The Need to Give: Free Software and the Nets
Date: Tue, 29 Sep 1998 00:24:43 -0700

A few weeks ago, O'Reilly Publishing sponsored an Open Source Developer Day. The conference took place in downtown San Jose, emerald city as ghostn town, in a hotel that only even partially fills up for conventions and conferences. In a ballroom-conference room with a raised stage for speakers and a few hundred filled seats, the big figures in open source came together to discuss the "movement." Eric Raymond, of Cathedral and Bazaar fame, was the keynote speaker. He focused most of his talk on the "enterprise market" and Linux's penetration of it. Linux, the phenomenon, has made recent notice in the straight economic press, in The Economist and Forbes, and so have several other of the free software projects. After
querying the audience to see who they were, and after discovering that the majority of the folks in attendance were hackers, Raymond delivered an entertaining tour through some of the more recent and significant achievements of Linux. Raymond's entire talk focused, nonetheless, only on the enterance of Linux as a serious player in the corporate enterprise server and high end computing markets. Linux's penetration of the highest ends of the computing world is an interesting story and it is one that at least can be measured somewhat. But the Linux phenomenon is much larger and includes a worldwide spread into PCs and even recycled 486's. The 486 market, for example, is of no financial significance in Silicon Valley at the moment but may prove to be of social and economic signficance globally.

There was little discussion by any of the participants of the larger social impact of free software based computing and networking. The rest of the day was spent discussing business models and the legal technics of licensing. A day of presentations and note taking on this or that approach was punctuated by Richard Stallman's declaration that John Ousterhout was a "parasite" on the free software movement. Ousterhout was up on the business models panel, describing his company, Scriptics's, planned support of the open source core of Tcl, the language he nursed to adolescence, and their simultaneous planned development of proprietary closed tools for Tcl as well as closed applications. Stallman walked up to an open microphone during the question period and said that it was interesting to see the behemoth of IBM, a representative for which was on the panel, sincerely entering in to the free software community by supporting the Apache project while John was planning on making the fruits of the community into closed and in Stallman's view, harmful, proprietary products.

Some people clapped while Stallman injected a little controversy into the affair. Others jeered. If Stallman hadn't goaded Ousterhout and, later, O'Reilly the "conference" may have ended as more a press conference than as a town meeting for the free software community. Some of the more official attendees were said to be embarassed by Stallman and he may not be asked back again. Most of the assembled seemed baffled and confused by the dissension and controversy. Many of the old timers just groaned, "Oh, there goes Stallman again." Some were worried that the hackers would be dismissed by the trade press because of Stallman's behavior.

A week later a VP from a software company thinking about going open source talked to me after he got a full report from someone they sent to the conference. "Stallman is a Communist", he said. "Stallman is no Communist" I said, laughing, thinking to myself that political economy was no where near any topic of conversation I heard all day. The closest Stallman ever got to talking about politics was to mention the U.S. Bill of Rights. "He's not even a Marxist," I interjected. The software community is not known for a highly articulated or nuanced view of political economics. Those in the proprietary software world and many of
the open source vendors are good proper businessmen and are not quite sure how to address a question that is not about either technical capacity or profits. Many did not even seem comfortable with the possibility that dissension and debate might be a good thing.

Stallman's very presence makes some in the free software communties uncomfortable. He's like an old cousin that shows up at the wrong time, is a little too loud and says things that other people think but that no one else will come out and say. Foremost amongst the traits that make the denizens of Silicon Valley uncomfortable is Stallman's contempt for the merely commercial. Stallman is, indeed, full of contempt for the merely commercial, for profit for profits sake, especially when that profit comes at the expense of the the free circulation of ideas and software. That contempt for profit for profit's sake is whatcontemporary executives, hip though they may be, find so consternating and odd. Expressing that kind of contempt in Silicon Valley is a little like going to Las Vegas and declaring contempt for gambling. But Stallman's antics make perfect sense within the context of free software and within the community of free software developers. It strikes me as a mark of consistency and mental precision that he persists in his strict interpretation of free software. His legally technical discussions of the GNU General Public License are brilliant expositions of what is known as, though not by RMS, the "viral" license. A "viral license" is one that legally binds the user to keep any modifications in the open source code free and open to further modification. The GPL has been very good, as they say, to Linux, as an example. The GNU project spent a considerable amount of time and money crafting a very clear and legally binding document that has served as a haven for many a free software developer. Linus Torvalds was spared the need, as many have been, to craft a license and set a precedent that would allow for the open and distributed development of his open source project.

Stallman's GNU project has done incalcuable good for free software. No one in the free software communities denies that. His insistance on no compromise can make many of them shift in their seats, however. He doesn't make the "suits" comfortable either. And he doesn't want to. Stallman doesn't carry a business card; he carries a "pleasure card," as he calls it, a card with his name on it and what appears to be a truncated personals ad, "sharing good books, good food...tender embraces...unusual sense of humor." The guy is clearly not looking for a job or a businessdeal. Friends perhaps or "community", but clearly not a business deal. But he's not against others making a profit from free software; he, in fact, encourages those that manage to make profitable businesses while contributing substantively to free software and free documentation for free software. He, like every other "hacker" in assembly at Open Source Developer Day that I managed to talk to, is a thoroughly pragmatic thinker. He knows that no business would come near free software if it did not offer a successful business model for them. He's just not willing to make compromises with those who would like to combine open source with closed and proprietary software. And his reasons are pragmatic; he reasons that if an open source project is canibalized or "parasitized" by the development of closed products it will be a detriment to the free flow of ideas and computing.

John Ousterhout's plans for Tcl are just plans at the moment. He'splaying with the possibility that he might be able to support the opensource development of Tcl while he develops proprietary tools on top of it. He, acknowledges that there will be some tension from the need for Scriptics to offer a return on investor capital and from the need for the community to have substantial evidence of support for the further free development of Tcl. Many future contributions could be lost if open source developers think that their efforts will only serve the closed products of Scriptics and not the community of Tcl users. Scriptics could lose funding, on the other hand, if investors think that everything will "just be given away."

Where's the business model? The source of the recent interest in the mainstream press has been the monetary success of companies that serve and support the free software communities. The large and growing user communities are spending quite a bit of money on support and commercially supported versions of free software products as well as documentation. The commercial Linux vendors are making significant revenues; C2net's commercial, strong encryption version of Apache will make the small company some $15 million dollars in revenue this year; O'Reilly will make over $30 million dollars in revenue on documentation of free software this year. Although the revenue figures are signficant for free software and peripheral businesses, they are, of course, dwarfed by the revenue figures for proprietary software. The synecdoche, of course again, for proprietary software is Microsoft or Bill Gates. Gates, whose personal fortune exceeds the combined wealth of the entire bottom forty percent of
the United States population, and whose corporation is the second wealthiest in the world behind the mammoth General Electric, represents the greatest monetary success ever achieved by any software enterprise.

As large as Microsoft looms, it would be a mistake to credit them withspurring or even inspiring the development of free software. Free software has it's own trajectory and its own history that both predates Microsoft and lies outside of Microsoft's ken. Free software is a child of abundance, of the free flow of ideas in the academy and in hacker communities, amongst an elite of developers and a fringe of hobbyists and enthusiasts. The communities of free software are fertile "phraities" that lie outside the normal bonds of business as usual and official policy. The fact that the abundance of free flowing ideas has reached a significant enough mass to support business models has less to do with presence of a clay footed proprietary monster such as Microsoft and more to do with the superior and more engaging model that free software offers users and developers.

Microsoft is, as Eric Raymond says, merely the most successful and largest example of the closed or proprietary model of software development. It is the model and not Microsoft in particular that open source and free software offer an alternative to. The free software model doesn't make near as much money; it makes better software. Enough people realize the superiority of free software to make it a threat to proprietary software in every niche it enters into. It's hard to think of anything that might threaten Microsoft, but the giant computer companies are like so many Romes who in their moments of greatest success already show signs of terminal decay. Perhaps Linux will threaten Microsoft's notoriously buggy cash cow NT as Eric Raymond predicts. Many apparent NT boxes out on thenetworks are already Linux boxes that simulate NT by using Samba. It is hard to tell, and in many ways it is beside the point.

Does the move to open source come from over-fullness, from a sense of abundance, or from desperation, from ressentiment? From within the communities of free software, the answer is obvious, the move to free software comes from an abundance. When a large, already commercial company decides to go open source, think Netscape, it is often seen as an act of desperation. The rising stars of the free software communities, the Cygnus's and the Red Hat Softwares had the community before the business model. It's much harder for a company to start with the business model and try to create a community. IBM did not open the source code for one of their products; they decided they wanted to leverage the community and the brand name of Apache. Netscape's competing web server has been put to rout by the superior market share of Apache as has everyone else.

One of free software's most significant contributions, in fact, has to be the way that it has made companies look for other business models besides proprietary software. Some of the most successful internet companies, in market capitalization at least, rely entirely on free software to serve their content. Yahoo is one prominent example of a Free BSD enterprise. The success of of Yahoo may hearten executives who are contemplating using free software to do their networking, but for the users of free software and for the developers of free software the process and the products are the motivation

Free software projects develop devoted communities that are extra-monetary. The metaphors of a gift economy are appropriate and useful here. After Marcel Mauss's Essai Sur le Don of 1920, after his series of readings on extra-monetary exchange and readings in ethnography and historiography, words, metaphors, and analyses of gift economies have spread into vocabularies. More recently, internetworked exchange of ideas and software has given fresh license to the use of the terms for gift economy. One of the terms for gift exchange that has had the most currency is the potlatch, the term that describes the gift-giving ceremonies of the Northwest Coast Tribes of North America. The potlatch is a "system for the exchange of gifts", a "festival," and a very conspicuous form of consumption. The potlatch is also the place of "being satiated:" one feels rich enough to give up hoarding, to give away. A potlatch cannot take place without the sense that one is overrich. It does not emerge from an economics of scarcity. Marshall Sahlins's Stone Age Economics of 1972 is, in addition to a study of gift economics, a critique of the economics of scarcity. Scarcity is the "judgement decreed by our economy" and the "axiom of our economics." Sahlins's and other's research has revealed that "subsistence" was never the central problem for humanity, except for underpriveleged classes within the developed market and now industrial and even post industrial cultures . Poverty, is as Sahlins says, an invention of civilization, of urban development. The sentence to a"life of hard labor" is an artifact of industrialism. The mere "subsistence scrabblers" of the past had more leisure time, time for ceremony, for play, and were afforded a greater nonchalance than the vast majority of the populations of our era.

But Sahlins's critique of the economics of scarcity needs to be distinguished from the "long boom," the specious celebrations of some kind of information or network economy that miraculously will save us from scarcity and failure. The ethnographic descriptions of communal consumption of surplus in ritual splendour are more a rebuke of the failures of progress to really deliver the goods than they are the description of something that is coming into being in the information age. In a similar way the gift-giving amongst an elite of programmers is more a model of how collaborative and distributed projects can both create wonderful results and forge strong ties within a networked economy than some representation of the successes of the information age as a whole.

What further developments can the free software projects achieve? A possiblity and a barrier was pointed out by Stallman at the Open Source conference. We lack good open source or free documentation projects for free software. Free documentation of high quality for free software is extremely important because free software develops rapidly and it needs timely and well crafted documentation. Tim O'Reilly has said that he already tried to copyleft a book on Linux but that it failed to sell. Perhaps it is time he tried a free documentation experiment again. The market is much bigger than it was even a couple of years ago. O'Reilly points out that writers don't want to copyleft their books as much developers want to participate in free software projects.

The phenomenon of free software is probably bigger than anyone of us realizes. We can't really measure it because all the ways of tracking these kind of phenomena are economic, and the "small footprint" operating systems, Linux and Free BSD, are passing along the much more numerous and impossible to track lines of those Eric Raymond called the "broke," those just like the people who built them. There are a few hints though. In August, broke the record for the largest ftp download of software for a single day, surpassing the previous record which was set by Microsoft for one of its Windows releases. All of's software is free and open source. also reports that much of the download is to points outside of the United States and the EU.

What will be the social and economic effects of free and open source computing? Do the successful collaborative free software projects prefigure other kinds of collaborative projects? Will the hau or the spirit of the gift of free software spread into other areas of social and intellectual life? I hope so. There is a connection between the explosion in the use of networked computing and the recent rise to prominence of free software. And this connection may prove to foster new forms of community and free collaboration on scales previously unimagined, but it certainly won't happen by itself. It will take the concerted efforts of many individual wills and the questioning of many assumptions about the success and quality of the collaborative, the open, and the freely given.