Behind Apple Computer's long-awaited Newton lie six years of painful and sometimes contentious work on a project whose originator never imagined it would be so crucial to the company's future -- and a product that Apple nearly didn't make at all.
This "personal digital assistant" began as a tiny "pirate" project hatched by one engineer who at first planned to leave Apple and start his own company to develop it. Only a clash of personalities among his likely partners kept the project at Apple, where it was nearly killed by skeptical brass, partly resurrected by a handful of Soviet programmers and had its core software entirely gutted and redesigned in a furious last-minute scramble over the past year.
The result, after at least three major retoolings of the original concept, is a pocket-size, battery-powered device that can read handwriting, compose and send faxes, receive pages, keep track of a user's appointments or clean up a rough sketch into a finished drawing.
And it's clear that those who have toiled for years to get Newton to market feel the pressure of delivering a product that many call the key to Apple's future.
"Everyone's greatest fear is that we've put too much weight on the Newton too soon," said Michael Tchao, 30, an Apple marketing whiz who joined the Newton team in 1990 and who helped convince project leaders the first Newton should be pocket-size.
But the engineer who first conceived of Newton never dreamed it would eventually become interwoven with a host of corporate partnerships and break Apple's historic pattern of keeping its prized technologies bottled up inside the company.
Steve Sackoman came from Hewlett-Packard to Apple in 1984 during the throes of Macintosh euphoria and quickly found himself in charge of Macintosh hardware development. For three years he oversaw follow-up products to the original Macintosh personal computer. By 1987, he had helped start five such products, and in the process became increasingly frustrated at dealing with the company's convoluted internal politics.
"You spent a lot of time spinning your wheels. Everybody's got their piece of turf to be defended," he said. "I decided the prospect of turning out Mac clones wasn't what I wanted to do with my life."
So in the spring of 1987, Mr. Sackoman told his boss, Jean-Louis Gassee, Apple products chief, that he wanted to quit and start his own company. Mr. Sackoman wanted to design a new kind of computer, one that had communications at its heart and that didn't have to conform to any established rules of personal computer design.
"The idea was to see if the personal computer could be rethought without carrying around a lot of baggage," he said. He envisioned a wide range of devices, from pocket-size to replacements for wall-size white boards. The one he settled on was a go-anywhere model. Users could jot down their thoughts on its screen with a pen and communicate with other such devices without a complex network setup.
The idea was so appealing to Mr. Gassee that he agreed to leave with Mr. Sackoman to start the venture. They soon found themselves on a plane to Boston, where they were to meet with three other men who had expressed keen interest in the idea. One of them was Mitchell D. Kapor, who had made millions as the inventor of the Lotus 1-2-3 spreadsheet. Another was Mr. Kapor's friend and software guru, S. Jerrold Kaplan, whose specialty was artificial intelligence.
In a room at the Marriott Hotel, near Lotus' Cambridge, Mass., headquarters, the five men discussed the idea. It looked for several weeks as if the company would get started, but it soon ran into snags. The largest problem was personality: Messers. Gassee, Kapor and Kaplan, all strong-willed, realized "they would butt heads" in the boardroom, Mr. Sackoman said.
When the deal fell through, Mr. Gassee persuaded a reluctant Mr. Sackoman to let him sell then-Chairman and Chief Executive Officer John Sculley on sponsoring Newton as an internal research project. Mr. Sackoman agreed but made a highly unusual set of demands: Newton would have to be physically separate from the rest of Apple; Mr. Sackoman was to have complete control over the environment of the venture; and Apple couldn't force Newton to be compatible with anything that had come before it.
Surprisingly, Mr. Sculley agreed, and in July 1987 Mr. Sackoman began assembling a team of engineers and programmers -- no marketing types were invited -- and moved into a converted warehouse in the middle of Apple's main campus in Cupertino, )) Calif.
The group went to work, and over the next two years the team of engineers gradually grew to a mere two dozen, including such Apple veterans as Steve Capps, who remains one of the chief programmers on to day's Newton project, and Mike Culbert, a hardware designer who also remains a key member of the project. The pirate team developed its first custom chips and by the end of 1989 assembled a plug-in Macintosh circuit board that could simulate some of the Newton's operations.
The Newton that Mr. Sackoman's group had incorporated two expensive AT&T; microprocessors and had the processing power a Sun Microsystems work station. It was slated to have a large and costly touch-sensitive LCD screen, and at least three built-in networking devices. It weighed five pounds, and given Apple's huge profit margins at the time, would have sold for between $5,000 and $8,000.
"There were no constraints, and one of the things that got lost early was cost," Mr. Culbert said.
At that point, Apple's board "didn't want to continue funding the project," said Lawrence G. Tesler, who was then in charge of Apple's advanced technology group (ATG) and soon to become the head of the Newton effort. The company brass wanted to put the money elsewhere because few of them believed Newton ever would work.
Sackoman, Gassee join forces
Mr. Sackoman, who had an agreement with the board that let the body kill the project but not change any of its designs, read the handwriting on the wall and in March 1990 resigned. A few months later, Mr. Gassee was purged and the two men joined forces to start Be Labs, a San Jose start-up working on a personal digital assistant of its own.
With Mr. Sackoman gone, it appeared as if Apple's three-year experiment would come to an inglorious end. Mr. Sculley told Mr. Tesler to scour the Newton project for salvageable technology and to get rid of the rest.
But Mr. Tesler asked Mr. Sculley first to let him see if he could come up with a way to make the Newton smaller, lighter and cheaper.
"I challenged the group to cut the cost in half," Mr. Tesler said.
Mr. Tesler said the engineers managed "within a couple of weeks" to come up with the rough outline of a modified Newton that Mr. Tesler figured could be sold for about $2,000.
"He gets a lot of credit for saving it from being broken up," Mr. Tchao said, noting that Mr. Tesler took a big career risk at the time to "reset" the Newton project.
One reason: Newton still had a big doubter in Albert Eisenstat, executive vice president and corporate secretary. He was convinced that near-perfect handwriting recognition was crucial to public acceptance of Newton, and equally convinced that neither the Newton group nor a one-man handwriting research effort in ATG was up to the task.
The Russians are coming
That's when the Russians came to the rescue.
A group of programmers in Moscow had written to an Apple engineer they knew, describing some handwriting recognition software they had developed for reading the addresses on envelopes and inquiring if Apple wanted to look at it.
A few months later in 1990, when Mr. Sculley and Mr. Eisenstat were getting ready for a trip to meet with top officials in the government of Mikhail Gorbachev, Mr. Tesler suggested they look up the programmers. They did, and although their software wasn't all that good, a Russian-speaking Apple engineer along on the trip concluded the ideas they had were sound.
Apple agreed to hook up with the programmers, who called themselves Paragraph, and Mr. Eisenstat dropped his major objection. Eventually, the Paragraph software became the heart of Newton's handwriting recognition system.
That still left the Newton group with a problem: They had plenty of technology but no real idea what consumers thought of it or who would buy it. So the group hired its first marketing executives -- forbidden under Mr. Sackoman's regime -- including Mr. Tchao.
With Mr. Sculley prodding the group to make a smaller device, Mr. Tesler and the engineers started working on one more the size of a steno pad than a standard notebook.
The group also began testing its idea with potential users. At first, all they had to show prospective users was a thick plastic housing the size of a sheet of paper, with weights inside to simulate bulk and a children's plastic-film drawing tablet for a screen. They asked people to use their imaginations more than any fancy technology.
7+ Users were fascinated with the concept.
But all of them said they wanted it even smaller.
Mr. Sculley shared the notion that Newton should fit in a pocket. So did Mr. Tchao and Mr. Capps, who got permission to start a kind of pirate effort of their own to investigate such a device. They began toying with electronic personal organizers from Japan, like the Sharp Wizard, and early pen computers like Sony's Palmtop.
But for the next year and a half, most of the group toiled on the steno-sized unit, while Mr. Tchao, Mr. Capps and a few others kept working on a smaller model that used almost entirely different software.
The group realized that, whichever device arrived first, the real business for the future wasn't in making and selling the hardware. Bucking the company's long-standing policy not to license its key technologies, the Newton group developed a plan to let other companies make their own versions of the Newton.
That, they reasoned, would quickly create a sizable market for a vast new array of businesses. Apple hopes to sell on-line information, communications and electronic mail services to Newton buyers and has started a publishing arm to produce information such as maps, encyclopedias and even games for the Newton.
And it has, for the first time in its history, licensed core software and design technology so other companies can manufacture their own versions of Newton.
It fell to Mr. Sculley, now the company's chief technical officer and head deal-maker, to enlist Sharp as the first licensee and as the manufacturer of the product. In early 1992, he began telling the outside world about Apple's concept of PDAs and the immense market he felt they would create. At the Consumer Electronics show in Las Vegas in January 1992, Mr. Sculley talked about PDAs; four months later, he decided to show Newton to the world at the Chicago version of the same show.
But, much to the consternation of the engineers, Mr. Sculley decided to take not the larger machine most of them had worked on, but the pocket-size model. And he brought Mr. Tchao to demonstrate it to the crowd.
The public reaction was instant. So was that of the engineers. Suddenly, after nearly three years, the group had to switch priorities to the pocket-size machine -- and they realized that they needed to rewrite nearly all of the Newton's core software.
Programmers and engineers took to working around the clock at Apple's new research and development center, living on the "Newton diet" of early-morning doughnuts and late-night pizzas.
How close did Apple cut it? The final programming was ready only in early July; there are some reports that last-minute fixes are being loaded at the factory. And the first units rolled off the assembly line just a week before its gala rollout in Boston, sending workers scrambling to check them and ship them in time.
So after six years, consumers finally get to try out what has eventually arisen from Mr. Sackoman's original idea, a product that must not only meet high marketplace expectations if it is to succeed but also win over Apple insiders as well.
"I think there are still some skeptics at Apple," Mr. Tesler said. "But a little skepticism is a healthy thing."