> Nevertheless, the revolutionary economics of FOSS are based on collaboration, and are incompatible with competition.
FOSS seems to mostly be about copying existing non-free work in some legal way. It only exists because it copies its competition.
As for economics . . . big players often back FOSS development but only in markets those companies aren't in yet and only insofar as the development directly benefits them.
If you want to compete with FOSS, just make something closed source that's successful. They will come.
You are partially right, but your characterization is overly broad. Yes, FOSS does often copy closed source, but when doing that it often innovates on top of it - not only in features, but in extensibility.
Many open source projects become so widely used and powerful, they become a reference in the field not only because they are free, but because of their quality or feature set when compared to closed source alternatives. Examples off the top of my head are ffmpeg, qemu and Blender.
> A lot of commercial software seems to mostly be about copying existing popular work in some legal way. It only exists because it copies its competition.
Fixed that for you. But not really.
Open source software exists because a community sees a use for it. Just like any other competitor, it competes on features, including price, support, and total cost of ownership. Free software doesn't necessarily mean no cost to deploy and support.
> The notion is that the software they build belongs to them. They possess a sense of ownership over the software, which comes with a set of moral and perhaps legal rights to the software, which, importantly, are withheld from any entity other than themselves. The “developers” enjoy this special relationship with the project – the “developers” being the special class of person entitled to this sense of ownership and the class to whom the up-and-coming source-available movements make an appeal, in the sense of “pay the developers” – and third-party entities who work on the source code are merely “contributors”, though they apply the same skills and labor to the project as the “developers” do.
That's... literally just proprietary software, with "source code available on request"?
That's every OSS product, though. It's like with Git - theoretically, the system is designed around a mesh topology, where there's an amorphous blob of clones exchanging patches, without any designated special repo; in practice, everyone pulls from and pushes to a designated "origin", creating a tree in which the root owns the canonical version of the source code. Same with any OSS project - there's always one designated organization or group that holds the moral, legal, and/or assumed right to project identity, which includes both the name and the source code.
Simple proof: fork any OSS project you like, make some changes, and start promoting it under the same name. See how long you'll last before first accusations of theft roll in.
OSS is a strange beast - the source code itself isn't something scarce you can make money off, but the identity of a project and associated reputation very much is. Which gives you the "first party"/"third party" developer split, regardless of whether the payment is in dollars or (with most OSS) just Internet clout.
In OSS projects the source code is available in general, not “on request”. On request means they can deny you the code, or provide you with a version that isn’t what they distribute.
Also, if you fork an OSS project and promote it under the same name you’d be (IMO rightfully) criticized, but if you fork and promote it under a different name, and clearly credit the original, I believe most people wouldn’t care. At worst your project gets no attention, or maybe the owner cares, but if your clone is superior people will move to it. This has precedent, albeit with dead projects, like tui-rs becoming ratatui; with projects that became closed-source, like MySQL; with small patches of big projects, like linux and firefox; and where the fork has a small audience. But active projects that aren't becoming closed-source probably tend not to have substantially different, popular forks, simply because the effort to maintain such a fork to par with the original is too much, and not adequately justified by the motive.
Moreover, it’s legal to fork an OSS project and start distributing it on your own. If the project is truly open-source and not copyleft, it’s even legal to close-source and sell your fork. It’s legal to directly copy parts of the open-source project into your competing project, with attribution, rather than having to “clean-room engineer” or find a different algorithm to avoid being accused of copyright. It’s legal to provide “third-party” support to the open-source project that is superior to the owner’s first-party support they are trying to monetize.
This seems to me the crux of DeValt’s argument: if you create an open-source project you don’t own the project identity. You own the name, but that doesn't mean much, because someone can fork the project with a similar name and everyone can informally know it as “new [your project’s name]”. You absolutely don’t own the code legally. And people will say and act as if you own the project, but only if you don’t abandon the project or try to make it closed-source; because it’s a lot more effort to hard-fork a project with an active maintainer, which will have no initial audience and constantly need to be synced or fall behind, especially when you can just create a pull request.
It’s true that most open-source projects have one repository and an “in-group” that are de-facto “owners”, but ultimately their ownership is conditional on people’s acceptance. If the people switch to a competitor who literally copies their code, the original owners have no power to stop them, as has happened to MySQL/MariaDB, Terraform/OpenTofu, and Redis/Redict. This is functionally different than legal ownership of “source available” code, which may be legally protected from derivation, and especially “open core” where there are parts of the code (blobs) you physically can’t modify or in some cases even use outside the main distribution.
"""
While open source was proven to be incredibly profitable and profoundly useful for the software industry as a whole, the economics of making open source work for one business are much different.
"""
"""
This logic [of "post-open"/"fair source"/"source available"] is rooted in a deeper notion of ownership ... The notion is that the software they build belongs to them. They possess a sense of ownership over the software, which comes with a set of moral and perhaps legal rights to the software, which, importantly, are withheld from any entity other than themselves. The “developers” enjoy this special relationship with the project ... and third-party entities who work on the source code are merely “contributors” ...
"""
"""
My well-wishes are contingent on any movements which aim to compete with open source stopping short of calling themselves open source.
"""
To me, this article is pointing out the intrinsic "cursed" problem of providing a startup mentality to an open source product. FOSS relies on giving freedom to replicate while (naive) startups are almost by definition looking for moats or monopolies to become unicorns.
There are ways of providing FOSS products, say as a by-product or ancillary project, but if a startup's main product is the source, it's almost antithetical to the business proposition. I think this point is only ever obliquely addressed and this article tackles it head on.
I also appreciate the articles focus on "open washing", where the new breed of "post-open" or "fair source" movements either engage in or stop just shy of trying to redefine the term "open source" to mean "source available" and not one that respects freedom.
It's pretty rare to see a post that I agree on so wholeheartedly.
The big question is though, how do you make business work in FOSS context. I'm glad that SourceHut is doing well, but I suspect it thrives partially due being small niche software. On the other end of scale you have FAANG style megacorps that do a lot of FOSS work too. But the middle-ground between these two extremes seems really difficult. Yet it feels desireable that non-niche FOSS development wouldn't end up being solely in the hands of megacorps.
I do see a connection here to how companies have been increasingly becoming hyperfocused on specializing what they perceive as their core domain, and outsourcing absolutely everything else.
It used to be that companies owned, and were proud of, their own headquarters, employed all the support functions personnel and so on. Contrasting janitors to CEOs is common trope, but who employs janitors anymore?
While this change has been economically incredibly profitable, I suspect there might be some long-term issues that crop up.
How does this relate to FOSS? In a sense software maintenance is lot like janitoring, and I do think one way of making the ecosystem healthier would be if companies would have larger in-house software divisions. So for example instead of the likes of Ford or Walmart paying to MS (or IBM) for stuff like Office suite, they would just have in-house teams working on LibreOffice.
> As for me, I still believe in open source, and even in the commercial potential of open source. It requires creativity and a clever business acumen to identify and exploit market opportunities within this collaborative framework. To win in open source you must embrace this collaboration and embrace the fact that you will share the commercial market for the software with other entities. If you’re up to that challenge, then let’s keep beating the open source drum together. If not, these new movements(source-available software) may be a home for you – but know that a lot of hard work still lies ahead of you in that path.
This article feels more like a critique of the new "source-available" trend in tech whereby existing OSS projects sell out and try to retroactively re-license themselves to be more closed-source friendly. I've been sort of aware of this phenomenon since there was a small upset when Muse Group bought Audacity, though it appears it is still GPL[1].
The real issue is that people just aren't thinking through the effects a license brings. They'll release things under open source licenses like the GPL because it's the 'expected' thing to do, or because they want others to be able to contribute to it, or what not, without thinking through what it allows others to do in turn (use it without paying, rerelease it themselves, fork the project, etc).
So a lot of companies start up thinking they can open source their code, and treat their project like it was under a proprietary license with source available, then realise the hard way that's not the case. And a lot of authors release stuff for free when they don't see the value of it, then realise years later that they could be getting paid for their efforts.
And the article is right that it's probably not going to be possible to outcompete open source with any new license or incentive structure. Once people are used to something being free to use without any limits, any less free alternative is going to severely struggle to gain traction.
> With the rise of GitHub and in particular the explosion of web development as an open platform, commercial stakeholders in software caught on to the compelling economics of open source.
It could also be that the grassroots of the dotcom era literally grew up and wanted to capitalize on their work, while still retaining their values. What we have today is something in-between closed source and open source. I.e. "commercial stakeholders" isn't a merged branch, but the linear evolution of the original one.
15 comments
[ 3.2 ms ] story [ 50.4 ms ] threadFOSS seems to mostly be about copying existing non-free work in some legal way. It only exists because it copies its competition.
As for economics . . . big players often back FOSS development but only in markets those companies aren't in yet and only insofar as the development directly benefits them.
If you want to compete with FOSS, just make something closed source that's successful. They will come.
Many open source projects become so widely used and powerful, they become a reference in the field not only because they are free, but because of their quality or feature set when compared to closed source alternatives. Examples off the top of my head are ffmpeg, qemu and Blender.
Fixed that for you. But not really.
Open source software exists because a community sees a use for it. Just like any other competitor, it competes on features, including price, support, and total cost of ownership. Free software doesn't necessarily mean no cost to deploy and support.
That's... literally just proprietary software, with "source code available on request"?
Source available is a type of proprietary software but a specialization thereof that DeVault is highlighting.
Simple proof: fork any OSS project you like, make some changes, and start promoting it under the same name. See how long you'll last before first accusations of theft roll in.
OSS is a strange beast - the source code itself isn't something scarce you can make money off, but the identity of a project and associated reputation very much is. Which gives you the "first party"/"third party" developer split, regardless of whether the payment is in dollars or (with most OSS) just Internet clout.
Also, if you fork an OSS project and promote it under the same name you’d be (IMO rightfully) criticized, but if you fork and promote it under a different name, and clearly credit the original, I believe most people wouldn’t care. At worst your project gets no attention, or maybe the owner cares, but if your clone is superior people will move to it. This has precedent, albeit with dead projects, like tui-rs becoming ratatui; with projects that became closed-source, like MySQL; with small patches of big projects, like linux and firefox; and where the fork has a small audience. But active projects that aren't becoming closed-source probably tend not to have substantially different, popular forks, simply because the effort to maintain such a fork to par with the original is too much, and not adequately justified by the motive.
Moreover, it’s legal to fork an OSS project and start distributing it on your own. If the project is truly open-source and not copyleft, it’s even legal to close-source and sell your fork. It’s legal to directly copy parts of the open-source project into your competing project, with attribution, rather than having to “clean-room engineer” or find a different algorithm to avoid being accused of copyright. It’s legal to provide “third-party” support to the open-source project that is superior to the owner’s first-party support they are trying to monetize.
This seems to me the crux of DeValt’s argument: if you create an open-source project you don’t own the project identity. You own the name, but that doesn't mean much, because someone can fork the project with a similar name and everyone can informally know it as “new [your project’s name]”. You absolutely don’t own the code legally. And people will say and act as if you own the project, but only if you don’t abandon the project or try to make it closed-source; because it’s a lot more effort to hard-fork a project with an active maintainer, which will have no initial audience and constantly need to be synced or fall behind, especially when you can just create a pull request.
It’s true that most open-source projects have one repository and an “in-group” that are de-facto “owners”, but ultimately their ownership is conditional on people’s acceptance. If the people switch to a competitor who literally copies their code, the original owners have no power to stop them, as has happened to MySQL/MariaDB, Terraform/OpenTofu, and Redis/Redict. This is functionally different than legal ownership of “source available” code, which may be legally protected from derivation, and especially “open core” where there are parts of the code (blobs) you physically can’t modify or in some cases even use outside the main distribution.
Copyleft licenses are "truly open source".
""" While open source was proven to be incredibly profitable and profoundly useful for the software industry as a whole, the economics of making open source work for one business are much different. """
""" This logic [of "post-open"/"fair source"/"source available"] is rooted in a deeper notion of ownership ... The notion is that the software they build belongs to them. They possess a sense of ownership over the software, which comes with a set of moral and perhaps legal rights to the software, which, importantly, are withheld from any entity other than themselves. The “developers” enjoy this special relationship with the project ... and third-party entities who work on the source code are merely “contributors” ... """
""" My well-wishes are contingent on any movements which aim to compete with open source stopping short of calling themselves open source. """
To me, this article is pointing out the intrinsic "cursed" problem of providing a startup mentality to an open source product. FOSS relies on giving freedom to replicate while (naive) startups are almost by definition looking for moats or monopolies to become unicorns.
There are ways of providing FOSS products, say as a by-product or ancillary project, but if a startup's main product is the source, it's almost antithetical to the business proposition. I think this point is only ever obliquely addressed and this article tackles it head on.
I also appreciate the articles focus on "open washing", where the new breed of "post-open" or "fair source" movements either engage in or stop just shy of trying to redefine the term "open source" to mean "source available" and not one that respects freedom.
The big question is though, how do you make business work in FOSS context. I'm glad that SourceHut is doing well, but I suspect it thrives partially due being small niche software. On the other end of scale you have FAANG style megacorps that do a lot of FOSS work too. But the middle-ground between these two extremes seems really difficult. Yet it feels desireable that non-niche FOSS development wouldn't end up being solely in the hands of megacorps.
It used to be that companies owned, and were proud of, their own headquarters, employed all the support functions personnel and so on. Contrasting janitors to CEOs is common trope, but who employs janitors anymore?
While this change has been economically incredibly profitable, I suspect there might be some long-term issues that crop up.
How does this relate to FOSS? In a sense software maintenance is lot like janitoring, and I do think one way of making the ecosystem healthier would be if companies would have larger in-house software divisions. So for example instead of the likes of Ford or Walmart paying to MS (or IBM) for stuff like Office suite, they would just have in-house teams working on LibreOffice.
> As for me, I still believe in open source, and even in the commercial potential of open source. It requires creativity and a clever business acumen to identify and exploit market opportunities within this collaborative framework. To win in open source you must embrace this collaboration and embrace the fact that you will share the commercial market for the software with other entities. If you’re up to that challenge, then let’s keep beating the open source drum together. If not, these new movements(source-available software) may be a home for you – but know that a lot of hard work still lies ahead of you in that path.
This article feels more like a critique of the new "source-available" trend in tech whereby existing OSS projects sell out and try to retroactively re-license themselves to be more closed-source friendly. I've been sort of aware of this phenomenon since there was a small upset when Muse Group bought Audacity, though it appears it is still GPL[1].
[1] https://github.com/audacity/audacity/blob/master/LICENSE.txt
So a lot of companies start up thinking they can open source their code, and treat their project like it was under a proprietary license with source available, then realise the hard way that's not the case. And a lot of authors release stuff for free when they don't see the value of it, then realise years later that they could be getting paid for their efforts.
And the article is right that it's probably not going to be possible to outcompete open source with any new license or incentive structure. Once people are used to something being free to use without any limits, any less free alternative is going to severely struggle to gain traction.
It could also be that the grassroots of the dotcom era literally grew up and wanted to capitalize on their work, while still retaining their values. What we have today is something in-between closed source and open source. I.e. "commercial stakeholders" isn't a merged branch, but the linear evolution of the original one.