What inspires open source programmers, defines their culture and sets the open source world apart from that of proprietary software development? That's an important question for understanding what drives the creation of monumental platforms such as Linux, OpenStack and Hadoop. It's also one that can only be answered through a careful look at the history of what open source leaders have called "hacker culture" and the "hacker ethic."
In most existing tellings, the history of hacker culture is clear-cut, and can be traced cleanly across the various stages in the development of what we know today as free and open source software. Eric S. Raymond's essay, "A Brief History of Hackerdom," which probably best encapsulates the prevailing thinking today on the origins of the culture behind open source, traces "the hacker culture as we know it" to a model railroad club at MIT in 1961.
Members of the club, who evinced a passionate commitment to tinkering, exploring and sharing knowledge, went on to staff the MIT Artificial Intelligence Lab, which was one of the early centers of the computing industry. The lab served to incubate hacker culture further and, when many of its members took jobs at other universities or in industry, they took hacker culture with them, spreading the spirit of discovery and creativity. The lab was also the place where Richard Stallman, who in the early 1980s founded the Free Software Foundation and the GNU operating system, worked in the early 1970s.
Raymond's narrative of the origins of hacker culture comes mostly from Steven Levy's 1984 book, "Hackers." However, Raymond, who was writing in the late 1990s, took the hacker story further by arguing that, as the computing industry became more commercialized during the 1980s, the hacker culture was marginalized, as the passion for tinkering and having fun that had predominated in earlier communities of computer programmers gave way to a focus on efficiency, profit and centralized authority. For Raymond, the rise of Linux, and the fantastic success that it and other decentralized open source projects enjoyed beginning in the late 1990s, signaled the "revenge of the hackers" and the resurgence of the hacker culture at the center of the computing industry.
The narrative that Raymond tells has much to recommend it. But it also has some apparent flaws, and getting to the bottom of them is key to understanding what really drives open source programmers, historically and in the present.
The Real Hacker Culture
One of the most obvious weaknesses in Raymond's account of hackerdom and open source is the clear disconnect between the principles that governed the community of MIT hackers in the 1960s and 1970s (and, in the 1980s, powered Stallman's free software movement) and those that became important to Linux developers in the 1990s. People such as Linus Torvalds, who launched Linux as a computer science student in 1991, were much less ideological in the way they thought about computing and programming than the people in the GNU movement were—and, arguably, Torvalds's laissez-faire attitudes were what assured the terrific success of Linux, as compared to GNU, which failed to produce a usable open source operating system kernel.
To be sure, Torvalds and his cohort thought of programming as a way to have fun—as the title of Torvalds's autobiography, "Just for Fun," makes clear—and they were committed to collaboration and cooperation with others. But they were not militant in their devotion to such endeavors, nor did they think of software development as a sort of ideological crusade. They did not, in other words, evidently seethe against the world of closed-source software development, or view what they were doing as a way to gain "revenge" against some kind of evil empire of proprietary software, as Raymond supposed. (If Torvalds seethed against anything in the early 1990s, it was Minix, an academic Unix clone whose source code was publicly available, and which hardly incarnated the principles that, according to Raymond, open source hackers were fighting against.)
It is also difficult to square an interpretation of Linux and the open source movement of the 1990s as the triumph of hacker culture with the fact that, since that time, the purest representatives of the MIT-era hacker culture—people such as Stallman—have remained on the sidelines of the mainstream open source community. (Actually, Stallman strongly disavows association with the open source movement, which he views as "missing the point" of the software-freedom principles that emerged from the hacker culture of the 1960s and 1970s.)
Instead, the open source world for nearly the last two decades has been dominated by individuals and companies that have done much to advance Linux development, but have shown little interest in the ideas that Raymond associated with hacker culture. Canonical has built business models that pair Ubuntu Linux with various proprietary products, and its founder, Mark Shuttleworth, made his fortune in the 1990s writing security-certificate software, not hacking in the free software or open source communities. Jim Whithurst, Red Hat's current CEO, who has led the company quite successfully since 2007, made his name before that as an executive at Delta Airlines, in an industry that is about as far removed from the epicenter of the hacker culture as one can get.
Last but not least, the whole premise of a monolithic hacker culture is murky and difficult to digest—not only because Steven Levy's account of its origins at MIT in the 1960s lacks documentation, and occasionally makes some fuzzy claims that are difficult or impossible to substantiate, but also because Raymond's definition of hacker values various substantially from Levy's—and neither definition accounts consistently for what leaders of the free software and open source worlds have actually done.
For instance, according to Levy, one of the tenets of the hacker ethic—which he describes in explicit terms in chapter two of "Hackers"—is to "Mistrust Authority [and] Promote Decentralization." Principled opposition to authority of any kind is absent from Raymond's description of open source culture, where, on the contrary, programmers' subordination to "benevolent dictators" is key to the organizational success of most software projects, according to Raymond. And a commitment to decentralization was not evident in, for example, the GNU project, where development was centrally coordinated, much more so than it has ever been for Linux (a fact that Raymond seemed to see as part of the explanation for Linux's success in the face of GNU's flagging progress).
For these reasons, the very idea that there is such a thing as a unified and consistent "hacker culture," which explains how open source programmers think, and what makes open source software and companies different from their closed-source counterparts, is a very tough sell.
Hackers, or Academics?
And, anyway, to the extent that it is possible to identify a distinct set of values that permeate open source culture—such as the importance of peer review and the openness of information—they are not very different at all from the principles that prevail (most of the time, at least) within the world of academia—which, like modern open source, coexists perfectly well with, and often complements, commercial endeavors.
It's also hard to imagine how academic values could not have exerted a strong influence on the figures who played important roles in laying the foundation for what became Linux and open source, since virtually all of them, after all, began their careers in universities. Even Unix itself, despite being born in the commercial setting of Bell Labs, saw its first real-world applications largely on university campuses, and one of its founders, Ken Thompson, was committed enough to academia to take a visiting professorship at Berkeley in 1975 after helping to create Unix in the late 1960s.
The notion that modern open source culture can best be understood in terms of its origins in academic settings, rather than with reference to the nebulous hacker ethic or culture that Steven Levy narrowly traced to a model railroad club, is not a new idea. Nikolai Bezroukov was perhaps the first to make the argument in a major public way, when he published a paper in the journal "First Monday" in 1999 which argued, among other things, that open source functions "more like a regular scientific community than some [open source software] apologists would like to acknowledge." Raymond flatly and aggressively rejected such claims, and contended in another essay that, far from being "genetically related," academia and open source simply happened to share some similarities "because they've both evolved the one most optimal social organization for what they're trying to do." I don't think that's true.
Of course, I am an academic by trade—though certainly not a computer scientist—so maybe my tendency to see the origins of open source culture in academia, rather than the hacker ethic Levy and Raymond struggled to define, is simply a reflection of my interpretive bias. I'd love to know what you all think, since, as I have noted, I am trying to understand these issues well enough to write a book about them. Feel free to share thoughts below, or to contact me privately.