Marketer's Dream, Engineer's Nightmare
By John Markoff
The New York Times
Cupertino, Calif. -- John Sculley, chief executive of Apple Computer Inc., traveled the world last year to preach about products he predicted would set his company apart from the rest of its industry.
The star of his show was a palm-sized computer that would read handwritten notes jotted on its screen with a special pen, convert them to text, store them for future use, even transmit them wirelessly to other computers. More than that, the computer would anticipate its master's needs like a secretary -- for example, automatically entering into the machine's electronic datebook an appointment time transmitted from someone else. Dubbed the Newton, the new device would arrive, he promised, not in the 21st century but well in time for a place under the Christmas tree in a year or two.
To those who heard Mr. Sculley and watched demonstrations of the Newton, it seemed it was almost ready to ship.
The truth was very different. While Mr. Sculley was proclaiming an era of "Newton Intelligence," the team designing the computer was floundering. For one thing, the computer language on which the Newton was supposed to be based was nowhere near ready; even as Mr. Sculley was touting the Newton, its design team was throwing out four years of Apple research and starting from scratch using another language.
The completion of the Newton, originally scheduled for April 1992, would ultimately be postponed until August 1993. And the computer would be far less ambitious than the one Mr. Sculley was describing.
The pressure to finish, exhilarating at first, eventually overwhelmed some of the young designers. After 18-hour days, some engineers went home and cried. Some quit. One had a breakdown and ended up in jail. One took a pistol and killed himself.
Suddenly people at Apple were calling the team a zombie-like cult. "What's going on over there?" they would ask.
Then, when all seemed lost last December, a half-dozen programmers, led by a software engineer named Steven Capps, worked heroically while their teammates took a Christmas break to turn the project around.
For all their efforts, though, the Newton has so far come up short. It has been ridiculed for its mistakes in translating handwriting and called unworthy of its price, $699 and up. Apple has sold little more than 50,000 Newtons in the four months since the computer hit the market; by contrast, the company's Macintosh computer sold 70,000 machines in just its first three months a decade ago, even though it cost three times what the Newton does.
The Perils of Innovation
What happened? Why has the Newton been a disappointment? The reasons are by no means unique to this product and offer a study in the pitfalls awaiting companies rushing to define a new generation of a product.
Perhaps the main problem was a chief executive who didn't really understand the technology he was championing or how much time it would take to make it work, and who in spite of his efforts to sell himself as a technology visionary never really felt comfortable with the engineers who could have set him straight. As a result, Mr. Sculley promised the computer buyer too much too soon. Furthermore, by trumpeting his coming product so aggressively, he actually encouraged others to work on competing products; some of these competitors were Apple engineers who had left after being alienated.
The story of the creation of the Newton is also a tale of the dedication, personal sacrifice, rivalries and compromise that occur before a radically new product finally makes it to market.
The Newton was born in 1987 of the frustrations of Steve Sakomen, an engineer who saw a future beyond the personal computer as it was then known. It was a future in which people no longer needed a keyboard or a mouse but rather just a pen-like stylus to operate computers. Mr. Sakomen went to Jean-Louis Gassee, who ran Apple's research and development.
The two decided that Mr. Sakomen would lead a small group -- free of Apple's corporate culture -- in rethinking the future of computing. He set up shop in a warehouse on the edge of Apple's campus here, about 50 miles south of San Francisco.
One of Mr. Sakomen's early hires was Mr. Capps. At 38, he is the consummate computer hacker, a whiz at writing elegantly simple programs who also had helped design a key part of the Macintosh software.
Mr. Capps is perpetually dressed in T-shirts, baggy shorts and a pair of checkered slippers. Like most hackers, he lived on the other side of the clock during many months of the Newton project. A normal workday would begin around 3 or 4 in the afternoon and last until first light. This owlish existence allowed him to work free of common interruptions.
Partly because the Newton team was under Mr. Gassee early on, it never achieved the close relationship with Mr. Sculley that Steve Jobs had with the Macintosh programmers when he led Apple.
Communications Gaps - A Chief Executive Who Didn't Relate
Mr. Sculley, who is 55, made an effort to get closer to the long-haired, unshaven programmers and computer designers who are Apple's heart and soul. But at first his visits were rare and, to the Newton team, he seemed aloof and uncomfortable. Never mind that Mr. Sculley was later to anoint himself chief technology officer at Apple. His background was not in technology. He had come to Apple in 1983 from Pepsico, where he was known as a marketing wizard. To the Newton team members, he never felt like one of them.
Many team members remember one painfully awkward afternoon meeting before Christmas 1989. Mr. Sculley, Mr. Sakomen and about 10 young team members sat in a semi-circle, hands clasped in their laps. The atmosphere seemed oddly formal. Finally, searching for some common ground, Mr. Sculley started asking about members of the original Macintosh team. Then, as the conversation ebbed, he abruptly left.
At the end of 1989, Mr. Gassee quit after a power struggle with Mr. Sculley; Mr. Sakomen followed. It seemed Mr. Sculley was on the brink of killing the project and supporting a rival group of Apple designers planning to work on a similar product at a start-up company backed by Apple.
That company, later named General Magic Inc., had grown out of Mr. Sculley's efforts to build industry alliances to develop new technologies. Whereas the Newton project had begun as an effort to design a powerful -- and expensive -- pen-controlled computer, General Magic had been working on a machine with some of the same functions but a lower price.
Wooed by General Magic, Mr. Capps was torn. With General Magic lay the romanticism of working with many of his friends from the old Macintosh team. And a stake in a start-up company that could soon be worth millions of dollars is attractive to engineers, who typically make a salary of $50,000 to just over $100,000.
With the Newton team lay Mr. Capps's feelings of loyalty to the young programmers. And Mr. Capps was uneasy about the star worship Mr. Sculley had helped create around two leaders of General Magic, Bill Atkinson and Andy Hertzfeld.
In early 1990, Mr. Sculley had tried to persuade the team that would become General Magic to remain with Apple and join with the Newton forces, but there was to be no alliance. Without the natural talent of his predecessor, Mr. Jobs, for inspiring an intensely loyal following, Mr. Sculley let his best designers leave for a company that Apple did not totally control and one that would soon obtain additional financing from potential competitors -- A. T. &T., Sony, Motorola and Matsushita.
Mr. Capps finally decided not to let the Newton die. In April 1990, using a Macintosh screen jerry-rigged to look like a notepad-sized computer, he demonstrated to Mr. Sculley, who had been trained as an architect, how the new computer could instantly transform a rough architectural sketch into a finished plan.
Mr. Sculley was captivated. "This is great stuff, you guys," he said. "Let's go for it." Mr. Sculley asked Larry Tesler, 48, head of Apple's Advanced Technology Group, to take Mr. Sakomen's former position as head of the Newton team. The team would grow to about 60 from 30.
But Mr. Sculley, under pressure from the fierce personal computer price wars and having a hard time differentiating Apple's products from other personal computers, warned that he wasn't authorizing a five-year research project. He gave the team two years to deliver a finished product. The deadline: April 2, 1992.
Conflicts - Easier Prophesied Than Produced
The Newton was saved. But the real trouble was only beginning. Computer development, which involves writing millions of lines of instructions that will breathe life into hardware circuitry, is never predictable. Within months of agreeing to Mr. Sculley's deadline, some Newton designers began to realize that a two-year timetable was not realistic.
Many kinds of Newtons had been considered, from pocket-sized computers to wall-sized computer drawing boards. The original vision of Mr. Sakomen and Mr. Gassee had been an $8,000 machine that was to be more powerful and, at about the size of a clipboard, bigger than the current Newton. That was later scaled back to a $4,000 machine.
Seeking a more reachable goal, Mr. Capps, with Michael Tchao, the Newton product manager, and Michael Culbert, a hardware designer, began plotting to radically redefine the Newton. Mr. Tesler resisted, so in order to make it fly the group had to get Mr. Sculley's blessing.
Out of a meeting between Mr. Tchao and Mr. Sculley aboard a corporate jet in February 1991 came a basic redefinition of the idea of the Newton. It would now be small enough to fit in the palm of the hand and it would cost less than $1,000. It would no longer have the full range of wireless-communications capabilities previously planned. But it would still convert handwriting to text, and play secretary by organizing, storing and retrieving that information.
At the time the goal seemed attainable. "We thought we could pull pieces out of the work that had already been done and then put them in the smaller box," Mr. Tchao said. "We learned it wasn't going to be that easy the hard way."
For one thing, starting over turned out to mean starting almost from scratch. For another, when the programmers at General Magic learned of the new concept of Newton they were outraged: their new product and the Newton would now be very much alike.
Mr. Sculley gently pushed the General Magic developers toward the second of their two technologies, a futuristic language to link computers. But in a May 1991 meeting, Marc Porat, a co-founder of General Magic, confronted Mr. Sculley. "You've turned on us," he said.
Relations between the two camps of developers, never warm, dove below freezing. Apple's failure to bring the two groups together left the company with a nightmare that could well become more frightening with time. Next year Motorola will introduce a palm-sized computer that will run software written at General Magic. Many think it will be the Newton's most serious competitor. A Sony machine using General Magic technology is also in the works.
Mr. Sculley first told the world about the Newton at the Consumer Electronics Show in Las Vegas in January 1992. True, it was little more than a rough prototype. But Mr. Sculley needed something fast. And it's easier to describe the future than to invent it.
In May 1992 Apple demonstrated the Newton for the first time, at the Consumer Electronics Show in Chicago. But Mr. Capps's demonstration made the machine look a lot slicker than it was; while it could do a few fancy tricks, it was still laden with bugs that made it unusable.
Mr. Sculley's design team was nervous. Mr. Capps and Donna Auguste, 35, who was managing the software side of the team, confronted their team leader, Mr. Tesler.
Mr. Capps and Mr. Tesler had never seen eye to eye. For one thing, Mr. Tesler's computer scientist's approach -- precise and systematic -- conflicted with Mr. Capps's informal, seat-of-the-pants hacker's style.
But they also disagreed on a fundamental component of the Newton. Mr. Tesler had been drawn to the project in part because the team creating the Newton's software was to use an advanced programming language called Dylan. When perfected, such a language would allow software designers to work much more quickly and to build an artificial-intelligence capability into the Newton.
Now, on that May day in Chicago, Mr. Capps and Ms. Auguste told Mr. Tesler that the team that had been developing Dylan at an Apple lab in Boston for four years was way behind schedule. Without paint brushes, Apple's designers couldn't paint. Throw out Dylan, they said.
Mr. Tesler admitted they were right. But giving up on Dylan meant starting over again, with a more conventional language. Newton was scaled back to what the team nicknamed "Newt," a computer without many of the planned features, but one that could be finished, albeit not until summer 1993, more than a year late.
The Toll - A Source of Pride But Also Despair
Meeting that new deadline, even using proven programming languages, would be difficult. The pressure was turned up.
In Silicon Valley, working long days on a breakthrough product is a source of immense pride. And the young members of the Newton team comprised the valley's best and brightest. But some on the Newton team found the pressure and the almost cultish group commitment that went with it intolerable.
Jerome Coonen, a math wizard and a Macintosh designer, left the Newton team in 1991. He had always hated what he called group-think.
After his second child was born he decided that being a full-time father would be just as rewarding as being a double-time Newton developer. "I looked at a baby being born and then at where this project was headed and where my career at Apple was headed and the whole thing seemed absurd," he said. Mr. Coonen left. He is still out on his own, as a consultant.
Several others also walked away. But others recommitted themselves.
Mr. Capps, who had done it all before when working on the Macintosh, found that he still had an amazing capacity for work. He converted his house into an office and worked there with two other Apple engineers, often around the clock.
Realizing that making Newton a reality would require a leader, Mr. Capps began to assume the role of a Steve Jobs. But while Mr. Jobs often drove employees by shouting, Mr. Capps inspired them with hard work.
He also tried to provide the vision. "People need a leader," he said. "Sometimes it doesn't matter in which direction you point as long as someone is pointing."
Mr. Tchao, who is 30, realized he was at the breaking point. Some nights he would get home to his Palo Alto cottage around midnight after 16 hours of nonstop meetings, and he would cry. "I remember there was this point where I said to myself, 'I'm not going to feel guilty, I just can't do it.' " He pulled back to a more normal schedule.
For others on the Newton team there was no such resolution.
Ko Isono, a 30-year-old programmer from Japan, married against his parents' wishes and brought his Japanese bride to live in Fremont, a suburb east of Silicon Valley. Like everyone on the team he was working incredibly long hours, and his wife was soon miserable, stuck all day and all night in a drab house in a strange land.
Mr. Isono was writing the portion of the Newton software that controlled how text and graphics were displayed on the screen. As the deadline closed in, he began to feel trapped.
Last Dec. 12, Mr. Isono came home, went into the bedroom and shot himself.
The suicide almost unhinged the Newton team. To make matters worse, within a week another Newton programmer had a breakdown. He attacked his roommate and ended up in jail on an assault charge.
Then came a turning point. Most of the team was given Christmas week off to decompress. A few remained behind. With Mr. Capps and Ms. Auguste managing, a half dozen programmers got a huge amount done.
As 1993 arrived, the difficult task of managing outside expectations about the team's progress fell to Mr. Tchao. Sometimes it meant equivocating. He couldn't tell the designers that they were far behind -- too discouraging. He had to tell the marketing people that progress was even slower than it really was -- he didn't want them to get too far ahead. To Gaston Bastaiens, the marketing executive who was his boss, he had to sound promising.
In the final months of development, the team leader, Mr. Tesler, rebelled. Refusing to compromise the product for the sake of a deadline, he would do things politically risky, like criticizing Mr. Bastaiens. Mr. Tesler was forced out. He was sent back to Apple's research group.
Compromises - It's Not Everything But It's Something
When the Newton was rolled out last August, the event proved an anticlimax. The product had already received a great deal of attention, and many sexy features, like wireless communications and some artificial intelligence, were not in the machine.
By contrast, the Macintosh introduction in 1984 had been a sensation. Apple had done everything it could to hide the machine. When it finally reached the market, many people, including competitors, were surprised by how many features it had.
Today, four months after the Newton's introduction, it is not clear whether the computer will succeed or fail. After a burst in the first two months, Newton sales have slowed to a trickle, Apple insiders say.
Yet not everyone thinks the Newton is a disaster. After an early crop of negative reviews, two widely read computer analysts, Richard Shaffer of Technologic Partners and Mark Stahlman of New Media Associates, have weighed in with more positive assessments. And the product will be improved, both from inside and outside Apple. Just last week, 1,000 outside software designers showed up at the first meeting intended to create new applications for the Newton.
The Future - As Boss Departs, Team Carries On
Whatever the final verdict, the Newton project contributed to Mr. Sculley's fall from atop the $8 billion company. Intent on pushing toward his vision, Mr. Sculley neglected his rear guard, the Macintosh line.
Mr. Sculley recently said he was uncertain whether his early, aggressive marketing campaign had been the right strategy. "You could build a case either way," he said. But he said the campaign was a means of inspiring the Newton team as well as a way to build outside interest. "We felt if we didn't have a stake in the ground, support would be difficult to generate inside Apple. And we were afraid of losing the Newton team."
Others say that one of Mr. Sculley's most glaring errors was failing to realize that his vision was outrunning the Newton team. "If you put a vision too far ahead, other competitors will step in and pull the market away from you, and toward producing a more coherent product," said James Moore, president of Geopartners, a Cambridge, Mass., consulting firm. Indeed, Tandy, Casio and Eo have introduced pen-based computers since Mr. Sculley's road shows.
Mr. Sculley has since left Apple to head a tiny company, Spectrum Information Technologies Inc., that offered him a big pay package.
Mr. Capps, by contrast, still labors for the Newton team. The team is trying to add many of the features missing from the first Newton, like better handwriting-recognition software, a brighter display screen, longer-life batteries and new wireless-communication features. The team is also working on new kinds of Newtons, like ones that can be incorporated in a desk telephone.
Mr. Tchao plans to work on Newton communications technologies. Ms. Auguste recently left the company for U.S. West. Mr. Tesler has become Apple's chief scientist.
As for the late Mr. Isono, the Newton now includes a dedication. Enter a special command and the screen says: "In Memory of Ko Isono."
GRAPHIC: Photos: Steven Capps and the Newton computer he helped build, at home above Silicon Valley. (Ed Kashi for The New York Times)(pg. 1); Michael Tchao, the manager who helped define Newton. Jerome Coonen, software engineer, left to spend more time with Jeffrey, 2, and Lucas, 5. (Photographs by Ed Kashi for The New York Times); Larry Tesler, team leader, wanted something better. John Sculley, the executive who promised too much. (Associated Press)(pg. 8)
Copyright 1993 The New York Times Company