The Innovators

by Walter Isaacson

  Whatever the legalities were, Jobs had a right to be angry. Apple had been more innovative, imaginative, elegant in execution, and brilliant in design. Microsoft’s GUI was shoddy, with tiled windows that could not overlap with each other and graphics that looked like they had been designed by drunkards in a Siberian basement.

  Nevertheless, Windows eventually clawed its way to dominance, not because its design was better but because its business model was better. The market share commanded by Microsoft Windows reached 80 percent by 1990 and kept rising, to 95 percent by 2000. For Jobs, Microsoft’s success represented an aesthetic flaw in the way the universe worked. “The only problem with Microsoft is they just have no taste, they have absolutely no taste,” he later said. “I don’t mean that in a small way. I mean that in a big way, in the sense that they don’t think of original ideas and they don’t bring much culture into their product.”116

  The primary reason for Microsoft’s success was that it was willing and eager to license its operating system to any hardware maker. Apple, by contrast, opted for an integrated approach. Its hardware came only with its software and vice versa. Jobs was an artist, a perfectionist, and thus a control freak who wanted to be in charge of the user experience from beginning to end. Apple’s approach led to more beautiful products, a higher profit margin, and a more sublime user experience. Microsoft’s approach led to a wider choice of hardware. It also turned out to be a better path for gaining market share.

  RICHARD STALLMAN, LINUS TORVALDS, AND THE FREE AND OPEN-SOURCE SOFTWARE MOVEMENTS

  In late 1983, just as Jobs was preparing to unveil the Macintosh and Gates was announcing Windows, another approach to the creation of software emerged. It was pushed by one of the diehard denizens of the MIT Artificial Intelligence Lab and Tech Model Railroad Club, Richard Stallman, a truth-possessed hacker with the looks of an Old Testament prophet. With even greater moral fervor than the Homebrew Computer Club members who copied tapes of Microsoft BASIC, Stallman believed that software should be collaboratively created and freely shared.117

  At first glance, this did not seem like an approach that would provide incentives for people to produce great software. The joy of free sharing wasn’t what motivated Gates, Jobs, and Bricklin. But because there was a collaborative and communitarian ethic that permeated hacker culture, the free and open-source software movements ended up being powerful forces.

  Born in 1953, Richard Stallman was intensely interested in math as a child growing up in Manhattan, and he conquered calculus on his own as a young boy. “Mathematics has something in common with poetry,” he later said. “It’s made out of these true relationships, true steps, true deductions, so it has this beauty about it.” Unlike his classmates, he was deeply averse to competition. When his high school teacher divided the students into two teams for a quiz contest, Stallman refused to answer any questions. “I resisted the notion of competing,” he explained. “I saw that I was being manipulated and my classmates were falling prey to this manipulation. They all wanted to beat the other people, who were just as much their friends as were the people on their own team. They started demanding that I answer the questions so we could win. But I resisted the pressure because I had no preference for one team or the other.”118

  Stallman went to Harvard, where he became a legend even among the math wizards, and during the summers and after he graduated he worked at the MIT Artificial Intelligence Lab, two subway stops away in Cambridge. There he added to the train track layout at the Tech Model Railroad Club, wrote a PDP-11 simulator to run on the PDP-10, and grew enamored with the collaborative culture. “I became part of a software-sharing community that had existed for many years,” he recalled. “Whenever people from another university or a company wanted to port and use a program, we gladly let them. You could always ask to see the source code.”119

  Like a good hacker, Stallman defied restrictions and locked doors. With his fellow students, he devised multiple ways to break into offices where there were forbidden terminals; his own specialty was climbing through the false ceilings, pushing aside a tile, and lowering a long strip of magnetic tape tipped with wads of sticky duct tape to open door handles. When MIT instituted a database of users and a system of strong passwords, Stallman resisted, and he rallied his colleagues to do so as well: “I thought that was disgusting, so I didn’t fill out the form and I created a null-set password.” At one point a professor warned that the university might delete his directory of files. That would be unfortunate for everyone, Stallman replied, since some of the system’s resources were in his directory.120

  Unfortunately for Stallman, the hacker camaraderie at MIT began to dissipate in the early 1980s. The lab bought a new time-sharing computer with a software system that was proprietary. “You had to sign a nondisclosure agreement even to get an executable copy,” Stallman lamented. “This meant that the first step in using a computer was to promise not to help your neighbor. A cooperating community was forbidden.”121

  Instead of rebelling, many of his colleagues joined for-profit software firms, including a spinoff from the MIT lab called Symbolics, where they made a lot of money by not sharing freely. Stallman, who sometimes slept in his office and looked like he shopped in a thrift store, did not share their money-seeking motivations and regarded them as traitors. The final straw came when Xerox donated a new laser printer and Stallman wanted to institute a software hack so that it would warn users on the network when it jammed. He asked someone to provide the printer’s source code, but he refused, saying he had signed a nondisclosure agreement. Stallman was morally outraged.

  All of these events turned Stallman into even more of a Jeremiah, railing against idolatry and preaching from a book of lamentations. “Some people do compare me with an Old Testament prophet, and the reason is Old Testament prophets said certain social practices were wrong,” he asserted. “They wouldn’t compromise on moral issues.”122 Neither would Stallman. Proprietary software was “evil,” he said, because “it required people to agree not to share and that made society ugly.” The way to resist and defeat the forces of evil, he decided, was to create free software.

  So in 1982, repelled by the selfishness that seemed to pervade Reagan-era society as well as software entrepreneurs, Stallman embarked on a mission to create an operating system that was free and completely nonproprietary. In order to prevent MIT from making a claim to any rights to it, he quit his job at the Artificial Intelligence Lab, though he was allowed by his indulgent supervisor to keep his key and continue using the lab’s resources. The operating system Stallman decided to develop was one that would be similar to and compatible with UNIX, which had been developed at Bell Labs in 1971 and was the standard for most universities and hackers. With a coder’s subtle humor, Stallman created a recursive acronym for his new operating system, GNU, which stood for GNU’s Not UNIX.

  In the March 1985 issue of Dr. Dobb’s Journal, a publication that sprang out of the Homebrew Computer Club and People’s Computer Company, Stallman issued a manifesto: “I consider that the Golden Rule requires that if I like a program I must share it with other people who like it. Software sellers want to divide the users and conquer them, making each user agree not to share with others. I refuse to break solidarity with other users in this way. . . . Once GNU is written, everyone will be able to obtain good system software free, just like air.”123

  Stallman’s free software movement was imperfectly named. Its goal was not to insist that all software come free of charge but that it be liberated from any restrictions. “When we call software ‘free,’ we mean that it respects the users’ essential freedoms: the freedom to run it, to study and change it, and to redistribute copies with or without changes,” he repeatedly had to explain. “This is a matter of freedom, not price, so think of ‘free speech,’ not ‘free beer.’ ”

  For Stallman, the free software movement was not merely a way to develop peer-produced software; it was a moral imperative for making a good society. The princi
ples that it promoted were, he said, “essential not just for the individual users’ sake, but for society as a whole because they promote social solidarity—that is, sharing and cooperation.”124

  To enshrine and certify his creed, Stallman came up with a GNU General Public License and also the concept, suggested by a friend, of “copyleft,” which is the flipside of asserting a copyright. The essence of the General Public License, Stallman said, is that it gives “everyone permission to run the program, copy the program, modify the program, and distribute modified versions—but not permission to add restrictions of their own.”125

  Stallman personally wrote the first components for the GNU operating system, including a text editor, a compiler, and many other tools. But it became increasingly clear that one key element was missing. “What about the kernel?” Byte magazine asked in a 1986 interview. The central module of an operating system, a kernel manages the requests from software programs and turns them into instructions for the computer’s central processing unit. “I’m finishing the compiler before I go to work on the kernel,” Stallman answered. “I am also going to have to rewrite the file system.”126

  For a variety of reasons, he found it difficult to complete a kernel for GNU. Then, in 1991, one became available not from Stallman or his Free Software Foundation, but from a most unexpected source: a twenty-one-year-old toothy and boyish Swedish-speaking Finn at the University of Helsinki named Linus Torvalds.

  * * *

  Linus Torvalds’s father was a Communist Party member and TV journalist, his mother a student radical and then print journalist, but as a child in Helsinki he became more interested in technology than in politics.127 He described himself as “good at math, good at physics, and with no social graces whatsoever, and this was before being a nerd was considered a good thing.”128 Especially in Finland.

  When Torvalds was eleven, his grandfather, a professor of statistics, gave him a used Commodore Vic 20, one of the first personal computers. Using BASIC, Torvalds began writing his own programs, including one that amused his younger sister by writing “Sara is the best” over and over. “One of the biggest joys,” he said, “was learning that computers are like mathematics: You get to make up your own world with its own rules.”

  Tuning out his father’s urgings to learn to play basketball, Torvalds focused instead on learning to write programs in machine language, the numerical instructions executed directly by a computer’s central processing unit, exposing him to the joy of being “intimate with a machine.” He later felt lucky to have learned assembly language and machine code on a very basic device: “Computers were actually better for kids when they were less sophisticated, when dweebie youngsters like me could tinker under the hood.”129 Like car engines, computers eventually became harder to take apart and put back together.

  After enrolling in the University of Helsinki in 1988 and serving his year in the Finnish Army, Torvalds bought an IBM clone with an Intel 386 processor. Unimpressed with its MS-DOS, which Gates and company had produced, he decided that he wanted to install UNIX, which he had learned to like on the university’s mainframes. But UNIX cost $5,000 per copy and wasn’t configured to run on a home computer. Torvalds set out to remedy that.

  He read a book on operating systems by a computer science professor in Amsterdam, Andrew Tanenbaum, who had developed MINIX, a small clone of UNIX for teaching purposes. Deciding that he would replace the MS-DOS with MINIX on his new PC, Torvalds paid the $169 license fee (“I thought it was outrageous”), installed the sixteen floppy disks, and then started to supplement and modify MINIX to suit his tastes.

  Torvalds’s first addition was a terminal emulation program so that he could dial into the university’s mainframe. He wrote the program from scratch in assembly language, “at the bare hardware level,” so he didn’t need to depend on MINIX. During the late spring of 1991, he hunkered down to code just as the sun reappeared from its winter hibernation. Everyone was emerging into the outdoors, except him. “I was spending most of my time in a bathrobe, huddled over my unattractive new computer, with thick black window shades shielding me from the sunlight.”

  Once he got a rudimentary terminal emulator working, he wanted to be able to download and upload files, so he built a disk driver and file system driver. “By the time I did this it was clear the project was on its way to becoming an operating system,” he recalled. In other words, he was embarking on building a software package that could serve as a kernel for a UNIX-like operating system. “One moment I’m in my threadbare robe hacking away on a terminal emulator with extra functions. The next moment I realize it’s accumulating so many functions that it has metamorphosed into a new operating system in the works.” He figured out the hundreds of “system calls” that UNIX could do to get the computer to perform basic operations such as Open and Close, Read and Write, and then wrote programs to implement them in his own way. He was still living in his mother’s apartment, often fighting with his sister Sara, who had a normal social life, because his modem hogged their phone line. “Nobody could call us,” she complained.130

  Torvalds initially planned to name his new software “Freax,” to evoke “free” and “freaks” and “UNIX.” But the person who ran the FTP site he was using didn’t like the name, so Torvalds resorted to calling it “Linux,” which he pronounced, similarly to the way he pronounced his first name, “LEE-nucks.”131 “I never wanted to use that name because I felt, OK, that’s a little too egotistical,” he said. But he later conceded that there was a part of his ego that enjoyed getting acclaim after so many years of living in the body of a reclusive nerd, and he was glad he went along with the name.132

  In the early fall of 1991, when the Helsinki sun started disappearing again, Torvalds emerged with the shell of his system, which contained ten thousand lines of code.V Instead of trying to market what he had produced, he decided simply to offer it publicly. He had recently gone with a friend to hear a lecture by Stallman, who had become an itinerant global preacher for the doctrine of free software. Torvalds didn’t actually get religion or embrace the dogma: “It probably didn’t make a huge impact on my life at that point. I was interested in the technology, not the politics—I had enough politics at home.”133 But he did see the practical advantages of the open approach. Almost by instinct rather than as a philosophical choice, he felt Linux should be freely shared with the hope that anyone who used it might help improve it.

  On October 5, 1991, he posted a cheeky message on the MINIX discussion newsgroup. “Do you pine for the nice days of minix-1.1, when men were men and wrote their own device drivers?” he began. “I’m working on a free version of a minix-lookalike for AT-386 computers. It has finally reached the stage where it’s even usable (though may not be depending on what you want), and I am willing to put out the sources for wider distribution.”134

  “It wasn’t much of a decision to post it,” he recalled. “It was how I was accustomed to exchanging programs.” In the computer world, there was (and still is) a strong culture of shareware, in which people voluntarily sent in a few dollars to someone whose program they downloaded. “I was getting emails from people asking me if I would like them to send me thirty bucks or so,” Torvalds said. He had racked up $5,000 in student loans and was still paying $50 a month for the installment loan on his computer. But instead of seeking donations he asked for postcards, and they started flooding in from people all over the world who were using Linux. “Sara typically picked up the mail, and she was suddenly impressed that her combative older brother was somehow hearing from new friends so far away,” Torvalds recalled. “It was her first tip-off that I was doing anything potentially useful during those many hours when I had the phone line engaged.”

  Torvalds’s decision to eschew payments came from a mix of reasons, as he later explained, including a desire to live up to his family heritage:

  I felt I was following in the footsteps of centuries of scientists and other academics who built their work on the foundation
s of others. . . . I also wanted feedback (okay, and praise). It didn’t make sense to charge people who could potentially help improve my work. I suppose I would have approached it differently if I had not been raised in Finland, where anyone exhibiting the slightest sign of greediness is viewed with suspicion, if not envy. And yes, I undoubtedly would have approached the whole no-money thing a lot differently if I had not been brought up under the influence of a diehard academic grandfather and a diehard communist father.

  “Greed is never good,” Torvalds declared. His approach helped turn him into a folk hero, suitable for veneration at conferences and on magazine covers as the anti-Gates. Charmingly, he was self-aware enough to know that he relished such acclaim and that this made him a little bit more egotistical than his admirers realized. “I’ve never been the selfless, ego-free, techno-lovechild the hallucinating press insists I am,” he admitted.135

  Torvalds decided to use the GNU General Public License, not because he fully embraced the free-sharing ideology of Stallman (or for that matter his own parents) but because he thought that letting hackers around the world get their hands on the source code would lead to an open collaborative effort that would make it a truly awesome piece of software. “My reasons for putting Linux out there were pretty selfish,” he said. “I didn’t want the headache of trying to deal with parts of the operating system that I saw as the crap work. I wanted help.”136

  His instinct was right. His release of his Linux kernel led to a tsunami of peer-to-peer volunteer collaboration that became a model of the shared production that propelled digital-age innovation.137 By the fall of 1992, a year after its release, Linux’s newsgroup on the Internet had tens of thousands of users. Selfless collaborators added improvements such as a Windows-like graphical interface and tools to facilitate the networking of computers. Whenever there was a bug, someone somewhere stepped in to fix it. In his book The Cathedral and the Bazaar, Eric Raymond, one of the seminal theorists of the open software movement, propounded what he called “Linus’s Law”: “Given enough eyeballs, all bugs are shallow.”138