What is the point of free and open source software?

  • July 25, 2022
  • Erik Cameron
  • 10 min read

At the risk of a big opening: Free and open source software (FOSS hereafter) is utterly pervasive. It is a global, decentralized public resource responsible for astounding economic output. It presents as ideas and ideals, but many (most?) of us have built careers on top of it, and things don't get much more concrete than that. This is why it's surprising people seem to disagree so much about what the point of it is. Is it a failure if FOSS's ideals miss the forest for the trees in the face of new technologies and markets? [Amr 2017] Or is it a mortal wound for the integrity of the development process to be compromised? [O'Meara 2017] Would we consider the movement healthy if it compensated the people who keep the internet running for free, [Xe 2021] or is any arrangement for which corporate profit is an input dead on the vine? [Horn 2020 and Baer 2021], from two very different angles!] Is its moral rot systematic and philosophical [Goodman-Wilson 2019], or is it just plain meanness [McAllister 2014]?

The real reason this post exists is a genre of thinkpiece devoted to FOSS' vital signs; especially whether it is broken, sick, dying, dead. There are no shortage of examples:

If you're looking for a complementary list of arguments that FOSS is alive, functional, critical but stable, thinking of going for a walk: They probably exist, but I didn't bring any because in fact, many of the pieces above conclude reports of its death were greatly exaggerated. Besides, my concern isn't to triage FOSS (well, maybe a little) but to consider the wide variation in the perspectives above as to what life or health even means in this context.

Many of you have already distinguished "free software" as an idealist movement from "open source" as a production methodology, and (quite rightly!) objected to treating them as the same thing.

In material terms—historically, socially, economically, technologically—they are impossible to separate. For all practical purposes, they function as two wings of a single coalition. Second, even put together, they don't put us in a position to evaluate FOSS as a real, incarnate thing, because neither explicitly deal with its success or failure from the perspective of labor, which is central to much of the contemporary debate. You can't really talk about one without the other, and even if you talk about both, there's more to it. (For a bracing analysis of the differences, see Melody Horn's widely-read 2020 piece "Post-Open Source", discussed below.)

That said, the conditions under which they might be considered successful or unsuccessful could vary wildly, so we'll consider them individually here. We'll start with free software and come back to open source in Part Two.

Free software

The term "free software" was coined by Richard Stallman at the Free Software Foundation (FSF) in 1986 to refer to software that satisfies certain kinds of individual liberty; specifically, the abilities to copy, analyze, change and distribute source code. As of 2022, according to FSF "[t]o use free software is to make a political and ethical choice to assert the right to learn and share what we learn with others." That is to say, to FSF, the point of free software is political and ethical, not technical. (And the ethics/politics themselves relate to knowledge, rather than labor.)

One illustration of this came in 2015 when Stallman decided he would rather prevent the Gnu Compiler Collection from supporting modern IDE features like symbol completion, than allow GCC front ends to be paired with free-but-not-copyleft backends like LLVM and Clang"Since LLVM and Clang are not copylefted, they invite nonfree extensions. They are a gaping hole in the defensive wall around our city." Allowing full access to GCC's internals would have obvious technical benefits, and as far as I know Stallman advanced no technical argument against the proposal. Disallowing it put the project at a significant disadvantage at a time when its influence was already waning. The message was clear: GCC exists to serve FSF, not the other way around.

We must note Stallman himself is a polarizing figure to say the least. The extreme position above was hardly unanimous among GCC developers. More recently, Stallman's personal conduct has seen him resign from MIT and FSF, though in 2021 he rejoined the former, resulting in an open letter signed (as of 6/2022) by over 60 organizations and 3000 individuals calling not only for his removal, but for that of the entire FSF board for taking him back. The letter would be remarkable in any case (the list of institutional co-signs contains august FOSS names like Mozilla, Tor and X.org, for example) but it also speaks directly to our question, i.e., what the point of FOSS is, and what constitutes success or failure:

"We, the undersigned, believe in the necessity of digital autonomy and the powerful role user freedom plays in protecting our fundamental human rights. In order to realize the promise of everything software freedom makes possible, there must be radical change within the community. We believe in a present and a future where all technology empowers – not oppresses – people. We know that this is only possible in a world where technology is built to pay respect to our rights at its most foundational levels."(emphasis mine)

Whether you agree or disagree, it's a clear philosophical and practical statement: The undersigned believe that fulfilling Stallman's own vision ultimately requires ideals beyond individual software liberties. (And when you when you look at how much GNU involvement the list of signees represents, it's clear the call is coming from inside the house.) At the very least it says the free software ethos can't be reduced to a single person or understanding, but rather represents a large and diverse bunch indeed.

With that in mind, how has the ideal fared? In 2022, free software is ubiquitous, but so is proprietary data, and data rights (especially the right to privacy) currently dominate the conversation about digital freedoms. Add to that the fact that most of that empire of proprietary data is built on free software, and the waters muddy further. Unscientifically, you might ask yourself, just how free do you feel, digitally speaking?

From a consumer perspective, [Tarek Amr] notes that the presumptions of the free software model were upended by the twin forces of cloud computing and hardware bundled with services. "Free software" per se presumably refers to my rights over my own computers; FSF is not advocating I have the freedom to view and modify what you're running. But that's exactly the relationship of most users to cloud services—I may own my data, but I effectively contract with Google to keep it safe for me, which makes it hard to see how I could have freedoms to the software managing that data. Furthermore, even if I did (and much of it is of course free software) the software itself is insufficient to make the final product "free," as in speech or beer, because the product is built on a network of proprietary data. So you don't have control over the data you own, and the things you want to do are dependent on data other people own.

On a related note from [Melody Horn], "most of the people using Linux right now are using it by accident, distributed as ChromeOS or Android, neither of which is free software. so Linux is a win for the free software movement but a useless one." Horn believes the aggressive posture of FSF/Stallman both drove away free software adherents and commercial users. The Linux project itself declined to update to the more-stringent version 3 of the GNU Public License, and, Horn claims, the risk of legal action encouraged commercial interests to back competing tools with more permissive licenses, like LLVM and Apple's accompanying Clang, which were already "eat[ing] GCC's lunch" by the time Stallman forbade their integration in 2015.

Both Amr and Horn declare free software dead, but for wholly different reasons:

"[M]ost FOSS advocates ... have already missed the new battlefield’s location. ... It’s perfectly fine to admit that Free and Open Source Software ideas are dead, since the compute environments and legal frameworks they were created in are also gone." [Amr 2017]

"and LLVM became at least as good as GCC, and a less risky decision for big companies, and easier to use to build new languages. so the free software movement's last technical advantage was gone. its social advantages also kinda went up in flames with the GPLv3, too: the software that was the foundation for the GPL enforcement lawsuits stuck with the GPLv2. ... the free software movement, in the end, burned itself out, by fighting for a tiny crumb of success and then turning around and lighting that success on fire. the death of free software tells us that we can't use a license to trick corporations into sharing our values..." [Horn 2020]

For Amr, FOSS has lost by winning on the wrong battlefield. He acknowledges the distinction between free software and open source but doesn't feel it material to his point, because to him they are conjoined, and taken together, winning at the wrong things. In contrast, Horn believes free software is dead and open source is not (more on the latter in Part 2) because free software became technologically and legally irrelevant; i.e., ze takes the things open source is winning at to be relevant in a way Amr does not.

What do they agree on? Primarily:

  • free software (individually or as a part of FOSS) is a community enterprise that has to serve the needs of the many in order to be alive or successful
  • the current state of affairs does not serve those needs
  • "free software" per se is focused on anachronistic problems
  • large commercial interests are the bad guys

For Horn, it's that while free software built some political capital, it was too "misguided and quixotic" to prevent commercial interests from destroying it, which they did by creating a seedless version ("open source") incapable of spreading further freedom. For Amr, it's vertical integration, and the fact that content has replaced code as the hostage of proprietary licensing. To that, I'd add that the mass harvest and monetization of personal data makes it difficult to think about digital freedom in terms of software freedom, especially when free software has not only failed to prevent that sort of abuse, but materially abetted it to the tune of untold sums of wealth.

We all also tend to read "free software" a bit broadly when laying down criteria, but narrowly when applying them. Amr's reading implies a projection of FOSS values (whatever those might be) into the world of content licensing and distribution deals. Horn deems free software a movement of "ideological diehards" focused on "esoteric" individual liberties, but (by way of hir quote of Kat Marchán) FOSS [is] unsuccesful due to failure as a "community endeavor." My own rejoinder (and yours, if you agreed with me) ironically implies FOSS has failed for not helping us keep things private, when free software's stated mission is enforced openness.

Maybe one or more of us is wrong. Maybe the values of the free software movement have nothing to say about consumer lock-in, sustainable community practices or personal privacy. It's not obvious how the FSF's stated goals—the right to learn and share what we learn—entails any of those things, at least until you start imagining how those rights would be asserted in the real world, at which point they all seem (to me) in play. In any case, the sheer volume and variety of things people expect from free software is itself a decent argument that it's alive and well.

If this has all seemed a bit soft and subjective to you, I encourage you (a) indulge in the soft and subjective once in a while, and (b) come back for Part 2 on Open Source, where we will get down to brass tacks about the economics and engineering behind FOSS.

(Part 2: Open Source)