer imaging of brain maps, but was deeply hurt when the work was published with what he felt was insufficient credit to his contribution. Eventually he left academia and in 1978 went to Apple (employee 51), but the aforementioned slight was amplified when, after Atkinson made invaluable contributions to Apple's Lisa (he was the author of its groundbreaking graphic routines and a key interface designer), the Apple brass decided to trumpet the project's managers in the press tour, and Atkinson's achievements went virtually unnoticed. (Apple did reward him by making him an Apple Fellow in 1983, an honor involving stock options and a degree of freedom.) Bill's problem with his employer's oversight was not so much ego, as a matter of his deeply ingrained sense of fairness. Bill has a radar for the personal angle, and the idea of one person gaining an unearned edge over another is loathsome to him. One-on-one he is an intense communicant. Bill is an eye-contact person, giving you total attention, really wanting to know how you are doing, how you are feeling. He hugs. And he thinks that "business as usual" is no excuse for not doing what's right. The second thing crucial to Bill is his need to get his products out into the world. He bears scars from those times when a project of his failed to reach the public. He loved the idea that Apple bundled his MacPaint with every Macintosh, and he was crushed when the company decided that his post-Mac project, a flat-pad communicating computer called Magic Slate, was too esoteric a product to begin developing in 1985. He went into a depression, not working for months, until one night he wandered out of his house in the Los Gatos hills, stared at the star-filled sky, and had an epiphany: In the face of the awesome celestial epic, what was the point of being depressed? All you could do, really, was use your abilities to do what you could to make a little part of the universe better. And Bill Atkinson went back into the house and began using his abilities to work on a new project that would become known as HyperCard. What he wanted most was for John Sculley to agree to bundle it with every Mac. No use developing a great product unless it shipped. Once he had that assurance, the deal was done. And, by providing an easy-to-use notecard metaphor that allowed people to set up, access, and manipulate diverse strands of information, HyperCard was a triumph. When Marc Porat approached Bill to join up with him, Bill was wary. Bill liked the idea, but his alarms went off at the idea of joining this guy. What changed his mind was the way Porat was going about things. He realized that for such an ambitious project to take wing, it needed to seek manufacturing and communication partners. John Sculley, in a break with Apple tradition, agreed. The partners he had in mind were Sony and Motorola. "That's what first brought me into the picture, when he started saying Sony and Motorola," says Bill. "The idea of Apple making a $200 anything was ridiculous to me. Apple couldn't make a $200 blank disk, and for this to become a widespread consumer product it has to eventually get down into that range. But when we talked about an open consortium that a lot of manufacturers would join in, I thought, 'Wow, this might really work.' And then I said, 'How can I help?'"
Andy A lot of people would have thought that Andy Hertzfeld would never have gone back to working for a company. He is a natural rebel, a free spirit. His ideas of paradise were embodied in the circuit designs of the Apple II computer, an object he loved passionately and used as a philosophical touchstone. It was like him: willing to take weird twists and brilliant nonintuitive leaps to pursue a nexus of art and technology. The genius behind the Apple II was Steve Wozniak; Woz did it to impress his friends and enrich the world. When Andy, as a computer science student at Berkeley, bought an Apple II in 1977, it changed his life. He played with it all the time, learning its secrets, writing nifty programs. A year later he went to work for Apple as a programmer, and a dream came true - he became pals with Woz. But as the company grew, Andy became increasingly disenchanted. His penchant for openness, for saying what he felt, didn't click with the corporate ethos. And when Apple had its first major layoff, in 1981, he was crushed. He would have left, had not Apple co-founder Steve Jobs literally scooped up his belongings and led him over to the small group working on the computer called Macintosh. Andy would help build its programming toolbox and much of its new interface. The Mac team wasn't like a company - it was a family. A transplanted Philadelphian, Andy found his home in the corridors of Apple's Bandley building where the Mac team built its masterpiece. His relationship with Steve Jobs - the digital impresario who alternately inspired and insulted the young, idealistic engineers to go beyond themselves and produce an "insanely great" design - had its stormy aspects, but was essentially a close collaboration. Jobs was like a dominating father to the team, with all the good and bad baggage that implies. Still, as the Mac headed towards completion, Andy came to understand that it would be difficult for him to thrive in a corporate environment. The bane of his existence was his immediate superior, Bob Belleville, a former Xerox PARC engineer several years his senior. One of the worst days of Andy's life occurred when Belleville gave him an unfavorable job review - at the same time Andy was working all his waking hours performing magic on the Mac. Andy left Apple in March 1984, even before Jobs, his mentor, was bounced out of the company. Working on his own, often without prospect of payment, Andy set about supporting the computer in which he had invested so much passion. Working as an independent programmer, he wrote Switcher, enabling Mac to handle more than one application at a time; besides licensing the program to Apple and others, he distributed it electronically, for free. Then he wrote Servant, an update to the Mac Finder that Apple eventually bought from him when it was producing the MultiFinder upgrade. He sped up the Mac graphics, and, for a song, sold the technique to Radius and Apple. He also wrote the initial software for Radius's full-page screen, as well as Thunderscan, which turned an ImageWriter printer into a low-cost scanner. Perhaps one of his least-known achievements was a graphically oriented television operating system, concocted for the Frox company, affiliated with frogdesign. Steve Jobs later tried to convince Andy to work for his post-Apple company, NeXT. But Andy wanted no part of a corporation. Until Bill told him about Pocket Crystal. "Bill called me to expose me to this idea," recalls Andy. "Marc was over at my house the next day. Frankly, the most exciting thing about the project initially was the opportunity to work with Bill. I saw Marc as being very bright, creative, and articulate, but I thought he might be...you know, not genuine. He exaggerated. He was such a good salesman I reacted a bit negatively. But then, in working with him, I finally saw his core values are right. He has a lot of integrity, and he wants to work for the right purposes, and that's what's important to me. We've had our frictions from the beginning but Marc exceeded my expectations." Andy is talking about the eventual alliance Porat pulled together: not only Sony, Motorola, and Apple, but also AT&T, Philips, and Matsushita. An all-star team of partners. Andy and Bill like to tell the story of how Marc arranged a meeting at Sony. "We went there, on a Monday or Tuesday, and we gave our sort of vision pitch, and our demo of our current software, which I think was at that time still running on the HyperCard - " "That was September of '90," recalls Andy. "Sculley brought us there and introduced us. We pitched them on our thing, and I remember at the end of the day, (Sony Managing Director) Norio Ohga asked us when we were leaving. We said Friday. He said, 'Well, here, we do things by consensus.'" The Japanese way. The Magic crew sagged at the prospect of waiting for months until word was given. "And then he looked at his vice president and said, 'I want consensus by Thursday.'"
Making Magic Happen Porat and his two prize recruits staked out a new company, selecting its name in honor of Arthur C. Clarke's maxim that "the best new technology is indistinguishable from magic." Apple had a 10 percent stake in General Magic, as did its first partners, Sony and Motorola. The company kept the remaining 70 percent. With Andy and Bill on hand, hiring the best talent was easy for General Magic. Former Apple people, of course, would form the core team. There was Dan Winkler, author of Hyperscript; Bruce Leak, head of the Quicktime team; and Phil Goldman, who worked on Apple's system software. There was Susan Kare, the artist who had drawn the icons, crafted the fonts, and perfected the look of the Mac interface. On the marketing side there was Joanna Hoffman, who was the original marketing person on Macintosh. Even Jane Anderson, John Sculley's personal publicist, came to General Magic. And then there was a whole new generation of engineers, people in their twenties who reminded Bill and Andy of themselves when they had worked on the Mac team. One, Darren Adler, had been the technical lead of Apple's System 7 team - at age 24. Bill and Andy weren't doing it on their own, they were spearheading a team of self-proclaimed Magicians. In its early days General Magic radiated a lot of excitement. It had a very nice interface that obviously drew upon Bill's HyperCard and Andy's Mac interface, with the unmistakable graphic imprint of Susan Kare. The basic screen looked like a desktop with various tools; on the desk was a postcard that one could fill out and send to anyone. Bill explains that the starting point was HyperCard, and from there they built a telecommunications interface: "The first concept we latched onto was Telecards, something dropping out of the sky and landing in your pocket." "Bill spent our first week working on it, and we had a postcard that looks very much like the one we're shipping today," says Andy. "Andy made a little server on the Macintosh," recalls Bill. "Within three days of having the idea of wanting to do a service, we had a HyperCard mock-up on the Mac with little Telecards and a server going that actually sent the mail," says Andy. "Then we came up with one of our seminal ideas - rubber stamps with semantics." "You want things to be emotionally and visually appealing, not just functional," says Bill. "Instead of raw text, you want some graphics to really help make it appealing, but even the simplest graphics editor we could think of has the problem that it takes time to draw - and an even worse problem that most people can't draw something beautiful, something they love. They can scribble, but they can't draw something attractive. The rubber stamps are an outgrowth of those little stickers kids use. People can arbitrarily assign properties to them - there are scripts inside them. For instance, the stamp marked URGENT has a script that marks it as such." "By the beginning of 1991, we came up with this idea of places in the software," says Andy. "We didn't have (interface features such as the) hallway or the Downtown street yet, but we had the idea of making each scene in the software look like a physical place. We lived with that for a while, but it still wasn't satisfying to us. We even had the concept of Downtown, but we didn't have the framework of Downtown, where it scrolls as you walk by. Finally in January of 1992, one of our engineers, Kevin Lynch, really had the seed that instigated this idea, with the hallway of doors, and the street Downtown." The interface has evolved to a geographical depiction of a portion of cyberspace. (The graphics are by necessity very simple, since low-cost liquid crystal screens cannot handle subtleties.) Those who have read Neal Stephenson's Snow Crash will instantly see its relationship to the "Metaverse" depicted therein - a virtual world where one can conduct all sorts of transactions, gather objects, and, above all, maintain a sense of place. To buy things, go Downtown and go into the electronic shopping mall. To scan newspapers, go to the Newsstand. Eventually, the local pizza shop will show up on your personal Main Street - go in, scan the menu, and place your order, hold the anchovies. And incidentally, the interface does not use handwriting recognition. You can use a pen or your finger to draw or write on the screen, but digital text is entered with a virtual keyboard - which, surprisingly, doesn't work too badly for short messages. "Actually," says Bill, "we spent a lot of time trying to make it so you type less, because we don't think that the typing is very good." Some critics will undoubtedly complain that the system looks more like a graphic Adventure game than a platform for conducting business. A toy, not a tool. But then, they said that about the Mac, too. And now even hard-core suits diddle their mouses and trip through icon space. Likewise, I think that the Magic approach will scale nicely as the market for hand-held out-of-body experiences grows. (It would be nice, though, to have more shortcuts built in for adept users - sort of a "beam me up, Scotty" button that intuits where you want to go at any point.)
The Telescript Solution Early in the development of the interface, though, the Magic people discovered that a crucial piece of the puzzle was missing. In order to realize its vision, General Magic had to pay more attention to the underlying protocols of communications technology - a black art that no one on board had really mastered. They required an entirely new way of processing messages. Fortunately, they came across Jim White, a legendary telecom programmer who, independently, had been thinking of a solution. White, the wizard who had developed the standard x.400 networking protocol, envisioned a network which was not merely a pipeline for messages, but a platform of its own - a seething system where every message was literally a computer program that executed itself as it went out on the Net. Each message would be a software agent, a sort of digital servant performing a task for the person who sent it out. Sounds scary, and actually it is. But if you could control the agents you'd have something terrifically powerful. "At that time (1990), we weren't thinking in terms of agents, we weren't thinking in terms of making the network into a platform," says Bill. "When Jim first brought up agents, I had in my mind knowledge-navigator types anthropomorphic, obsequious little sci-fi agents. But then I realized that he wasn't talking about these human personalities so much as traveling programs that can take my will with them." The system is called Telescript, and it's the most interesting part of General Magic technology. In previous network architectures, a message was like a letter zipped down a chute - nothing happened to it between the time it was sent and the time it was received. But in Telescript, the network isn't a system of chutes but a sort of beehive, with the messages actively poking around and doing things. The closest thing to it is Postscript, which changed the way printers worked by substituting rigid output from a computer application to a vibrant page description language that allows for the open-ended capabilities that come from computer programming. Marc Porat, with typical flourish, has described Telescript as "machinery that gives people the magical ability to project their desires into cyberspace; that ties together networks of all sorts into one seamless, invisible, central nervous system..." Telescript turns out to be the ace up the General Magic sleeve - but it wasn't invented there. "Jim had it pretty much conceptualized before he came to Magic - it was more like, 'Does he trust us enough to give us this ground-breaking idea?'" says Andy. "At first, Bill and I didn't appreciate it for what it really was. I had the light- bulb experience when he gave a video show to Motorola. It just hit me that an incredible shift would happen if we could have this Telescript communication infrastructure and I came to work the next day a changed person. I had thought the most important thing we were doing was making this user interface, and boy, it's not." "It's the enabling piece," explains Bill. "The real simple kind of example is instead of calling up Dow Jones every morning and saying, 'How's Apple stock, is it over a hundred yet?' I'll send an agent out. I don't even have to know what an agent is, I just fill out a little form and this agent goes out and lives at Dow Jones, and six months later, a card drops in my pocket that says, 'Apple stock just broke a hundred.'" "The beauty of Telescript," says Andy, "is that now, instead of just having a device to program, we now have the entire Cloud out there, where a single program can go and travel to many different sources of information and create sort of a virtual service. No one had conceived that before. The example Jim White uses now is a date-arranging service where a software agent goes to the flower store and orders flowers and then goes to the ticket shop and gets the tickets for the show, and everything is communicated to both parties." As with its spiritual cousin PostScript, very few people will actually program code in Telescript. It's invisible. But everyone who uses a Telescript network will enjoy its advantages. "There will be some guys that do the Telescript, but the bulk of the users here will be dropping a stamp on something or using a form," says Bill. "So to request a meeting with someone, you get this form, you just fill in the blanks and send it. The guy receiving it presses 'Yes I will attend,' 'No I won't attend,' or writes a counteroffer, sends the reply back. If it's 'Yes I will attend' the device puts the meeting in your datebook. The user is not going to be aware of the Telescript underneath." Some folks have noted that those Telescript agents bear an uncomfortable kinship to computer viruses - autonomous programs that do their jobs beyond the control of the person who launched them. Jim White, acknowledging the truth of this, has installed all kinds of restrainers into Telescript to control potential rogue agents. First, the language is built in a manner that makes it next to impossible for viruses to get out and infect other computers (in programmer's terms, it's interpreted instead of compiled). Second, each agent has a "permit," so a person can't send a billion of them out into the Cloud, intentionally or otherwise. Third, no agent can exist without being positively identified, through digital authentication technology, with its originator. Since these agents will be performing important tasks - arranging trips, spending money, maybe even zipping medical records from one doctor to another - we can only hope Jim White has it right. So, General Magic's technology finally came into focus. It would have two products, both software systems that the company would license to others. The first is Telescript. The second, the visually oriented operating system, is called Magic Cap, an acronym for Communicating Applications Platform. (Sometimes Bill and Andy wryly add that the third General Magic technology is "Alliance Management," Porat's job.) Though Telescript can be used without Magic Cap - supposedly Telescript will find its way onto personal computers and personal digital assistants like the Newton - the two presumably work best together. This double dose of magic will live on devices sold by Sony and Motorola later this year. (Alliance members Philips and Matsushita will soon come up with their own hardware devices.) These hand-held communicators will link up to AT&T's brand new Telescript-oriented division, PersonaLink Services, described by AT&T's flacks as "the foundation or 'host' for an electronic community - a sort of town center for...customers: merchants offering their products and services, and people meeting, shopping, working, and playing." Initially, the cost will be at the Newton level - roughly $700. Probably once you get all the accessories and cards you want for the device, the thing will run you close to a thousand bucks. This is too high for the everyday users that, Bill and Andy hope, will one day form the bulk of their customers. The idea is that by the time the millennium kicks in, we'll be buying the doodads at Kmart for about two hundred bucks apiece. It turns out that the General Magic mission is a long-haul deal, and not only because the initial cost limits the user base to gizmo nuts and corporate users. "We're prepared for a relatively slow start, especially from our Macintosh experience, where it took until 1986 for the Macintosh to find its legs," says Andy. "Same thing here. The screens aren't really good enough to satisfy the customer we have in mind these days. It's too low-contrast. You have to kind of take this leap of faith. They don't run long enough on batteries. And the wireless communication infrastructure just isn't here yet. It's not even a question of being too expensive. It's just the infrastructure isn't filled out." "The data isn't available," says Bill. "It's the chicken-and-egg problem," Andy says. "But you gotta start somewhere and we have to create the pull to get the wireless infrastructures deployed, to get the screens to improve. We have to show there are good ways to use these screens. So in our over-optimistic way we just do our thing and the rest will come."
Newton Not that there weren't bumps along the way. One of the most traumatic events during the development process was the spiritual abandonment of the one Alliance member that seemed the most reliable: Apple Computer. Apple had a team developing a computer called Newton. Its leader was former Mac member Steve Capps, formerly a close friend of Andy and Bill. Emphasis on formerly. Newton was originally supposed to be a much different sort of machine than the General Magic devices: bigger, less a communicator than a computer. Price point: $2,000. But around 1991 or so, the Newton took a hard left towards General Magic country: a communicating personal digital appliance. By the time Apple marketing started publicly talking about Newton in 1992, it was difficult to determine exactly what would differentiate a Newton from a Magic device. Meanwhile, Apple distanced itself from Magic, as Sculley began touting Newton as his personal contribution to the paradigm shift. "When Apple first started saying they were going to have a personal communicator and reinvent telephony," says Bill, "We started thinking, 'Gosh, they've changed their story to Marc's.' I worried that Newton would taint the notion of personal communicators; people would say, 'They don't work yet.'" "That thing is not a communicator in any sense of the word," says Andy of Newton. "Yet they're positioning it as a communicator and I really think that maybe they got that language from us." Apple's turnaround was one thing, but Andy took the shift as a personal betrayal on Steve Capps's part. Andy thinks Capps should have explained things to him and Bill. For the past two years Andy has avoided his old friend. Now, months after the Newton's appearance to decidedly mixed reviews - which confirmed General Magic's early decision to eschew not-ready-for-prime-time handwriting recognition - Andy is thinking of reconciliation. "I need to talk to Capps about righting things, because I am confused about certain ways he acted," says Andy, who still professes that he wishes Newton well. Not that he thinks Apple's personal digital assistant is a winner. "I feel awkward about contacting Capps because if he asked me, 'What do you think of the Newton interface?' I can't say, 'Oh I think it's really good,' because that would be lying. But I still love him as a person." The statement is pure Andy - full of emotion and honest to a fault.
How to Change the World Now that they've designed the technology, Bill and Andy love to talk about its implications - how, in effect, General Magic can change the world. "I'd be interested to see what happens in schools when enough of the kids have these," says Bill. "You'd have things like the teacher giving out homework by beaming it up to a bubble on the ceiling which beams it out to other people, and kids are actually able to collaborate by beaming something down to one of their friends. They could arrange a lunch date in the middle of class. Without passing notes." Andy is bouncing in his chair, eager to chime in. "Another thing that's interesting is that we'll get reviews of things. You know, I love arts, music, and books. The reviews are all going to have buttons on them now: 'Buy It.' Or, you know, everything gets intermingled, everything can be a link to something else." "Or like I'm listening to the FM radio and I hear a song I like," says Bill. "I press a question mark and I instantly see the name of the artist and the other songs on the album and the price of the album. And there's the damn BUY button, and the next day a CD will show up in my house." "The General Magic unit will have a radio on it?" I ask. "Yeah, it'll be like a Sony FM Teletouch - it has an FM radio in it, with subcarrier reception so it can get the data at the same time," says Bill. "And what'll happen is, you'll be buying music directly from a specific direct-mail CD place and they can afford to discount - no middleman, so they can afford to sell it at half the going rate, I can buy $7 CDs instead of $15 CDs. And you know what's cool for the radio station? The radio station gets tagged, because this device knew what station you were listening to, and what time. So in your order, they include the information, and the CD guys can send a fee over to the radio station to subsidize them for playing it." "That's called payola," warns Andy. "It could lead to radio stations tomorrow with no commercials," says Bill. "But the content will be a commercial," I argue. "What happens with the radio stations, what are they going to play? If they get a cut of the sales are they going to play the widest possible music, or just the most popular?" Bill thinks about it. "There will be a station that plays the hit discount music, and there'll be another station for guys who want to hear classical." "There's a whole question of advertising and how that'll work in an interactive world," says Andy. "Like coupons, for example," says Bill. "You know, you see these coupons in the newspaper, '10 cents off on this product.' Why do they do those coupons? It's just because they're trying to monitor advertising. Everybody spends tremendous bucks on advertising, they want to know which advertising campaigns are working. So they establish a flow of coupons into the store, get a regular base line going, kick in an ad campaign, and see if it goes up or down. Okay? But with a Telescript device you can much more directly tie into customer behavior. Lets say they've got an ad on the radio about a car. You press a question mark and there's a coupon: 'Mail me a brochure.' They know exactly which version of the ad was more effective than the other version." This changes the entire relationship between broadcasters, listeners, and advertisers, I remark. Bill agrees. "Radio guys work for advertisers, their business is to deliver customers. But all this 'push' advertising is on the way out. For example, when people get cable and stuff, and start seeing TV without advertisements, they don't go back to it. And people will start to really want a 'pull' kind of advertising, they're gonna reject the inundation of push advertisement. The world's gonna go more to narrowcast for specific markets, and it's gonna go more toward pull of information rather than unwanted push of information." "One possibility is you will actually get paid to look at the ad," says Andy. "The device knows when you've opened the mail. It can't know if you've read it, they'll pay you if you just open it." "They might even pay you more if you open it and leave it open for a minute than if you open and close it right away," says Bill. Finally, the compulsively honest side of Andy Hertzfeld leads to a disclaimer: "I feel compelled to say that all of the stuff we're talking about happens without General Magic, without Telescript, without Magic Cap. It's just that we can make it happen faster and we can put a human face on it. Make it so ordinary people like it. That's our mission."
Pressure Drop But in November, as the rollout approaches, things are getting a little hairy. When they started General Magic, Bill and Andy felt that they and their fellow senior citizens in the forty-plus demographic would not engage in the "90-Hour-A-Week-And-Loving-It" frenzy of the Macintosh experience. Ten years removed from that craziness, each has other interests. Andy is involved in a long-term relationship and spends a lot of time with his enamored's 11-year-old son; Bill has two daughters upon whom he delightedly dotes. A lot of Bill's energies are also consumed overseeing construction of a stunning new house. And, since they both have hit the big 4-0, there is the spectre of stamina-gap. Indeed, General Magic is not as totally consuming as Macintosh, where no outside lives existed. But it does require a maniacal devotion, especially in these crucial days. Andy in particular has been coding like a madman. "The pressure right now, really for the last six months, is incredibly intense," says Andy. "I feel we're in an X-ray machine - you see everyone's bones sticking out. The pressure strips away the superficial, and you see your flaws. The pressure does that. It also makes it hard to relax. I've had to make sacrifices in parts of my life for it, but we believe it's all worth it - we're attempting something really important. I feel much more responsibility to the team than I felt on Macintosh. Then I luxuriated in being a troublemaker. I was a thorn in Steve's side in some sense, because any little problem I found, I magnified. I turned it into a big deal - 'I'm going to quit if you don't fix this!' And that created a big problem for Steve. Here, I'm responsible. I still find the problems but I can't complain, because I have a responsibility to all these people. And I guess it's more adult." "Actually, lately, I've sort of taken over the thorn-in-the-side role, and you're the one who fixes it," says Bill. "At the beginning I was programming a lot and Andy was programming a lot and now I'm doing very little programming and almost all design kibitzing and testing, and I go around and work with each of the guys, say, 'Well, what are you working on?' And helping in design areas." "Bill guides and shapes my work on a daily basis, or even sometimes an hourly basis, whenever I have a new design," says Andy. "I'm pretty good at coming up with things to try out. Bill's much better at figuring out which ones of those are the real paths we should pursue. I'm great at starting things, and I'm not so great at finishing, because I keep wanting to invent, and so one of the things Bill really does is steer me and help me work on the right things. I constantly count on him to prioritize what I do." "Andy is fearless and undauntable." "I'll try anything." "He has no fear. And no fear of huge amounts of work." "I don't need to have a design all worked out before I start it, and some people would consider that a weakness. But that's how you make things happen. If you waited until you had perfect clarity, you might never start. "Some people sit in a car in a parking lot and turn the steering wheel left and right, left and right, aiming the steering wheel just perfect before they start the car," says Bill. "Andy starts the car and starts moving, and then gets his hands on the steering wheel. And fortunately, he steers good enough once he's got into it that he doesn't knock over too many pylons."
"Are you guys going to be here ten years from now?" I ask Bill and Andy.
"Yeah," says Bill.
Andy looks startled. "At General Magic?"
"Oh, I don't know," says Bill.
"I don't think I'll be at General Magic," says Andy, "because I'm better at starting things.... There'll be new adventures - "
"I would guess that General Magic will have been taken over by weird people who don't know what they've got," concedes Bill.
"It's not so much that," says Andy. "I don't know if I have the wherewithal inside me as a person in his forties to try and start another platform. I think it's most likely for me to want to go work as an independent artist."
"I see you ten years from now with five kids," says Bill. "At least three. I see you working more on personal stuff."
"I'd love to write software that teaches kids about technology, that would be a fun area," says Andy. "I want to write a novel; that's very important to me. And I'm definitely around at General Magic the next couple of years, because it's so exciting to me."
Butterflies Fast forward to a night several weeks later, January 1994. The Yerba Buena Room of the San Francisco Marriott. A cavernous auditorium - the front is almost imperceptible from the back, as if the room contained an internal horizon. On the stage a dozen technicians are setting up the first public display of General Magic technology, scheduled for the next morning as a keynote for the Macworld Expo. A table holds two Macs, two palm-size prototype devices (with video cameras overhead to capture their screen images), and a futuristic scrim with flags reading "Telescript" and "Magic Cap." A far cry from the pirate flags that flew over the Bandley building as the Macintosh was under development. Waiting out the long delay while the stage is set for the final run-through (which would go so badly that a really final run-through would be arranged for the early morning) are Andy Hertzfeld and Bill Atkinson. Andy, wearing a sweatshirt and looking like he's spent too many hours at the keyboard, is feeling philosophical. He says he's been thinking about his former Apple supervisor Bob Belleville. Working in a position of authority at General Magic has changed his perspective, he says. "I'm older now than Belleville was at Mac," he notes. Maybe he's been too harsh, he thinks. Andy has also softened his stance on Steve Capps. "I talked to him a couple of weeks ago," he says. "It went okay. I don't think he was out to hurt us or anything." Capps is going to be at the auditorium tomorrow, because both he and Andy will appear at a Mac retrospective on the stage directly before the General Magic show. One potential participant declined the invite, though: Steve Jobs. "I guess he doesn't want to do anything that seems to diminish what he's doing now," says Andy. He wishes it were otherwise. Bill has counted how many chairs are in a row and how many rows there are. The room holds over three thousand, he calculates.
"Are you nervous?" Andy asks him.
"Not really," he says. "Are you?"
"Yeah, a little."
For someone like Andy, whose default energy borders on nuclear, his heart must be fluttering like a hummingbird. He says he wishes it were the next day, and over already. Bill, who is savoring the moment a little more, assures him that when the demo begins he will be so swept up in it that everything will be fine. A more serious worry is whether the software will survive the demo. Two nights before, during a run-through, everything seemed to work perfectly. But Andy detected a subtle problem, and began tinkering. The next run-through was a mess - the system crashed. Andy had to fix it that morning. On the stage, something goes wrong and the whole setup needs rebooting. Andy leaves to get a bite to eat. Bill drifts over to talk to some other people. It is going to be a long night. But by morning, the future will come into focus, as Andy and Bill once again perform magic: The demo, despite a few bugs (a reluctant microphone, a lousy video pickup of Andy's hand-held unit) wins over the crowd. Of course, the big question is still hanging: Have Andy, Bill, and their fellow magicians matched their boisterous success of a decade past? No way to get an answer today - the ultimate success of General Magic depends on a combination of Alliance management, luck, and the appearance of either a chicken or an egg (I'm still not sure which one comes first). But for now, there's good news: The technology is almost done, and it is the coolest thing in years. Once the bugs are out and the price comes down (in two years? three years? four?) it may well fulfill the cathartic function in our lives that Mark Porat described to Andy and Bill four years ago. No wonder that at the conclusion of the demo, dozens of technophiles, pundits, and horizon-watchers crowd around two smiling former Mac wizards - they've just learned why 1994 may look like 1984.