Keeping the Open in Open Source
While peer production is now well-known after the pioneering work of Yochai Benkler and the evident success of open-source software, the governance of these communities is the subject of much less debate and study. Yet, evaluation of governance is crucial, as there can't really be any real peer production if the allocation of resources is not the result of the social relationships themselves, but driven by commercial interests and command. Indeed, many commercially driven and managed projects may well be parading as open source, but if they are managed in a classic way, the potential for social change is quite limited. George Dafermos from Delft University of Technology has been one of the few researchers specializing in the governance of free software communities and, thereby, has established more clear criteria for genuine peer production.
Michel Bauwens: You're well-known in the P2P Foundation community for your research of governance of free / libre and open source software (FOSS or FLOSS) projects. For those who are unfamiliar with your work, please tell us about your background and research.
My journey in the free open source software (FOSS) milieu spans more than a decade. As a researcher, I've participated in several studies of FOSS of which, perhaps, the best-known is Management and Virtual Decentralised Networks: The Linux Project, one of the first analyses of large FOSS projects from an organisational perspective. I've been involved in copyleft activism and have contributed to some projects as a consultant on FOSS licensing. In the last five years, however, my research work has centred on the organisation of the development of FreeBSD, one of the largest and oldest FOSS projects, and I'm currently preparing a PhD dissertation on the topic at Delft University of Technology. Like other social scientists, I consider FOSS to be an emerging paradigm for the organisation of collective activities and a laboratory of experimentation in structures for the decentralised development of technology.
Michel Bauwens: At the P2P Foundation, we distinguish peer production, peer governance, and peer property. Would you agree with this trilogy, and is it a fair assessment to say you are one of the few researchers specializing in the peer governance of peer production communities? Do you agree that there is a specific modality of governance which applies for peer production communities and how does peer governance relate to democracy?
The approach of the P2P Foundation is, without doubt, extremely useful in probing peer production, as it calls attention to key dimensions of this phenomenon. What sets peer production apart from other modes of production is precisely its mode of governance and property. To put it differently, what is special about this phenomenon is the way in which people participate in the production of a good and organise collectively their efforts, as is the way in which the goods that result from their activities are distributed. It is, therefore, obvious that peer production is characterised by a definite modality of governance, which consists of the self-selection of tasks by participants who make decisions collectively, based on consensus. And that, of course, has a striking resemblance to the concept of direct democracy.
As to whether I am one of the few researchers focusing on practices of peer governance, I'd say that, though researchers of peer governance are still few compared to those studying the governance of traditional organisations, peer production is no longer an exotic research field and the number of social scientists turning to peer production has been continuously rising over the last 10 years. No doubt, awareness of the promise of research in peer production is spreading in the academia.
Michel Bauwens: In Charles Leadbeater's We Think, he makes the very strong claim about the governance of open-source and online collaborative communities: "What they are not, ever, is egalitarian self-governing democracies." Do you think this is correct – why or why not – and how should we judge the peer governance practices of such communities? Can you tell us your views on the tension between equality and hierarchy in peer production? What do you think of the concept of benevolent dictator?
I haven't read Leadbeater's book, but similar arguments have been voiced by others. Usually, the thrust of the argument is that informal forms of organisation – such as those encountered in many peer production projects – are susceptible to the formation of administrative cliques, thereby breeding corruption and abuse of authority.
However, cliques can be found in all realms of life and, as many sociological studies have shown, the formation of cliques is endemic even in bureaucratic organisations despite all their written rules and formal roles. Most importantly, the problems that this critique attributes to the (often informal) form of organisation of FOSS projects are much more prevalent and destructive in formal organisations. Formal 'democratic' organisations like trade unions and political parties, even those that proclaim their egalitarian ideals, to say nothing of governments elected through general suffrage, are a hotbed of corruption and arbitrariness in the exercise of administrative authority. This should not be construed as implying that all forms of collective organisation are doomed to degenerate into corruption and abuse of authority.
The desire to realise democracy is not futile. Rather, the problem is that real democracy, that is, that mode of governance which is characterised by the unmediated participation of all community members in the process of formulating problems and negotiating decisions, is unattainable once a group is split into a fraction that decides and commands and another that obeys. Such structures make a travesty of the notion of democracy. It is from this vantage point that we must gauge how democratic peer production communities are. We need to look at how decisions are being made: Is decision-making based on the direct participation of all community members or is it concentrated in the hands of some of them who rule over the rest in the name of some real or imaginary community? That is the key question. Looked at from this perspective, the governance of peer production communities is often far more democratic than that of traditional 'democratic' organisations.
The concept of the benevolent dictator is very interesting. And it illustrates well the tension between hierarchy and equality in peer production to which your question refers. Let it be allowed me to elucidate it with an example. The archetypical benevolent dictator is Linus Torvalds, founder and leader of a well-known FOSS project, Linux. Torvalds is the 'dictator' of Linux in the sense that he is the ultimate authority when it comes to deciding which code contributions will become part of the official Linux release. His authority is not exercised over Linux developers, but over their contributions. He cannot tell them what to do, how to do it, or when to do it. That is the reason why, as some studies have remarked (for example, Ruben van Wendel de Joode's PhD dissertation titled Understanding Open-Source Communities: An Organizational Perspective), his influence over the behaviour of individual developers is marginal.
In fact, for his decisions to be received as legitimate, they have to be consistent with the consensus of the opinions of participating developers as manifest on Linux mailing lists. It is not unusual for him to back down from a decision under the pressure of criticism from other developers. His position is based on the recognition of his fitness by the community of Linux developers and this type of authority is, therefore, constantly subject to withdrawal. His role is not that of a boss or a manager in the usual sense. In the final analysis, the direction of the project springs from the cumulative synthesis of modifications contributed by individual developers.
To recap, the tension between hierarchy and equality has to do with the degree of control exercised by project administrators over the changes and modifications contributed by the base of developers. This tension is stronger in projects which, like Linux, invest one developer (or a sub-group of developers) with the authority to accept or reject code contributions by the community of developers. Other large FOSS projects, however – FreeBSD, for example – allow their developers to make changes directly to the codebase, with no need to filter them through a gatekeeper. Consequently, in these projects the tension is remarkably weaker.
Michel Bauwens: Peer production involves communities of contributors, paid and unpaid; for-benefit associations such as the FLOSS Foundations; and participating companies. How do these three players relate to each other? Is there a point when the role of companies distorts the p2p and commons dynamics to such a degree that we can no longer speak of peer production?
Conceptually, peer production is a mode of production directed neither to market exchange, nor governed by corporate hierarchies. Hence, the motivation underlying peer production projects, as well as the orientation of production in them, is radically different from that characteristic of for-profit companies. Although many contributors to FOSS projects are professionals working in the IT industry – and some of them receive payment from their employer for their work on FOSS – as a general rule, participation in FOSS projects is voluntary and unwaged.
It makes sense, then, to view the relationship between those three organisational entities you mention in your question as follows: A peer production community produces a product, which, on account of being universally free to use, modify, and re-distribute, attracts firms which wish either to use it internally or to exploit it commercially in the market. To contribute to the sustained development of such a product, some of these companies give back to the community, for example, by donating money and equipment. Others offer prominent FOSS developers a paid job to keep on doing the same thing they were basically doing until then as volunteers. The role of FLOSS Foundations is often that of interfacing between FOSS projects and firms, providing a channel of communication that firms can use to maintain contact.
I think we can no longer speak of peer production when there is no peer governance or peer property. If tasks are not self-selected by participants, but assigned to them through a managerial hierarchy, if the managerial process is not collective, but based on the separation of those who make decisions from those who do the work, then what we are looking at is not peer production.
Nor can we speak of peer production outside the commons, that is, when the production process results in a product that is not freely available to all those who want to use it. Hence, companies may well contribute to the commons (for example, as many software companies do, by releasing software they have developed internally under FOSS licenses) or to peer production communities (in much the same way that some software companies contribute code to FOSS projects), but they cannot really manage the peer production process. Trying to do so is bound to elicit fierce opposition from participants. Even if a company launched such a project with the aim of distributing production requirements across the network, the attempt to exercise heavy-handed control over its development would undoubtedly deter outside participation, thereby destroying the peer production model.
Michel Bauwens: What do you see as the relation between peer production and the emerging sharing economy, which includes collaborative consumption, that is the main theme of Shareable? Are both expressions of something broader yet?
I don't think that collaborative consumption or the sharing economy is something new. Borrowing books from a library is a practice that spans millennia. And it is not only libraries. The concept of 'free-shops' is a very old one: people bring things they want to give away such as clothes and furniture, which others can take away for free. What I find very important is the wide spectrum of motivations underlying the sharing economy. To some people, sharing is a life principle, a moral value in itself. To others, it's merely convenient – what matters to them is to have access to some resource when they need it, rather than actually own it. This accounts for the rise and success of such commercial services as car sharing.
In Portland, car share cars can be found at designated lots, and get free and exclusive parking at several locations throughout the Central City. Photo credit: Lightpattern Productions. Used under Creative Commons license.
The fact that sociality is hard-coded into our genetic code is also a strong force behind collaborative consumption and the sharing economy. Watching a football game in an empty stadium is not fun, nor is pub-crawling alone. That is, essentially, why ideas such as book clubs and reading groups have been successful: by making book-reading a collective activity, they make it so much more fun and interesting. Much the same could be said of housing: Many people, especially younger ones, turn to collective cohabitation because they find living alone intolerable. Another common motivation is mutual aid by means of pooling resources, as exemplified by computer clubs in the 1980s and 1990s (such as the legendary Homebrew Computer Club) and hackerspaces (a.k.a. hacklabs and makerspaces) nowadays.
In my opinion, the sharing economy has been feeding on and catering to some fundamental human needs and desires since the very dawn of civilisation. That does not mean it is not related to peer production. On the contrary, peer production reinforces collaborative consumption and the sharing economy. It does so in two main ways. It creates and evolves technological infrastructures which transform the scope of increasingly more activities and the context in which they take place, thus turning them into peer activities.
To illustrate, consider how users of P2P file-sharing networks have redefined the consumption of cultural goods like music as an essentially peer activity. Most readily visible, however, is the way in which peer production promotes the sharing economy by means of enriching the commons. If we think of the commons as a shareable economic infrastructure, then we can see clearly how peer production liberates the sharing economy from constraints long imposed upon it by exclusive property regimes.
It is undeniably true that both the sharing economy and peer production are concrete expressions of an alternative path of economic development – and, more broadly, of an alternative model of social organisation – which enables new ways of looking at problems and solving them. The sharing economy is inseparable from a mode of association of humans in the process of exchange and consumption that promotes mutual aid and sociality, while peer production incarnates a mode of association of humans in the production process characterised by the spontaneity and informality of mutual friendships and communal relationships. In terms consonant with classical political philosophy, we could say they are both expressions of a negation of the dominant form of social relations in the present.
Michel Bauwens: Please also give details about your insights into the similarities and differences amongst the FLOSS Foundations. Now that open hardware is developing, to what degree is the logic of peer production in software also moving to the field of open hardware? Do you see the same social dynamics and institutions emerging, or are there differences? Are there also open hardware foundations with a similar role?
Most large FOSS projects set up foundations to handle administrative tasks that are not directly related to the product development process. For example, the FreeBSD Foundation organises conferences and developer events and is responsible for the financial (e.g. fund-raising) and legal aspects of the FreeBSD project, as well as for interfacing with corporate vendors. Other FOSS foundations have a somewhat greater role in project governance: the Apache Software Foundation (ASF) is indicative of a FOSS foundation that is used as a meta-governance layer for the nearly 100 different projects developed under the aegis of Apache. That means it does not try to influence the development process of these projects, but to ensure they converge on some fundamental organisational and legal aspects of the wider Apache project. The function of the Free Software Foundation is even more extensive, including also the organisation of activist campaigns for the rights of technology users. I am not aware of any open hardware foundations with a similar role, though it is possible that some may exist.
Although we cannot foreclose the exact organisational form of open hardware projects in the future, I would expect similar institutions to emerge in that field, as the normative principles underlying its development are much the same. Institutions grow out of principles and operate to continue them. And, so far, the principles permeating the development of open hardware are indistinguishable from those animating the development of FOSS. Like FOSS, the development of open hardware exemplifies a mode of production that is directed neither to market exchange, nor governed by managerial hierarchies. Thus, like FOSS, the community of open hardware projects is quite likely to evolve institutions designed to safeguard their autonomy from managerial hierarchies and prevent their commercial co-optation. The main difference between FOSS and open hardware pertains rather to the ease with which digital products like software can be modified and re-distributed, as compared to physical products. For the time being, this factor prevents the adoption of the distributed development model of FOSS beyond the stage of design in the production of products that cannot be digitised. However, the combined effect of the diffusion of advanced reproduction technology through Fab Labs and the rapid development of desktop manufacturing systems like Rep Rap which aim to eliminate the need for large-scale, expensive industrial infrastructures, may eventually make it as easy and cheap to hack a wheel as any piece of software.
Neal Gorenflo: What might geographic communities aspiring for a peer economy learn from the communities building software through peer production? What obstacles do you see in translating an online economic mode to an offline context like a town or city?
Peer production communities teach us that a different way to organise human labour – and human activities, more broadly – is possible. Peer production organises enthusiasm, not obedience. It shows that managers and supervisors are not necessary and that collective activities, even those involving the participation of hundreds of persons, can be coordinated in non-hierarchical frameworks. In fact, there is hardly a principle of traditional business management which is not contradicted by the working existence of peer production projects.
To get but an idea of the difficulties that the application of the peer production model to the context of a city entails, we need to take a look at the conditions making that practice possible. First, peer production lies outside the realm of necessary labour: The livelihood of the vast majority of FOSS developers does not depend on their FOSS work as they participate voluntarily in their free time. Second, the means of peer production are massively distributed: The cost of buying a computer compared to that of setting up a car factory is trivial. The instruments and machinery used in most industries are far more expensive and, therefore, inaccessible to most people. Now, it is obvious how difficult it is for these conditions to generalise across the entire population of a town or city. Regardless of their pretensions, our societies are based on forced labour; most people work jobs they do not like just for the sake of money, which they need to feed, shelter, and clothe themselves. So, instead of experiencing work as a creative activity, they perceive the institution of work as a necessity, not as a free choice.
By contrast, peer production thrives on volunteer participation. This suggests that implementing the peer production model in the context of a city requires a radical change of mentality and a wholly new logic of organisation, something akin to what some theorists and researchers of peer production call 'hacker ethic' – an ethic that celebrates creative activity as an end in itself and embraces the principles of self-determination and self-expression in work as fundamental to the organisation of human activities. In short, a peer-produced city is inconceivable without a diffuse Do-It-Yourself ethic; it can only result from a situation in which people make things because they want to use them and not because they are told or paid to do so. This may sound unrealistic, but it has been attempted time and again with great success, though in a scale somewhat smaller than that of a modern metropolis.
Consider, for example, the vibrant boroughs of Kreuzberg in central Berlin and Christiania in Copenhagen. Whereas both areas were largely left to decay by municipal authorities, some people took it upon themselves to infuse them with a new lease of life. Empty and dilapidated buildings were squatted and renovated and so made habitable again. All sorts of infrastructures – from electricity and water down to community services catering to those less able to look after themselves – were bootstrapped from the ground up without any assistance from the local authorities. So, for example, if a house was in need of renovation, instead of calling a company, squatters enlisted the help of their friends and neighbours and did it themselves. It is only through such a generalised practice of social self-management and mutual aid that a peer-produced city can be created.
The difference between this paradigm of bottom-up organisation and the dominant model of social organisation is massive. The same could be said with respect to the mode of property in peer production and that which prevails in contemporary societies. The creation of a peer-produced city presupposes that technical infrastructures are freely available to all city dwellers who want to use them and necessitates a shift away from exclusive property regimes towards structures based on collective ownership and homesteading. Given that the institution of a peer-produced society implies such a radical change of power, it is to be expected, of course, that attempts in this direction will meet strong political resistance from vested interests. That much is certain. Yet, despite the political obstacles that check its flight toward universality, peer production evinces the dynamism of an idea whose time has come. Whereas a hundred years ago the offline application of peer production was essentially limited to the case of some professional and scientific associations, the thrust towards its offline implementation is now being fed by a burgeoning spectrum of experiments.
Members of the Open Source Ecology movement building a Rep Rap 3D printer. Photo credit: Marcus Calabresus. Used under Creative Commons license.
Consider, for example, the far-reaching implications of such enterprises as Marcin Jakubowski's global village, which is but an attempt to build a peer-produced village in rural Missouri. In a nutshell, Jakubowski's global village is made up of a group of people who share the willingness and determination to create an autonomous and self-sufficient geographical community without waged labour and hierarchical organisations. So, the first thing they do is to buy a patch of land. Then they start producing all the materials and tools required to build the village itself – what Jakubowski calls the 'obal village construction set' which means everything from bricks and bulldozers to CNC Circuit Mills and 3D printers. Now, the point is that this prototype village has proven to be working very well. And its success has spawned a number of similar initiatives across Europe, thus turning the vision of such peer-produced villages into a concrete reality.
Neal Gorenflo: What initiatives or events would you point to that demonstrate that social production of software is influencing the larger culture? This is a theme explored by David Bollier in Viral Spiral, but I'm curious about the direct links you see from these communities to other arenas of life.
The influence of peer-produced software is so pervasive that is is nearly impossible not to notice its wider cultural impact. Peer-produced software was the main source of influence that shaped Jakubowski's aforementioned global village. It was an influence for the worldwide network of independent media centres known as Indymedia, which sprung to life in 1999 with the aim of helping counter-globalisation protesters coordinate their actions. Peer-produced software has revolutionised – via weblogs – the field of journalism. It has inspired collaborative efforts such as the Human Genome Project, which are reinventing the modus operandi of the pharmaceutical industry. It has motivated an open-source approach to the development of physical products, known as open design. It has prompted an ever-increasing number of companies to experiment with models of 'open innovation' and 'user innovation' thereby promoting the participation of end users in the product development process.
And there is more. Much more. The most popular encyclopedia of our time is Wikipedia, the user-created encyclopedia on the Internet. Increasingly more universities put educational material they use (like lecture notes and syllabi) online, making them freely available to anyone with an Internet connection. A well-known such experiment is MIT's OpenCourseWare. The same happens with scientific periodicals – papers published in so-called open-access journals are freely available to anyone, not just academics. Similar molecular processes are underway in the artistic realm: The Free Culture Movement campaigns for the free availability of cultural goods. All of these initiatives and projects are in some way influenced or inspired by the success of FOSS. They borrow some or all of its principles and apply them to a different context. The list is incredibly long. Of course, some ideas – like the Free Beer, whose recipe is freely available to encourage its users to modify it and redistribute it – may appear to have a lesser cultural impact than projects like Rep Rap, which aims to develop a 3D printer machine (about the size of a desktop computer) that will allow people to manufacture most of the artefacts used in daily life at their own homes. But when we look at them as a whole, we see clearly how profound the cultural impact of peer-produced software is.
In my opinion, the combined effect of all these projects is to incubate a culture of experimentation and Do-It-Yourself. Peer production is the harbinger of a new culture that encourages people to contemplate the idea of how a society based on the principles of peer production would be organised. We can already see this happening; concepts like that of open-source governance and open-source democracy (popularised by Douglas Rushkoff, among others) sound the death knell of traditional party politics by encapsulating the demand for broad and unmediated participation in political life.
Neal Gorenflo: What do you see for the future of free and open-source software? What are the threats internally and externally to its vitality and, also, the new conditions that might expand its production and use?
I am convinced that the open-source model will spread far beyond the software industry, thus revolutionising the scope of decentralised collaboration. That is not a prediction, for it's already happening. Among the forces giving this process a strong impetus is the refusal of regimented work environments and waged labour that the actions of an increasing number of people attest to. The image of the 'corporate man' is increasingly seen as repulsive by the young. So, they choose something else. Some are living off the welfare state (in countries like Sweden and the Netherlands where that is feasible); others are attempting some micro-enterpreneurial activity, whether in the formal or informal economy. But the underlying motivation of all is to escape work alienation. Their desire is to re-design the process and content of their work in accordance with their own dreams of flat hierarchies and tasks through which they can express themselves and develop their individual personality.
To a large extent, the development of FOSS feeds on these desires. If people participate in FOSS projects, it is because they find it creative, pleasant, and fulfilling. Students of FOSS like Pekka Himanen are definitely right in regarding the 'hacker ethic,' that is, the work ethic of FOSS developers as diametrically opposed to the dominant ethic of work under capitalism, the famous 'Protestant ethic.' The hacker ethic, in contrast to the Protestant ethic's emphasis on the disciplined obligation of work as a duty, exalts the joy and autonomy inherent in intrinsically motivated activities. Therefore, it is little surprise why it appeals to the growing number of people who have become disillusioned with the reality of work inside corporate hierarchies.
I don't think there is something in the dynamic of FOSS that undermines its vitality. The pivotal role of the FOSS community in the continued development of the Internet testifies to its internal sustainability and dynamism. Rather, the vitality of FOSS is threatened by external factors. Legislation which criminalises tinkering with technology suppresses the motive of experimentation, effectively restricting the scope of participation by end users and hobbyists. The entertainment industry, spearheaded by film and music companies, has been lobbying hard in this direction for the past 15 years and it is quite probable that consumer electronics will soon join this campaign. Draconian intellectual property regimes have the same effect: They discourage decentralised development and restrict the participation of end users and hobbyists. But though the scientific recognition of their harmful effect on innovation is widespread, it has so far been unable to influence the political and legislative process. In any case, the conflict between the community of FOSS projects and the cultural-industrial complex is only going to become more fierce over time and more varied in its manifestations. It will play out not only in the legislative arena, but also in the cultural and economic realm.
The Wikimedia table at the 2011 Open Source Conference in Prague. Photo credit: Packa. Used under Creative Commons license.
Neal Gorenflo: some claim that the FOSS community produces few original products, that this mode is good at creating free versions of products innovated privately. What is your view?
It is a common misconception, indeed. Some people reckon that FOSS technologies are but open re-implementations of functionality embedded in proprietary software, but they only look at this or that project and fail to see how diverse and rich the ecosystem of FOSS actually is. Let me illustrate with an example. Several academics with a social sciences background have told me the same. Why? Because they use their computers mostly to write papers and prepare presentations, lectures and so on. So, for them, 'computing' basically comes down to using Microsoft Office-like software. They know that Microsoft Office predated OpenOffice, so they conjecture that FOSS is but a re-implementation of already extant functionality. But that is, basically, a generalisation they make out of a limited number of cases.
The World Wide Web is the first-ever fully functioning implementation of a hypertext system (rather than just the notion of how such a system would work) and, perhaps, the greatest FOSS innovation, considering the impact of the web on our lives. Its core components were open source right from the beginning and its development thrived on the contributions of a loosely coupled community of fellow enthusiasts spread all over the world. It is not hard to find more examples, even well-known ones: It was Microsoft which emulated, and still emulates, the functionality offered by the Mozilla browser. (For instance, Mozilla had tabbed browsing years before Microsoft's Internet Explorer.) Gnutella and Freenet, as the first fully decentralised peer-to-peer networks, are yet another. Or BitTorrent, which has revolutionised file-sharing and is used at any given instant of time by more people than Facebook and YouTube combined. BIND, the most widely used DNS software, and Sendmail, which routes the majority of email, not only created but still dominate their respective fields. And when you look at software with a more narrow appeal, then you come to realise how rich and diverse FOSS is. Many FOSS tools used by IT professionals in their daily work still have no proprietary counterpart. To say nothing of what, in my opinion, is the greatest innovation of FOSS – the process through which it is developed, a mode of production that is radically different from that encountered in the proprietary software industry.
Neal Gorenflo: The interplay of FOSS to the market economy is complex and nuanced. What are the most common misconceptions about this relationship?
A rather common misconception is that FOSS is incompatible with the market economy, derived from the view that the free circulation of FOSS on the Internet strips it of value in a market for selling and buying software. Some people think there is no money to be made out of products which are freely shared. What they overlook, however, is that the market economy includes a major component of services and the service industry may well develop services atop a product that is freely available. For firms like IBM, which divested its former manufacturing facilities and evolved into a consulting firm, there is nothing disruptive in FOSS; customising FOSS to the needs of their corporate clients forms the basis of their business model. In a service economy, giving away a product for free makes sense if that serves the purpose of rolling out a host of profitable services. And, so, FOSS is supportive of software as a service industry. The same could be said of companies which use FOSS as a basis of developing their own software products. What they do is take functionality developed through peer production and then use it as a springboard for creating their own products. For instance, that is what many companies selling statistics software do.
The only misconception that is, perhaps, more widespread than the above one is that FOSS is perfectly harmonised with the market economy. The market economy cannot be sustained without a mode of production that is directed to market exchange. It is the consequence of producing for a market, rather than for immediate use. FOSS, by contrast, is the result of a mode of production geared to the satisfaction of the needs of the producers themselves; FOSS is made to be used, not to be sold. Contributors to FOSS projects develop software to 'scratch their own itch' and, by extension, make it freely available to others.
This practice forms the kernel of a sharing economy: Its principle is that what one produces over and above one's personal needs, s/he shares freely with others. That, of course, runs counter to the imperatives of the market economy for hoarding and exploitation of the surplus product. Although FOSS opens up some entrepreneurial opportunities, the continued expansion of the mode of production characteristic of FOSS – and of the sharing economy it sustains – is antithetical to the organising logic of the market economy and so undermines its foundations.