One by one Paul’s older brothers and sisters moved away from home, but Paul stayed on in the converted garage for all the seven years he studied at the university. He had enough money to rent his own place, and he certainly wasn’t lazy. He had spent part of the summer before college as a meter reader for Boston Gas. He loved the job, perhaps too much. On one of his first days, he read eight hundred meters. In union parlance, he did “two books,” and the union boss took him aside and said, “From now on, you’ll do one book.” After a while, he was put on “special reads,” which meant he had to find his way inside places where the meters hadn’t been read in six months. Sometimes he had to break in. The job had excitements. It was always possible that he would find someone inside with a gun. It was fascinating, too, like a tour of the innards of the city, all sides of it, from townhouses on Beacon Hill furnished in mahogany to restaurant basements in Chinatown and Southie haunted by enormous rats.
He left that job reluctantly. He went on coding part-time for his brother during most of freshman year while also playing in the jazz band. He loved the teamwork, the call and response of playing in a band, and especially the complex task of arranging music, which felt akin to the programming he was doing in his classes, indeed much more like computer programming than math had ever seemed.
He took most of his classes at night, and as many in computer science as he was allowed. During the days, he worked full-time, at a string of programming jobs seven years long. Soon enough he realized that, in spite of first impressions, the computer science faculty at lowly UMass Boston deserved his respect. Some were still in the midst of distinguished careers, and most had practical experience at companies. Unlike many computer science programs, theirs focused not on theory but on the actual engineering of software. Paul’s jobs added up to a supplemental course, like clinical rotations in medical school.
It was the 1980s, and computer technologies were still a source of wonder and anxiety to the American public, but they no longer belonged exclusively to specialists in white coats ministering to huge mainframe machines in sealed-off rooms. The invention of the microchip—the innards of a computer etched onto tiny slices of silicon—had made possible a great variety of machines and uses for them. Other devices had all but replaced centralized computer systems like the one Paul had broken into at Boston Latin. There were supercomputers, minicomputers, workstations, and personal computers, and all that machinery required the services of software engineers. During the 1980s, their numbers grew annually by more than 25 percent. Computers were spreading even into the humble corners of American life, into car repair shops and living rooms and grade schools, as well as into labs and government buildings and corporate skyscrapers. The computer had—all of a sudden, it seemed—become America’s essential tool, favorite new toy, and universal scapegoat. Everywhere in the land, one now heard the phrase “Sorry, the computer’s down.”
To many parents of kids in junior high and elders in business and education, it seemed as if the new technologies had arrived with a human user interface included, with a generation genetically coded to understand how computers worked. In his first programming job, Paul was hired as a temp doing menial chores, but ended up rewriting the company’s administrative software. He had wanted to learn what office work was like. After about six months, he quit and signed on at a medical device company, where he wrote a program to control a centrifuge. He also spent a year and a half with a defense contractor, writing software to control a high-speed camera in an air force spy plane and a user manual to go with it; and another year and a half at the minicomputer company Data General, where he performed coding jobs for a team of computer scientists with PhDs.
In 1987, Paul finished that practicum, and he also got his bachelor’s degree. He was twenty-four and ready, he felt, for bigger things. His favorite professor at UMass was a computer scientist with a doctorate named Bob Morris. He worked part-time for a large and growing company called Interleaf. It made software for the computerized creation of documents, and it had within its ranks, Paul gathered, a small group of extraordinary programmers. Paul asked Morris if he could join them. Morris said Paul wasn’t ready yet. “All right, whatever,” said Paul, and he went to work for one of Interleaf’s small competitors. Like the great majority of the fourteen thousand software companies founded in the 1980s, this one was doomed to roughly the same life span as a salmon’s, but it lasted long enough for Paul to learn some of the new art behind electronic publishing. Meanwhile, he went on studying at UMass. The PhDs he had met at Data General inspired him. He wanted an advanced degree, too.
By the time he started graduate classes, Paul knew his way around the society of programmers. He felt he was one of them, and yet he sometimes felt like an anthropologist in their midst. He knew coders so unworldly that they needed a mnemonic to guide them at a restaurant table. They’d murmur “BMW” to remind themselves that it was their Bread that lay on the small plate to the left of their Meal, their Water in the glass to the right. He could navigate the wider world better than many programmers, and he felt comfortable among them. He felt that all of them were a bit peculiar: “We’re all introverts, we’re all nerds, we’re all slightly awkward.” And he could more than hold his own technically, which was what really mattered in the society of coders.
At UMass Boston, Computer Science consisted mainly of two rooms, aboveground but subterranean in feeling, a little windowless world of concrete floors and concrete-block walls painted industrial white. There was a lab stuffed with equipment and next door a fairly large room filled with tables and chairs. This was where grad students hung out with computers and each other.
Paul’s main friend there was a thin young man with long blond hair named Karl Berry. Paul could usually find him sitting in front of one of the new Sun-3 workstation computers. The machines were named “red,” “blu,” and “grn.” In this society, abbreviation was a useful habit, and naming machines was one of the things that some coders thought of as “a fun event.” Karl struck Paul at first as one of those programmers who had trouble looking other people in the eye, a difficulty that Paul had tried to overcome in himself. Karl had a deep, strong voice, though, and strong opinions. He could be curt and impatient when discussing technical matters, but he struck Paul as very smart and principled and, once you got to know him, downright friendly. Karl was also a very good programmer, one of the two or three best who hung out in that room. Maybe in another setting Paul would have decided to make him a rival, but probably not. Paul never had the feeling that Karl wanted to compete with him. Karl felt more like a colleague, like a fellow member of a band that played the rarefied music of computer code. Karl was the first member of Paul’s band, as it turned out, a charter member of his twenty-pluses.
They were nearly the same age. Both had their first encounter with computers in seventh grade, though Karl’s early romance was more innocent and chaperoned. His family was living in Palo Alto for a year, one of the few places in the late 1970s where a junior high school student was apt to find a formal class in programming. Learning to code had felt immediately comfortable to Karl, as it had to Paul, like developing an inborn trait. You wrote the program just so, and in no time you saw the correct answer appear, an answer worked out not by you but by the machine under your command. This was powerful stuff, magical in the sense of Arthur C. Clarke’s third law of prediction: “Any sufficiently advanced technology is indistinguishable from magic.”
After Palo Alto, Karl’s family had returned to Canton, in upstate New York, the home of St. Lawrence University, where his mother taught music. On holidays and college vacations, Karl would use a plastic card to jimmy open the door to the university’s computer lab, where he spent hours reading manuals and trying out what he learned. What he later called the social niceties didn’t come easily to him, but that didn’t matter with computers. Exploring them didn’t require that he deal with other people and their complexities. Writing software seemed to occupy a place between pure mat
hematics and the experimental sciences, between the search for a platonic ideal and the search for theories that jibed with observations of nature. He referred to writing software as “this funny business.” It was the creation of something “notional/fictional/imaginary” that nonetheless had tangible effects. Coding was infinitely precise and yet imperfectible. Programs got written and revised until they did their jobs as well as possible. It was, Karl thought, a field for people who liked “the idea of good enough instead of perfect, the idea of play around in your head instead of play around in the mud.” If your program failed, it was because you hadn’t tried hard enough, not because information was withheld or missing. Programming computers felt, Karl would say, like entering “a universe unto itself.”
He had been a model student in high school, especially compared to Paul, and went on to Dartmouth to study the almost brand-new field of computer science. One day in 1983, his professor handed him a tape that contained some new software, saying, “You’re a bright guy, Karl. Go and install this stuff.” The tape contained Donald Knuth’s typesetting system TEX and the handbook Knuth had written for it, The TEXbook. It began: “GENTLE READER: This is a handbook about TEX, a new typesetting system intended for the creation of beautiful books—and especially for books that contain a lot of mathematics.”
Karl loved it, the clarity of Knuth’s prose style and the central idea behind the TEX software: that via a computer one could create art. And you could use TEX free of charge and change it however you wanted, so long as you didn’t call your modified version TEX. This wasn’t very unusual then. As Karl would put it, “Universities were still about sharing knowledge (instead of cashing in on it).” Karl would also say that he had his best luck socially through the computer. He had met his first girlfriend on one of the early online news and discussion groups. She was a typographer and a TEX enthusiast.
In the 1960s, IBM created a complex operating system called DOS and all but gave it away—to Microsoft, then a small company. In the years since, software had become an increasingly valuable commodity, and companies were turning the programs their employees wrote into private estates, surrounded by walls of secrecy, restrictive copyright licenses, and eventually patents. Meanwhile, a movement of programmers had begun to grow up in opposition. It ultimately became what amounted to a large, informal commune of great significance, which advanced the state of the art, kept the Internet and World Wide Web free for all to use, and produced those great libraries of code that anyone could borrow from, including entrepreneurs.
The progenitor of this movement was a legendary programmer named Richard Stallman. He founded something called the Free Software Foundation. Karl came across it in its early days, and was drawn to it at once. “Share and share alike”: that was how Karl summed up the philosophy. He argued the case with Paul sometimes, down in the computer room.
“I won’t invest a lot of R&D effort in my company if someone can just rip me off by copying all my ideas,” Paul wrote in one email to Karl, arguing for limited software patent protection. In another, he wrote: “I don’t like the idea of throwing out the baby with the bath water, killing software patents all together.”
Karl replied that it would be far better to have no software patents at all than the thousands upon thousands of patents that claimed ownership of software that by rights, and even by statute, should have belonged to everyone. Many patents were so broad that they encompassed elements fundamental to programming, elements long in the public domain. And there were so many of those kinds of patents that programmers often had no way of knowing which ones they might be violating as they wrote their code.
The argument widened at times. Paul wrote, for instance: “Capitalism American-style says that a million people all acting selfishly and *without regard for others* somehow create something reasonable.”
“I disagree that that is the case,” replied Karl.
In Karl, Paul had met someone who stood for an ideal. Paul was an idealist, too, but also a pragmatist, and a lot less innocent than Karl. It would have been hard to imagine Karl dealing pot or shooting bottle rockets at tailgating cars. At one point Karl wrote, on the wisdom of abolishing software patents: “I hope I can convince you. If I can’t convince you, I despair.”
Paul hung out with Karl and his girlfriend—she was a fellow master’s student—on many evenings in the main computer science room at UMass. To Karl, it seemed as though a lot of the other students there were learning this trade just to land jobs with good pay. He and Paul didn’t talk about that sort of thing. They traded thoughts on the craft of programming and on their other enthusiasms, which they were combining with programming. Paul just then was very interested in writing synthesizers, code that could turn a computer into a musical instrument. Karl and his girlfriend were fascinated with typesetting and the grand suite of software that Knuth had created for digital typography.
At the end of the master’s program, students were divided into teams of four or five, each team to work cooperatively on creating a substantial program. Paul and Karl and his girlfriend asked that they be placed on the same team, but the professor demurred. They were his best students, and he wanted them spread around. The course was supposed to teach a process. The product didn’t really matter. But it wasn’t very long before Paul went to the professor and told him that his teammates didn’t know what they were doing. Why didn’t he let Paul do the project by himself? He could do it in half the time the team would take. “You don’t understand what this course is about,” the professor answered. “I could do the project in half the time, but I’m not going to, and neither are you. What you are going to do is teach your team and get the project done on time.” Paul complied, happily enough.
Afterward, he and Karl both stayed on in Boston, but they went separate ways. Karl and his girlfriend got jobs at the Free Software Foundation, where they went to work on creating a digital family of fonts, a very challenging task in that era. Paul went to his professor Bob Morris and asked him again if he could have a job at Interleaf. This time Morris said, “Okay. You’re ready now.”
4
Boston’s social structure had accreted over four centuries. By tradition, certain routes to success were blocked for someone like Paul. In one world, people pronounced “wheelbarrow” as if the word were “wheelbarrel,” and like Paul got married in a Polish Catholic church in Dorchester. And in the other, girls learned excellent posture and grew up saying “toe-mah-toe,” not “tamayta,” and a lad went to country day school, then prep school, and then, if his grades were merely decent, Harvard, where he would be examined during “punching season” and invited to join a final club—the invitation guaranteed if it had been his father’s club and he was a “legacy.” A family address in Louisburg Square, a brass plate on the door, a father who took you to Brooks Brothers or maybe his tailor to buy your first suit. Money lay behind the whole construction, of course, but when it came to membership, manners mattered more than money. If you didn’t have a trust fund awaiting you at twenty-one, and even if your parents had fallen to shabby gentility, you could still land a job at an institution like the State Street Bank and Trust, and work among mahogany and the models of the sorts of sailing ships that had founded the old fortunes still being managed there. You’d be hired in the certainty that your boss could take you out to lunch at the Somerset Club and wouldn’t have to ask the steward to outfit you with a proper jacket and tie. You would carry in your accent the assurance that you knew the difference between the fish fork and the salad fork, and of course you would never have to murmur “BMW.”
One could feel excluded from that world even after being admitted to one of its favored institutions. In his autobiography, Tom White wrote about going to college in the 1930s: “I did not like attending Harvard—especially the first two years when I was a commuter. I had a poor self-image to start with and felt inferior to some of the good-looking, self-assured wealthy young guys. There was a separate house for commuters on a side street o
ff Harvard Square. Commuters used it to study, have a meal or a snack, or just rest for a while. I even hated going in there. I just hated the fact that I was a townie I guess.” In later life, Tom made the acquaintance of a scion of one of Boston’s founding elite, Alexander Forbes. “Sandy was class of 1932 at Harvard, handsome guy, a real Brahmin, just like the Cabots and the Lowells, spoke only to God,” Tom wrote. The last time Tom ran into Forbes, the man was pushing ninety. He said to Tom, “Things change as you go through life, and we don’t feel about you people the way we used to.” You people. By then, Tom wrote, he felt only amusement.
One can overstate the exclusionary power exercised by Boston’s Anglo-Saxon elite. There had long been ways for a young Irish Catholic to rise. Tom had found one, in construction. “In my business,” he liked to say, “all you had to be was low bidder.” There was politics, of course. And thanks to recent economic history, there was another and broadening route up, the technological exception. You could have the wrong accent and no table manners, and be possessed by psychological oddities or worse, so long as you belonged among Knuth’s 2 percent, born to program computers.
Enthusiasts imagined that digital technologies would serve as a great democratizing, equalizing force. At worst this was sheer fantasy, and at best just one of many distant possibilities in the kaleidoscope of the future. But there was one social revolution that computers certainly performed. Out on the highways that half encircled Boston—Route 128 and I-495—great clutters of new computer-related companies were arising. Most of the buildings had all the grace and style of grids on graph paper—quickly built, inexpensive, functional. The founder might be an immigrant’s son with a knack for engineering, his office suite decked out in wallpaper that vaguely looked like wood paneling. Downstairs, in cubicles and windowless labs, engineers were producing stuff that was in turn producing immense new fortunes. For those in the business of making or selling it, computer technology really was a new way to rise. Merit counted more in the business of computer programming than in most professions. And for boys, at least, the road from software engineer to software entrepreneur, from wage earner to company founder, was already well marked out, there for the taking when Paul arrived at Interleaf in 1989.