F. The Creation of the Open Source Idea
¶ 1 Leave a comment on paragraph 1 0 One of the difficulties for Bentham’s premise that people are rational and self-interested is that, on the surface of it, they sometimes are not. Most people at times clearly like to do work for something larger than themselves, for example. Soldiers, athletes on teams, mountain climbers, family members in moments of celebration or crisis, and law students on the law review often recall the intense energy and feeling of solidarity that comes from working on behalf of the group, frequently with an amount of effort that is clearly not in one’s obvious self-interest. People will often say, in fact, they have worked harder in these situations than in contexts where they were working purely for their own gain. It is only because utilitarian reasoning has become so taken for granted in our culture that the same people who speak fondly of such moments of working for the group can turn around and say things like “everybody knows that welfare discourages initiative” or “if people can’t make money from it, it won’t get done.”
¶ 2 Leave a comment on paragraph 2 0 Yet more than a few computer engineers know from personal experience that sometimes people will do things even if they could make more money from doing something else, and, as we saw in chapter four, a number of those inside the internet engineering community saw computer networking as a case in point. There are times when some of the best work is done, not to maximize profit, but out of passion or commitment to something larger. In the early 1990s, as the internet spread like videowildfire and became coupled in the popular imagination to an unregulated free market, some of these engineers started to raise their voices in protest. Michael and Ronda Hauben, for example, published Netizens, an important book that compellingly detailed the numerous ways in which the internet was born of and embodied, not capitalist self-interest, but forms of spirited and deliberately collective action. Unix, they showed, was designed from the ground up to enable collaboration and the sharing of interoperable software tools so as to encourage collective improvement of the system. The internet appeared, not just because of various nonprofit arrangements, but because of its deliberate design, on both a technical and social level, as an open system built for shared collective effort. Usenet and other nonprofit collaborative communication systems both spread much of the knowledge that made the global internet possible and taught a generation of technical professionals the value of online, citizenly collaboration.
¶ 3 Leave a comment on paragraph 3 0 It is perhaps because of the overwhelming dominance of the taken-for-granted utilitarianism in American society that those with different views sometimes feel compelled to leap to the other extreme. The Haubens took their crucial observation―that some of our most advanced technologies like the internet did not emerge solely from self-interested, profit-motivated contexts and logics―and used it to then insist that nonprofit behavior is obviously morally and technically superior. They defined and valorized the “netizens,” not as people who merely used the internet, but people who “understand the value of communal work and the collective aspects of public communication . . . people who care about Usenet and the bigger Net and work towards building the cooperative and collective nature which benefits the larger world.” “The so-called ‘free market’,” they argued, “is not a correct solution for the problem of spreading network access to all.” Never mind the ideologically blurry tradition of Vannevar Bush, with its technocratic faith in the ability of experts to elegantly mix public and private efforts or all the back-and-forth movement between private and public contexts characteristic of many of the engineers that the Haubens lionized. In the first half of the 1990s, faced with the spread of the bizarre claim that the triumph of the internet was somehow a triumph of free individual initiative and of the market, the Haubens seized on the opposite claim: Netizens presents a picture of computer communications as a nonprofit communitarian utopia, threatened by ignorant capitalist managers.
¶ 4 Leave a comment on paragraph 4 0 This stance of reacting to utilitarian dominance with communitarian purism is shared by Richard Stallman. Today he is a hacker hero, but in the early 1990s his effort to make a free and open clone of the Unix operating system was known only to narrow computer engineering circles (and those readers of Levy’s Hackers that made it all the way to the end). Stallman, upset by the restrictions that ensued from the commercialization of systems like Unix in the late 1970s and 1980s, set out to create, not just another version of Unix, but an alternative software universe in which the free sharing of software code was required by those using it; towards that end he created the Free Software Foundation and a unique form of copyright license, the General Public License or GPL, that specified that a piece of software could be freely shared and used provided that whoever distributes it also freely distributes the source code and any modifications to that source code. Instead of, say, releasing software code into the public domain, copyright was retained, but for the purpose of maintaining free and open distribution, not for the purpose of preventing others from selling the software. It was less a nonproperty or public domain approach than a kind of antiproperty. In 1991, when a Finnish college student named Linus Torvalds, as a way to learn about operating systems, began building an experimental operating system kernel, he early on copyrighted it using the GPL, so that he and collaborators could also use software tools created by Stallman and his colleagues.
¶ 5 Leave a comment on paragraph 5 0 In research and academic environments, where one’s prestige and job security often depends on open publication and collaboration, this made a certain kind of sense. But in a neoliberal world that was both in love with high technology and that seemed completely stuck on the assumption that innovation only sprang from the unfettered pursuit of profit, Stallman’s approach was so different as to be almost invisible. In its second issue in 1993, Wired did run an article titled “Is Stallman Stalled?” that briefly summarized Stallman’s approach. The author quotes Stallman, “I don’t think that people should ever make promises not to share with their neighbor,” and then spends close to half of the article discussing problems with Stallman’s effort. “Things seem to have bogged down,” the article states, noting delays in software production and problems with financing. The tone, particularly in the context of Wired at the time, is that of the knowing adult speaking of an idealistic child. If there is any success here, the article suggests, it is in the for-profit efforts to support Stallman’s programs, like Cygnus Support. The article is subtitled, “One of the Greatest Programmers Alive Saw a Future Where All Software Was Free. Then Reality Set In.” “Reality,” the reader must assume, is the world of self-interest, profit, and the market; sharing is a naive ideal. Wired was in effect telling its readers to move on, nothing to see here; the action is all in the romantic entrepreneurialism of start-ups.
¶ 6 Leave a comment on paragraph 6 0 In the next few years, as the business community took Wired‘s admonition to heart and plunged the American economy into the stock bubble, those with doubts about the “new economy” mantra quietly went about their business, while looking on helplessly at the dotcom rush. Stallman, Torvalds, and many other engineers continued tinkering with their systems. Scholars like Lessig and Boyle kept looking for ways to use the new technological situation to justify rethinking some core principles about law and society. The Haubens and others enamored of the communitarian moments at the internet’s roots kept promoting their vision, most often on the internet itself. But the world still seemed completely enamored of markets. The stock market continued to rise, and market practices proliferated wildly across the world, appearing successfully in places never before thought possible like Russia and China and in industries like telecommunications, where the market-driven mobile phone rapidly leap-frogged the regulated or state-owned land line telephone systems in many parts of the world. The many and crucial nonproprietary aspects of the internet, therefore, remained largely ignored. In spite of the best efforts of those like the Haubens, the internet’s history of nonprofit development, the many aspects of its development that involved various degrees of deliberate openness, the fact that much of its rapid success had to do with its open, welcoming structure and design: all these details could not penetrate the fog of the market enthusiasms of the early 1990s.
¶ 7 Leave a comment on paragraph 7 0 But there remained the problem of Microsoft. Microsoft’s dominance was still galling from the point of view of the programmer who wanted to believe that programming was something like art deserving of recognition. The stock bubble, as we have seen, could not have existed without the narrative of the romantic programmer; it would not have been compelling enough to take off without all the tales of rebel-hero programmers like Andreessen, which offered the hope of fusing the pursuit of wealth with the pursuit of self-expression and rebellion. Yet, as the 1990s progressed, Microsoft only increased its control over the operating system market and by 1996 started making dramatic inroads into Netscape’s share of the internet browser market. Major corporations like IBM and Apple found themselves gasping for breath in the face of Microsoft’s dominance.
¶ 8 Leave a comment on paragraph 8 0 What it took to sever the dominant culture’s association between the internet and the neoliberal market was Linux. By 1994, Torvalds and fellow tinkerers improved their new operating system kernel to the point where it could be combined with Richard Stallman’s work to make a rather effective complete software system, which began to look attractive to some software professionals. On the one hand, it was a familiar system to many experienced engineers already working in the area, as it was derived from Unix. On the other, it overcame some of the problems associated with the existing fragmentation of Unix into competing systems, and, because it was open, it allowed for easy tinkering and improvement. Finally, Linux went through a period of rapid technical improvements at the same time that Microsoft’s operating system dominance became so complete that it could afford to release software that was merely good enough in areas where it dominated and focus its vast resources on areas where it had competition, like the browser market. The average consumer booting up Windows 95 would not likely notice. Computer engineers, however, did; beginning with sheer technical admiration for Linux coupled to annoyance with the technical limitations of Microsoft’s software, they then began to wonder what the moral of this story was for the economic organization of software production.
¶ 9 Leave a comment on paragraph 9 0 This was the context in which a Unix programmer and gadfly named Eric S. Raymond wrote an essay called “The Cathedral and the Bazaar,” which he delivered at programming conferences in 1997 and which went on to circulate beyond the internet into the offices of key business executives and copyright lawyers, initiating a ground shift in corporate strategy. This article was passed around at Netscape in the run-up to its decision to open source its browser in January 1998. Not long after that, Apple would decide to base its next generation operating system on an open source foundation, and IBM and Sun Microsystems would adopt open sourcing as a key strategy. Soon, Linux-based companies like Red Hat would become stock market darlings.
¶ 10 Leave a comment on paragraph 10 0 It is significant that the arguments of “The Cathedral and the Bazaar” are not communitarian. In contrast with Stallman or the Haubens, Raymond dismisses appeals to altruism out of hand. The central rhetorical accomplishment of the piece rather is to frame voluntary labor in the language of the market; the core trope is to portray Linux-style software development like a bazaar―the archetype of a competitive marketplace―whereas more centralized and controlled software production is portrayed as hierarchical and centralized―and thus inefficient―like a cathedral: static, inefficient, medieval. (While Raymond seems to have had previous efforts in the Unix world in mind when describing cathedral-style software development, it is a safe bet that many readers thought of Microsoft.) Raymond thus disarticulated the metaphor of the market from conventional capitalist modes of production and reconnected it with a form of voluntary labor, of labor done for its own sake.
¶ 11 Leave a comment on paragraph 11 0 Part of what makes this curious reversal work is Raymond’s focus on programmers’ motivations. For an essay about such a dry and technical topic as the management of software development, there’s an awful lot of reference to the internal feelings, psychological makeup, and desires of programmers. (Subsequent discussion of open software seems to have maintained some of this focus.) Almost like a hip Entwicklungsroman, Raymond presents a first-person account of his own experiences in software development, during which he tells the story of how he became converted to the Linux software model. This narrative of personal revelation is interspersed with numbered principles or aphorisms, the first of which is: “Every good work of software starts by scratching a developer’s personal itch.” Because Raymond’s audience is in the worlds of business and law, he immediately sets out to reconcile his psychologically tainted portrayal of motivation with a utilitarian one. “The ‘utility function’ Linux hackers are maximizing,” Raymond continues, “is not classically economic, but is the intangible of their own ego satisfaction and reputation among other hackers.” Raymond goes on to draw analogies with fan subcultures, wherein the enhancement of reputation among the other members of the community is understood as a key motivation.
¶ 12 Leave a comment on paragraph 12 0 Much of the piece is devoted to explaining why a wide-open approach to software development involving frequent borrowing and sharing of code, early and frequent releases of updates that have the effect of involving users in development, and attention to maintaining positive social relations among participants all combined in the case of Linux to create better software. But it’s crucial to the essay’s effect that Raymond frames the motivation to write software as something born of a not entirely rational fascination or ambition (an itch), of a desire to have one’s accomplishments recognized not with money but with the psychological satisfactions of acclaim. One could of course criticize this as both an empirical description and as a philosophical argument, but what’s significant is, first, how the dream of having one’s “itch,” one’s inner passions, acknowledged by a community of the like-minded is a characteristically romantic structure of feeling and, second, how much Raymond’s statement of the problem, whether or not it is philosophically coherent, resonated with the computer culture at large and had some impact on the larger business culture in a way that communitarian or managerial arguments have not. In the minds of quite a variety of people, this vision of passionate programmers provides a much more appealing way to deal with the perennial industrial problem of monopoly than something like industrial policy or antitrust law.
¶ 13 Leave a comment on paragraph 13 0 Raymond publicly presented “The Cathedral and the Bazaar” in September 1997, after which the essay began to circulate. At the time, the executives at the Netscape corporation, after riding high on the early stock bubble, had been sensing doom. In 1997, they had been watching Microsoft’s share of the web browser market rise from negligible to almost half of the market, while Netscape was making almost no income and Microsoft’s profits from its other software was setting records. So in January 1998, they announced they would publish the source code for the browser for free distribution, and in February they invited Raymond and several others to a meeting to help them plan their new strategy. Those in attendance at the meeting saw this as an important opportunity to get the corporate community to take the free software community seriously and towards that end chose to follow a pragmatic path of using the term open source software (instead of Stallman’s term free) and of emphasizing technical advantages rather than ideals―which qualifies as an act of rhetorical genius. A few weeks later, Raymond and others would found the Open Source Initiative to support these efforts, and the organization would be involved in strategy shifts at Apple, IBM, and other companies. Today, Linux and other forms of open source software are central to many different businesses world wide and can be found on everything from cell phones to massive research computers. If one had predicted this chain of events in 1994, one would have been dismissed from almost all directions as hopelessly naive.
¶ 14 Leave a comment on paragraph 14 0 The argument here is not that this single essay by itself directly caused major corporations to adopt new strategies; rather, Raymond’s essay helped promulgate a way of understanding software development that played a key role in the corporate shift. It was a necessary but not sufficient part of the conditions of possibility of the move towards open software. Of course, these companies had an economic interest in the new strategy, especially given the Microsoft monopoly. But the economic conditions behind the change had been in existence for several years; an economic explanation alone cannot explain why these companies made their policy changes all within a roughly one-year period (1998).
¶ 15 Leave a comment on paragraph 15 0 In different times, open source might have seemed unremarkable; when Bell Labs gave up its patents on the transistor as a result of a consent decree in the 1950s, this was viewed as a reasonable solution to a complex problem, not egregious theft. But in the first half of the 1990s, paying for labor that produced something that could only be given away for free would have been considered deeply irrational. In 1997, the embrace of open source―as modest as it was―marked a profound shift in the common sense of U.S. political economic thinking in the high-tech realm. After the arrival of the open source movement, the neoliberal assumption that more-property-protection-is-better was no longer unassailable.
¶ 16 Leave a comment on paragraph 16 0 So how did this happen? Traditional economic reasoning does offer explanations of why a company might use open source software. A consumer device manufacturer like Tivo makes its money selling devices, not software, so using Linux makes sense not only because the software is free but because its open character allows Tivo to easily modify it according to its needs and to draw on the global support of the Linux community. Companies like Red Hat or IBM can make a solid profit by offering technical support to companies using their software, while allowing the software itself to be freely copied and shared. And when faced, as many high-tech companies in the late 1990s did, with the huge market power of a company like Microsoft, taking a gamble on open source software might have seemed like the only alternative. The only way to compete with Microsoft’s market power was to offer a platform that was free to consumers and that, because of its free and open character, could create a united front among Microsoft’s competitors.
¶ 17 Leave a comment on paragraph 17 0 But these economic reasons for turning to open source cannot explain why the shift happened when it did. All of these economic factors were in place in 1994, and yet no major company even briefly flirted with the idea of open software at the time. It was not until the fall of 1997, with the circulation of Raymond’s essay, that the idea could even begin to get attention, and then in 1998 and 1999 the concept all of a sudden became relatively mainstream. It took Raymond’s articulation of free software within a romantic individualist structure of feeling―and its appearance against the backdrop of the Microsoft problem―to lay the conditions for mainstream acceptance of the idea.
¶ 18 Leave a comment on paragraph 18 0 The acceptance was not instantaneous, of course, and as of this writing is not universal. Libertarianism of whatever variety is premised on the idea of private property, and so it is not surprising that many of the libertarian faith at first scoffed at open source. For example, in late 1998, once open source had gained some attention in the media, Wayne Crews of the Competitive Enterprise Institute published a critique of the open source movement (as part of a series titled “C:\spin: An Occasional Commentary on Regulation of High Technology―From an Undiluted Free Market Perspective”). Arguing that “like free love, open-source code is fun, but it’s probably not a way to run the world,” Crews wrote that,
¶ 19 Leave a comment on paragraph 19 0 for the most part, the prospect of becoming fabulously wealthy, not the desire to give things away, drives software innovation. Nearly all “freeware” programs―whether word processors, image editors, games, or browsers―pale beside superior commercial versions. Even Netscape’s release of the source code for Navigator―applauded by the open-source advocates―wasn’t a fundamental embrace of their doctrine, but an effort to create a pipeline for offering other, more profitable services. Conveniently ignored also was that the Netscape giveaway occurred after the IPO that made multimillionaires out of its founders.
¶ 20 Leave a comment on paragraph 20 0 Other critics were even less circumspect. Bill Gates and Steve Ballmer at different times insinuated that there is something communist about Linux. Forbes magazine scoffed at the open software “movement’s usual public image of happy software proles linking arms and singing the ‘Internationale’ while freely sharing the fruits of their code-writing labor.”
¶ 21 Leave a comment on paragraph 21 0 But such dismissals came across as shrill and tinged with desperation. They did not have the same kind of force that they might have had earlier in the same decade. Here was a technically sophisticated operating system that in some contexts seemed better than Microsoft’s products. As Windows users grew accustomed to the “blue screen of death”―the end result of a system crash in Windows, a frequent occurrence in the mid-1990s―the sheer technical quality of Linux stood as a glaring refutation of one of the central claims of the neoliberal argument about intellectual property; here was better software created without the incentives of property protection. And, after elevating disheveled programmers to the status of cultural heroes in the earlier 1990s―remember that it was Netscape’s publicity strategy to foreground its young programmers as opposed to its managers―the culture found it was exactly those heroes who were now increasingly celebrating something that seemed to point in a very different direction. In the intellectual space created by readers and writers of Wired, one had to recognize in Raymond’s rendition of open source much of the same Byronic attraction that had driven the magazine’s rhetoric in its earliest days; scoff at open source and you might start to look like one of John Perry Barlow’s dinosaurs, one of the old suits who didn’t get it. The romance of open source was too alluring to ignore, and the previous few years had taught the culture that this might not be a fleeting trend; the romantic allure of the hacker had already influenced the global economy by way of its role in the stock bubble.
¶ 22 Leave a comment on paragraph 22 0 The full intellectual structure of a point of view is often revealed, not just in the statement of high ideals, but in what people define as pragmatic―in those moments when someone claims that it is time to be sensible, to make compromises, to take a middle road. So, for example, in response to Wayne Crew’s market-based dismissal of open source, fellow libertarian Esther Dyson offered a path towards reconciling with the movement. “There’s a fundamental misunderstanding here,” she wrote,
¶ 23 Leave a comment on paragraph 23 0 Open source software may be freely available, but someone *is* responsible for it. Most of the support for OS software is paid for; that’s how (many of) the hackers’ paychecks are funded. There is a lot of value―and money―floating around the world of OS. And yes, Netscape’s use of OS to make its other services attractive is a legitimate, acknowledged and sensible business model. . . . (It seems to me that there are religious extremists on both sides of what ought to be an argument about business models, not morality.)
¶ 24 Leave a comment on paragraph 24 0 If two years earlier Dyson felt comfortable stating that the market simply “works and is moral,” now, in her mind, it was time to move away from blanket statements about morality. Now it was about business models, about being sensible, about striking a middle path. This tone of moderation would allow someone of Dyson’s convictions to maintain a friendly public stance towards open source.
¶ 25 Leave a comment on paragraph 25 0 But that tone of moderation would also allow the loosening of the link between the romantic view of individual freedom and market libertarianism. For the first time in more than a decade, it became possible in business culture to seize the glamorous position of rebellious high tech while also supposing that, Thatcher notwithstanding, perhaps there is an alternative to the market.
¶ 26 Leave a comment on paragraph 26 0 The only way to try for ideas like that is by having lots of ideas―or by having the engineering judgment to take other peoples’ good ideas beyond where the originators thought they could go. . . . Andrew Tanenbaum had the original idea to build a simple native Unix for the 386, for use as a teaching tool. Linus Torvalds pushed the Minix concept further than Andrew probably thought it could go — and it grew into something wonderful. In the same way (though on a smaller scale), I took some ideas by Carl Harris and Harry Hochheiser and pushed them hard. Neither of us was `original’ in the romantic way people think is genius. But then, most science and engineering and software development isn’t done by original genius, hacker mythology to the contrary. The results were pretty heady stuff all the same―in fact, just the kind of success every hacker lives for! And they meant I would have to set my standards even higher. [Raymond, “The Cathedral and the Bazaar”]