The Politics of Open Infrastructures - 2. From Free Software to Open Source
2. From Free Software to Open Source: Traversing the Terminologies, Values and Ethics of an Open Knowledge Infrastructure
Renée Ridgway
©2026 Renée Ridgway, CC BY-NC 4.0 https://doi.org/10.11647/OBP.0528.02
If, as Langdon Winner famously states, the politics of technology are about ‘ways of building order in our world,’ then hacker and geek politics are geared toward reordering the technologies and infrastructures that have become part of the fabric of everyday life ( Coleman 2011: 515).
1. Introduction
The coinage Free Software goes back to the 1980s where it was once valued for its ‘peculiar form of potentiality,’ not per se a thing or technology or license, but something that had possibility (Kelty 2013). 1 At that time, the ethics of Free Software, which is ‘open’ and not proprietary, were immanent to nascent computer science departments worldwide and people came to Free Software as the solution, enacting ‘disruption in the creation, circulation, distribution and control of knowledge’ (Kelty 2013). Simultaneously Free Software embodied a politics of liberalism, with its roots anchored in UNIX, situated within the hybrid academic-corporate culture of Bell Labs, which were critical of software development by corporations, organisations or consultancies (Kelty 2008). Eventually, this politics took form in the acronym FLOSS ( Free/Libre and Open-Source Software), articulated by Richard Stallman, who opposed proprietary software ‘hoarding’ (Kelty 2008: 99). Here, the ‘L’ signified ‘libre,’ meaning ‘free,’ not as in ‘free beer’ but ‘free speech,’ standing for the ‘free flows of information’ and taking no position on economic flows (Schneider 2021: 47). Stallman’s conception of Free Software was a pushback against the ‘piracy problem’ that Microsoft’s Bill Gates had already raised with his infamous ‘Open Letter to Hobbyists’ (1976) that complained about the sharing of software through hobbyist magazines.
In 1991, a Finnish computer science student (Linus Thorvalds) developed and maintained a ‘small kernel that could run alongside free software tools’ for his personal computer at home that came to be known as Linux (Driskoll 2015: 274). This operating system eventually se... Simultaneously, Raymond focused on the rise of the internet economy and making Free Software ‘into something that made more sense to investors, venture capitalists, and the stock-buying public’ (Kelty 2008: 109).
At the same time, the Free Software movement became affected by economic factors, namely the bursting of the dot-com bubble, and it morphed into many things: ‘a problem, a danger, a job, a calling, a dogma, a solution, a philosophy, a liberat... One of Kelty’s contribution was to elucidate that by 1998–1999, on the one hand ‘ Free Software’ was about rights and liberty whilst open source was concerned with ‘political agnosticism or nonideological commitments to “fun,”’ yet the practice is identical, which is an advantage as ‘they are all nonetheless contributing to the same mysterious thing’ (Kelty 2008: 112).
Twenty-five years later, this chapter enquires whether these practices and politics still ring true, along with how and when F(L)OSS and open source could be considered ‘open (knowledge) infrastructures’ and what types of ethics or values are (still) embedded within them. It unfolds by first examining the historical transformations and semanti... A critical discourse analysis interweaves excerpts on empirical data gathered on ethics and values from semi-structured interviews structured by Star and Ruhleder’s ‘dimensions’ (1996) or Star’s ‘properties’ (1999) of an ‘emerging infrastructure.’ By elucidating the techno-infrastructure through the... ideology, showing how these approaches differ on an ethical, rather than technical, level. In this way, the Free Software ‘movement’ of past decades resurfaces through the makers/practitioners of today, who, according to the author, could be considered a ‘ recursive public’ (Kelty 2008). The chapter contributes to critical infrastructure studies and STS literature by demonstrating that F(L)OSS is an open (knowledge) infrastructure when it is open, political, and collectivised, not only through open-source code but through the people who enable the infrastructure to exist, enacting maintenance and care as a commons of potential.
2. The Politics (or Lack Thereof?) of FOSS Initiatives
As explained above, the social and technological practices of F /OSS exhibit certain elements of the liberal tradition, where the emphasis came to be placed on ‘knowledge, self-cultivation and self-expression as their ethical philosophy’ ( Coleman 2013: 3). However, it was through cultural positioning and the ‘rearticulation of free speech principles’ that orchestrated the dynamics of what Gabrielle Coleman describes as the ‘political agnosticism’ of FOSS (2004: 509). This denial of formal politics or lack of political intention ignited a way to reconsider intellectual property rights by directing it ‘toward the protection of free speech, instead of its “conventional use” of securing property rights’ ( Coleman 2004: 508). For many Free Software enthusiasts, code is considered speech and the ‘L’ (libre) signifies the right to access and create free information flows. Therefore ‘free speech’ was seen as ‘free code,’ and as McInerney notes, the ‘source code,’ which is the recipe for software, invites users to modify it (2009: 209). Composed of people, ‘skilled programmers, security researchers, hardware builders and system-administrators,’ FOSS opened up a window on to the liberal values of free speech, though a ‘cultural-based technical practice: that of computer hacking’ ( Coleman 2004: 508).
In this way, FOSS was informed by computer techniques that constitute hacker values. Free Software programmers took pleasure in building and configuring whilst communicating and collaborating as ‘forms of value by inhabiting technology’ ( Coleman 2011: 512). Other values that came into being through Free Software included ‘competitive mutual aid, avid free speech principles, and implementation of meritocracy’ all while challenging the intellectual property establishment ( Coleman 2013: 3). The power of Free Software was to make these values, or principles, into material objects and thereby able to be ‘manipulated, reconfigured, tested and torqued’––radically open to change and promoting process over product as an infrastructural and material strategy for a new world (Kelty 2013). Creating alternative software licensing and establishing hacker values motivated their ‘commitments to freedom of information, geeky production, culture, and action,’ along with instantiating new ways for ‘collaborating, organizing, and protesting’ ( Coleman 2011: 513, 515).
As shown by Ben Birkinbine’s research (2020), the revolutionary changes ignited by FLOSS and other commons-based peer production were co-opted by these corporate structures and strategies. Although Free Software was an amazing human achievement, with Liberty removed from the equation, in Raymond’s ideal ( open source), the innovative nature of software production was through the ‘development model’ that would lead to mass revenues through the stock market (Kelty 2008: 109). In his 2015 article, Birkinbine focuses on the acquisition of Sun Microsystems by Oracle and shows how corporations became increasingly involved in FLOSS projects (and nowadays even more with the development of GenAI), yet these clashed with the community ‘ethos’ of FLOSS. Coleman articulates the ethos as a ‘belief in productive freedom’ that includes the right to ‘autonomously improve upon their peer’s work, refine their technical skills, and extend craftlike engineering traditions’ (2013: 3), without a corporate infrastructure dictating the terms and licenses. Giacomo Poderi’s ethnographical study of a FOSS videogame community reflects this positioning by articulating the ‘relationship between continuous design and software development distributed infrastructure’ (2014: 78). Echoing Coleman’s ‘belief in productive freedom,’ Poderi emphasises the ‘user-driven and self-organizing construction processes’ of the FOSS commons through ‘collective and bottom-up management,’ drawing on Star (1999) and Star and Ruhleder’s (1996) ecological approach to infrastructure (2014: 80).
3. When Is an Infrastructure a Complex and Relational Ecology?
In her seminal text, ‘The Ethnography of Infrastructure,’ Susan Leigh Star proposed the ecological as a ‘methodological perversion,’ putting forth those people and things which had been neglected with the studying of ‘boring’ things, such as infrastructure (1999: 379).
Star insightfully argued that infrastructures are not only ecological, but also relational and material constructs made by humans (1999: 382). Star draws on Gregory Bateson, who claimed that ‘What can be studied is always a relationship or an infinite regress of relationships. Never a “thing”’ (1978: 249), to show how infrastructures are impacted by other things. With technical systems that serve as larger infrastructures, it is often not the infrastructure as a whole that becomes visible, but the specific component that matters to a given user: the rail to the engineer; the stairs that become barriers for wheelchair users; the water system that is essential to a cook; or, for a plumber, the very same water system, understood as a ‘target for repair’ (Star 1999: 380). Often seen as the invisible infrastructure hovering in the background supporting other forms of work, Star uses the example of computers to comically state that if we were to conceive of them ‘more modestly as symbolic sewers’ instead of information highways, ‘this realm would open up a bit’ (1999: 379). Because for Star, if an information system is investigated without also exploring its ‘standards, wires, and settings,’ the researcher can overlook the ‘equally essential aspects of aesthetics, justice, and change’ (1999: 379).
During the mid-1990s, with the rise of decentralised communication technologies that encompassed large swaths of land, there was a growing need for shared standards as well as ‘situated, tailorable and flexible technologies’ ( Star and Ruhleder 1996: 112). In... Worm software was designed and constructed for biologists with an intention to ‘create infrastructural tools for research’ and to ‘transform local laboratory organization,’ whilst simultaneously competing with the expansion of finding information on the internet with search browsers (Gopher and Mosaic) ( Star and Ruhleder 1996: 112). Additionally, Star and Ruhleder built upon the scholarship of Yrjö Engeström, who asked ‘When is a tool?’ to demonstrate that a thing becomes a tool in practice and when it is connected to an activity (1990). During their study, Star and Ruhleder subsequently enquired ‘When, then, is an infrastructure?’ (1996: 113).
In order to map out the ‘emergence of infrastructure,’ Star and Ruhleder came up with a list of ‘dimensions,’ which are without an ‘absolute boundary on an a priori definition’ (1996: 112–113)—what Star later called ‘properties’ (1999). Star and Ruhleder’s findings showed that not many biologists ended up using Worm, but also that infrastructure does not just reside in the background but ‘emerges for people in practice, connected to activities and structures’––it is a relational concept. In her 1999 article, Star reflects back on this research, where she imparts that even though they ( Star and Ruhleder 1996) were following certain principles of participatory design, they began to ‘see infrastructure as part of human organisation, and as a problematic as any other’ (1999: 380). In the following sections, I apply these dimensions or properties to structure interviews (epistemological encounters) with a range of software developers, geeks, and hackers to find out if and when F(L)OSS emerges as an ‘ open infrastructure.’
4. Epistemological Encounters
Infrastructures are not only complex and relational ecologies. The system’s development process is also about gathering communities together who have very diverse approaches, for example regarding the values and ethics of building computing infrastructure and software. This chapter therefore draws on fourteen semi-structured interviews conducted both face-to-face and virtually from December 2023 to September 2024, each lasting approximately 1.5 hours. It also consists of historical analysis, tracing categories of infrastructures, delving into the development of technologies and reading literature (archival texts, articles, and books). During this period, these interviews were augmented by participant observations—informal and ad hoc conversations at various F(L)OSS conferences through introductions by previous ‘informants’ that adhere to the ‘snowball’ method. As with Judy Wajcman’s study on calendars in Silicon Valley, the research is located within the ‘broad ecology and cultural life’ (2018: 1280), in this case people living in F(L)OSS and open-source development communities. Also taken into consideration is that people ‘make meanings based on their circumstances’ and world experiences that are also ‘inscribed into their judgments’ concerning technology and the ‘built information environment’ (Star 1999: 383).
This type of fieldwork could be considered a ‘distinctive mode of epistemological encounter’ that offers a means to reflect on the process during and after the ‘material labor of fieldwork,’ which occurred whilst engaging with the ‘material labor of writing and rewriting’ (Kelty 2008: 18, 20–21). In the beginning, the research was about the values and ethics of Free Software as ‘a kind of complex technical practice’ (Kelty 2008: 21). Echoing Kelty, who ‘could not have identified “ recursive publics” as the object of the ethnography at the outset’ (2008: 18, 20–21), during the coding of the data and writing this chapter, quite unexpectedly it came to mind whether theses informants could also be considered a ‘ recursive public,’ which this chapter will return to below. Therefore, as with much research writing, the reflection and writing process enables one to ‘reorient it into a question that will have looked deliberate and answerable in hindsight’ (Kelty 2008: 20–21).
Furthermore, when choosing informants the aim was to be inclusive of professions (designer, software engineer, academic, non-profit director), gender (non-binary, women, men), continents (South American, North American, European), ethnicities (Brazilian, Indian, Finnish, Belgian, American, German, British, French) and financial interests (grant awardee, freelancer, self-employed, employee). The informants are anonymised, as agreed upon before the interview took place, and assigned a number. After conducting the epistemic encounters with the informants, the data was then ‘coded,’ based on what Star and Ruhleder (1996) define as the ‘steps toward an ecology of infrastructure’ by way of the specific dimensions/properties with small additions or adaptions (1999). Although they were not defined as values as such, (re)reading these dimensions or properties brought to mind many of the ethical considerations that the informants mentioned and that structure the content of the interviews.