Your customers, potential competitors, and even hyperscalers are all possibly the same people at different points. AGPL works because corporate lawyers are horrified by it, not because it actually discriminates against hyperscalers. If you find something that even sane lawyers wouldn't tolerate, then you have greatly limited the benefits of being open source in the first place. If you make a license that actually discriminates on user or use case, then it's not open source.
If you want to protect your project from being resold by potential competitors, do not release it as open source.
I think this problem might solve itself, though. Slowly but surely, companies and power users have become very wary of VC funded companies making big promises and big open source releases, with the knowledge that there is rarely a plan for sustainability and that there is a good chance if they stand on that rug it could be pulled later. Soon, if trends continue, the advantages that you once got from announcing something as open source will start to evaporate and turn into a liability as people start seeing ahead to the eventual "but of course we have to be able to monetize this eventually" stage.
The way I see it, a project can always be open sourced later on once there's a way to do it and ensure the company can remain sustainable. For the flagship product of a company, especially a VC-funded company, not starting open source is the ethical thing to do.
> For the flagship product of a company, especially a VC-funded company, not starting open source is the ethical thing to do.
I can't even..
Seriously, I don't understand where your argument is coming from. Because, if you look at it from the greater good perspective, commercial open-source is one of the only venues to build high quality software that can be freely self-hosted, modified and built-upon. Yet, you basically push for people building closed source software, due to what I understand is dogmatic believe of what open-source should mean.
> If you make a license that actually discriminates on user or use case, then it's not open source.
Based on a definition, bought and sponsored by the hyperscaler lobbies. Why the hell would discriminating against hyperscaller selling the product makes it non "open-source". 99.9999% users of open-source are not hyperscaler wanting to host and sell the product and will get value from the project.
Firstly, no it wasn't. The history of the open source definition and OSI are interesting, including hijacking a term that was already in use, but the hyperscalers had nothing to do with it. I'll avoid restating Wikipedia, it's enough to say that's not right.
Secondly, you can call it whatever you want but "open source" licenses that discriminate against user and use case are useless. Even if I just want to use something like a data structure implementation from your "open source" release in some unrelated project, I now have to inherit all of this baggage about your competitors. That just doesn't make any sense. The endgame of that is an ecosystem of open source that ultimately serves absolutely nobody except for maybe startup PR needs.
> The endgame of that is an ecosystem of open source that ultimately serves absolutely nobody except for maybe startup PR needs.
We self-host n8n, which by definition not open-source, and love it. It serves us. Not just n8n PR. That is the case for almost all self served products for non-commercial reasons
This is extremely confusing. I'm just talking about how it would harm the open source ecosystem if we allowed things that were not open source to be advertised as open source (and how it's bad when something is unopened after people start relying on it being open). I am not claiming that only open source software is good for society. Like, I don't really think it's a coincidence that GNU, DSFG and OSI all remain in relative agreement for decades over what makes something truly free or open source. They weren't all bought and sold by hyperscalers. There's reasons for the freedoms that are required to meet the definitions.
On the other hand, I think you can release shared source and closed source software that is still plenty useful and beneficial. For example, I am a very big fan of how Unreal Engine is licensed. Yes, it isn't "open source" and it isn't marketed as such. You can't take Unreal Engine components and go use them elsewhere even if you're not competing with Epic Games. Still, they provide some extremely powerful and useful software free of charge to basically all independent game developers. I think that's fantastic.
But that's all aside from what open source is. Unreal Engine isn't open source and it doesn't do anything for the open source ecosystem. Which is fine, because the entire world doesn't revolve around open source.
There is clearly a tension here where a bunch of people want to call their software open source without actually believing in what open source actually means. And this is absurd because it is betraying the fact that it's a marketing term. It seems like a bunch of startups realize that if they're trying to market to certain kinds of developers, if the software isn't open source, they won't touch it.
You seem to believe that we shouldn't create closed source software but at the same time directly advocating for it—that's what it is that's what you're describing. If you want a more practical and less idealistic reason why it sucks when software companies prevent people like AWS from using their software, it's because the actual users of your software, the customers of AWS, wish to pay AWS to host it for them. They would like to hire the lovely folks who work at AWS to host and manage the software for them and the license prevents that. You are preventing your users from doing what they want with the software. And that's the rub, that's why it's important to not discriminate based off of use.
Open source means Amazon pillages you and takes your margin. Period.
That just isn't going to work anymore.
The hyperscalers know this and they want you to keep building software this way. They'll write a managed version of your thing and collect all your money.
Fair source would be "customers can use this in an unlimited way as long as they don't sell a managed version".
That's what we need to do. Carve out the ability to make money for the originators doing the work.
Open source has turned us all into serfs working on giant kingdoms we don't own.
NOTE ALSO, the OSI has been so corporate captured that their definition of open source AI is not at all open. You shouldn't trust them at all.
I find it odd that lots of people talk about the OSI being corporate captured but I don't defer to them as to what 'counts'. I couldn't care less what the OSI does or approves of. Any software that grants users the four freedoms is Open Source—it's the same criteria to be Free Software which is decidedly not at all corporate captured and hasn't ever changed. The very first freedom of Free Software the freedom to run it for any purpose. It's not some small detail.
>> There is clearly a tension here where a bunch of people want to call their software open source without actually believing in what open source actually means.
So much this. If Open Source is a marketing term then companies are free to call whatever they do "Open Source" and the term becomes meaningless. You may as well call the software "Free Range".
On the other hand if Open Source is a defined thing, built around a specific definition, and the 4 freedoms, and so on, (hint: it is) then the marketing term could lead you places you don't want to go.
So there's a generation of programmers who want to "redefine" Open Source to suit their preferences. Dilute the definition until it's meaningless. I'm strongly against this.
To be clear everyone is welcome to use any license, with any restriction they like. But if it doesn't conform to the OSI definition don't call it Open Source.
Stop trying to turn a technical specification into a generic marketing term.
The term Open Source was coned because ESR, Bruce Perens and co. didn't like the term "Free Software" - they were afraid it will scare the software business away. At that epoch it was logical as software was mostly sold as a product, even a physical one. One of the arguments of OSI is you can build your services around OSS, just like Red Hat did.
Fast forward a few decades and it turns out you no longer can because any software you build in this way will be used by Amazon and they will earn the money, not you. From this PoV, pure OSS is almost hurting web-based software ecosystem development.
>> The term Open Source was coned because ESR, Bruce Perens and co. didn't like the term "Free Software" - they were afraid it will scare the software business
away.
Sure. They didn't completely align with Free Software, so they coined a new term for a new kind of license.
Current companies are free to do the same thing. If they don't like the Open Source or Free software licenses they are free to coin their own term, write their own license. Nobody has any problem with this, and every company is free to write whatever license they like.
What they don't get to do is co-opt the term "Open Source" and then change the meaning of it to something quite different. That's not ok.
I completely agree that Open Source is not necessarily a viable business strategy for most businesses. The solution to that is to not be Open Source.
For some reason, some companies though think this is not ok. They want to use the marketing term "Open Source", as if it means whatever they want it to mean. They want to build on the goodwill generated by decades of Open Source developers, and then bait-and-switch that goodwill at some point. That's no ok.
>> it turns out you no longer can because any software you build in this way will be used by Amazon and they will earn the money, not you.
You make this sound like this is something new. Whereas Open Source has always been used by big companies to make money. MacOS is based on NetBSD. That's over 25 years old now.
So yes, if you want to make a software business, and you plan to sell your product, then releasing it for free and allowing others to use it commercially is probably not a great business strategy. So if releasing it as Open Source is contrary to your business goals then, you know, don't release it as Open Source.
Framing it this way implies that open source is "flawed" specifically because it was trying to eschew the ideological aspects of the free software movement and make it more appealing to businesses and underscore its technical advantages rather than ethical. However this oddly common viewpoint ignores that the free software movement doesn't really define "free software" much differently than the open source software movement defines "open source software"; free software licenses still can't discriminate against different users or use cases.
> “Free software” does not mean “noncommercial.” On the contrary, a free program must be available for commercial use, commercial development, and commercial distribution. This policy is of fundamental importance—without this, free software could not achieve its aims.
What Amazon has done is of course bad for a lot of kinds of businesses like Elastic, but neither open source nor free software is about business models or how software development can be sustainable. For obvious reasons the perspective of VC-funded startups is overrepresented and overrated here of all places, but outside this space nobody cares. You license out your software to a community in hopes that it will help your business, and it might; you can get your software into package managers and Linux distributions easier if it is OSI/DFSG approved. The fact is, it's not a bug that this allows Amazon to go and use it and monetize it. Amazon uses and monetizes Linux and associated projects every single day, and it is ultimately beneficial to Linux and many open source projects. Best I can say is maybe you could argue Amazon can abuse their market position to compete unfairly with others, but that is truly irrespective of software licensing.
So what I make of this is simple: you shouldn't open source your flagship product unless your business model has a robust way to remain sustainable. In doing so, you are making a gamble that nobody else can monetize it better than you. Using an unfair CLA to trick people into contributing to a product that you already know with near certainty will be eventually closed source is just a dick move, and not knowing any better than to avoid this circumstance is negligent. Yes, as a proponent of the benefits of open source software, it is a bummer that we can't just have nice things. However, the last few years of CLA-based rugpulls have done great harm to people's trust in open source projects and startups based on open source, and I ultimately don't think we're better off for it.
> Stop trying to turn a technical specification into a generic marketing term.
It's already gone.
OpenAI, OpenArt, OpenWeb, OpenSea, OpenGov
The OSI is fucked too. The OSI definition of "open source AI" is horrible as you can't replicate or easily modify the systems without the data and tooling that the OSI definition doesn't require.
If AI eats software and the OSI definition of "open" AI wins, then we've already been captured by big business marketing.
I would love to hear your explanation for why you think that statement is an invocation of "no true Scotsman" of all things. Are you sure you know what that means?
> that enables your customers to use you freely [...] There are such licenses
There are such licenses only if you change the definition of "freely" to fit the narrative. Historically, "freely" (as in "free software") means granting end-user four essential software freedoms:
- The freedom to run the program as you wish, for any purpose (freedom 0).
- The freedom to study how the program works, and change it to make it do what you wish (freedom 1). Access to the source code is a precondition for this.
- The freedom to redistribute copies so you can help others (freedom 2).
- The freedom to distribute copies of your modified versions to others (freedom 3). By doing this you can give the whole community a chance to benefit from your changes. Access to the source code is a precondition for this.
If I can't redistribute ("resell") software, or can't run it and let others access it for a fee - it's not "use freely" anymore.
It is also an example of why I, as an end user, prefer FOSS licenses. The possibility of forking. The redistribution clauses there do not prevent forks, but definitely make them less likely, so you remain dependent on a single supplier.
The problem is that it creates a lot of uncertainty. For example, with this license, what usage is commercial? Can you distribute it to customers who buy other things from you? Can you offer services related to it - e.g. can you install it for a customer?
If you modify the software under what terms are your modifications distributed? Do they have to be distributed under the same terms? There is not transfer of copyright so the original authors cannot distribute the modified version commercially as part of their enterprise edition, that means that every modified version is a fork that cannot be merged back in?
As I said in another comment discouraging forks is a disadvantage for users as it leaves you dependant on one supplier.
The AGPL achieves the aim of deterring the likes of Amazon from providing the software as a service (at least without contributing their changes back, in practice the will not use it at all) without these sorts of problems.
I think that's perfectly fair. The community is quick to put whoever builds commercial OSS software on a cross the moment they change their license to ensure they still have a competitive advantage. Instead, we should encourage commercial OSS companies. COSS companiesare one of the only venues for creating high-quality OSS projects that you can self-host.
> I think that's perfectly fair. The community is quick to put whoever builds commercial OSS software on a cross the moment they change their license to ensure they still have a competitive advantage. Instead, we should encourage commercial OSS companies.
Complete agreement there. I'd like to laud NetBird for using AGPL rather than one of the recent VC-fueled proprietary-with-source-available licenses.
> I personally think the definition of open-source is problematic (and clearly biased by the lobbies of hyperscalers).
Open Source has existed since before "hyperscaler" was a concept that existed, and before Software as a Service was a going concern. Its definition has not in any way been affected by the lobby of an industry that didn't exist when it was defined.
One rationale for not changing the definition of Open Source is an issue of Schelling points / focal points ( https://en.wikipedia.org/wiki/Focal_point_(game_theory) ). Right now, we have a common definition of Open Source; if everyone could put their pet restriction in ("no military", "no SaaS", "no AI", "no nuclear power"), we'd end up with a hundred variants and no ability to collaborate and share code across projects.
> Why aren't n8n or MongoDB considered open-source?
In the case of MongoDB, it's because the SSPL requires that all the software used to offer the network service is also licensed under the SSPL. That prevents it from being used to write Free Software by mixing free programs and libraries that use a different license, even if they are free.
So, for example, if your network service supports managing MongoDB instances, and it includes Caddy or Nginx, then you're not complying with the license, as Caddy and Nginx aren't released under the SSPL and you cannot relicense them.
> Why does requesting that others not sell your product make the project not open-source?
Because requesting them to not do that makes your program proprietary, and thus non-free by definition.
The OSI definition doesn't allow that kind of restriction, mainly because it's all about keeping software as free and open as possible.
But the thing is, commercial open source companies play a huge role in making great open source tools, especially ones you can self-host. Without them, a lot of the software we rely on wouldn't even exist. People often push back when these companies change their licenses, but they forget the reality. Big cloud providers can make tons of money off open source projects without giving anything back. That's a tough spot for the folks.
I'm sure that in the nearest future we will have some COSS licenses :) Well, as an open source contributor I hope so
> The community is quick to put whoever builds commercial OSS software on a cross the moment they change their license to ensure they still have a competitive advantage.
Yes, people object to pulling a bait and switch by taking something that was open source and then making it not open source.
> Instead, we should encourage commercial OSS companies. COSS companiesare one of the only venues for creating high-quality OSS projects that you can self-host.
We do encourage that. We also discourage commercial fake-OSS companies.
Part of the point of Open Source is that the software isn't completely tied to a single company. If software is under one of those more restrictive licenses and the company goes under, the software is dead. If software under an open source license is developed by a company that goes under, one or even many other companies can continue working on it. This also applies while the company is alive, too; as you note, commercial companies developing open source software is a good thing, and preventing parallel development or forks is bad for the ecosystem.
I'll put it rather bluntly: present-day open source is largely free labor for SaaS companies.
SaaS, meanwhile, is the least open and least free model of software distribution, significantly less open or free (as in freedom) than closed-source commercial software you run yourself. This model, SaaS, is powered from the ground up by open source, and most SaaS gives little or nothing back. Some SaaS is not much more than a management and UI layer built around pre-existing open source standards and code.
Something is very wrong if open source exists largely to enable the least free model of software distribution. Open source as currently conceptualized is stuck in the pre-SaaS eras of the 1980s and 1990s and refuses to adapt to what "free" and "open" mean in the new landscape.
It doesn't help that the OSI is fully captured by companies like Google, Amazon, Microsoft, and Meta with a vested interest in promoting the SaaS and cloud-first model. If local-first ever gained traction it would be a threat to not just their SaaS products but their incredibly lucrative cloud businesses.
>> I'll put it rather bluntly: present-day open source is largely free labor for SaaS companies.
I've used Free Software and Open Source over 20 years and have never paid a SaaS company a dime.
OTOH the most valuable software that I regularly use is Free Software as opposed to Open Source Software. So maybe the OSS really is primarily free labor for SaaS?
I'm an OG open source user too. I installed Slackware Linux with kernel 0.99.15 from a stack of floppies as a teenager.
You're part of a very, very small minority. I'm talking about the majority of developers and what the majority of users experience.
What most users are experiencing today is an aggressively non-free non-open zero-privacy rent-seeking software environment that is enabled by open source under the hood. This seems contrary to the stated goals of free software.
>> What most users are experiencing today is an aggressively non-free non-open zero-privacy rent-seeking software environment that is enabled by open source under the hood. This seems contrary to the stated goals of free software.
That's why GPL is the preferred license for user freedom. HN is by its nature full of people who want to develop stuff for profit. Maybe it's time to stop hanging around here.
Very much so. When you run Oracle DBMS you have your own data in your custody. You also have the software and the ability to spin it up in the future. It’s a binary blob but there’s VMs and longer term emulation for that. I can run MS-DOS software for 16-bit x86 just fine on my ARM64 Mac. This includes unmodified commercial closed source software from decades ago.
With SaaS you usually don’t have your data let alone the code. The app cannot be run by anyone but the ones running it. Old versions can’t be run. If the company goes away it’s gone forever along with the data.
A lot of SaaS is tied to a specific cloud with its specific managed services, so it can’t even be moved between cloud providers without significant effort.
Honestly I think it's fair. Really the only people affected by this change are people creating proprietary forks, everyone else benefits from this change.
I just wish there was a way to ensure that the company itself doesn't do a proprietary fork.
Simple. Don't have contribution agreements. Contributors maintain their copyright so as they can prevent relicensing. Mind, FSF requires copyright for their submissions (I believe), but they're, arguably, a "good actor" in this space.
But, if the code base becomes a patchwork of contributors, it can become difficult to relicense.
Definition of freedom has changed in cloud era. The idea behind GPL was, if you modify your code and distribute it, you need to contribute back the changes, with the wider goal of increasing the commons. All the legalese was just to make sure that the idea worked.
In the era of cloud, distribution needs to include distribution over wire, because so many apps are now run in the cloud. And that's why AGPL. It preserves the spirit of GPL in modern times.
Yes, this idea that you must “contribute back” is itself an infringement on privacy, and is why copyleft licenses are falling out of favor of things like MIT.
Not the OP, but here’s how I see it: freedom doesn’t exist in a vacuum. Whether it’s in society or in software, real freedom comes with some responsibility. You can’t just take, take, take and expect the ecosystem to stay healthy. That’s not how any of this works.
Same thing applies to free software. If we actually care about keeping that world alive, there’s gotta be some duty baked in—some expectation to contribute back, or at least not strip-mine the commons and bounce.
MIT’s blown up mostly because the big players don’t need the broader free software community anymore. They’ve got the scale, the headcount, and the cash to build and maintain their own internal ecosystems. So from their POV, permissive licenses like MIT are perfect—no obligations, no copyleft, no friction.
But let’s not pretend that’s “freedom” in the idealistic sense. It’s freedom to extract, sure. Freedom to integrate and forget where it came from. For a lot of people on HN, that’s fine. But if you care about the sustainability of the broader ecosystem? Then yeah, we’ve got to talk about duty, not just rights.
If you are distributing an AGPL software as a derivative work, you must also distribute the source. You can make as many modifications for internal use and not distribute them, if you're not making a derivative work.
> They see this as bad, because they think you should be forced to publish your modifications to the software you use internally, even if you don’t want to.
This is incorrect. You are only forced to publish if you are creating derivative works. See how overleaf uses propietary git integration with AGPL overleaf.
That article makes a lot of dubious claims. The central claim is that by breaking links to the source code in development on your own machine with no one else using it, then you are breaking the license by not showing a notice offering the source to yourself.
This is typical of software developers trying to interpret law. Can you imagine someone explaining to a judge that they are suing for a breach of license terms under the circumstances. "So, you are saying he did not give himself access to the code on his laptop?"
Even if that nonsense was correct, there is a dead easy workaround. Run a server with the code on it bound to localhost and you then have your network server for all users interacting with the code (yourself!). Not needed, just an additional layer of proof the claim "it is impossible to comply" is false.
Edit: to add, I am also not impressed by the author's other blog posts, such as a moan about not having PRs for FOSS projects accepted for good reasons (if you dig down into it). Lots of other complaining and nonsense too.
When given the choice, there are very few cases where I don't pick the open source product. The peace of mind it won't be shutdown on you, the confidence the company has in their code quality, or the option to ignore migrations and stay on previous versions is worth way more than featureset deltas.
Licensing talk is confusing to someone not steeped in it. I talked to Claude about it 1-2 weeks ago when it was first announced and it was framed as a reinforcement of FOSS ideals.
I actually made the jump from tailscale -> netbird last month. Definitely more work and learning, but much more aligned w/ my perspective of self-hosting and open-source software. (Yes I thought about headscale but the YouTube reviews of netbird won me over).
Going from tailscale in the cloud -> self-hosting netbird required a richer mental model so I had to learn more about routing, why ZTNA is better, what is OIDC, etc
I do think that netbird's documentation is easier to read than tailscale's, but the tone does still assume a solid foundational networking background in places.
My uneducated guess is that the product is appealing to networking professionals, but a growing number of current/former tailscale users that are otherwise new to networking, but familiar w/ self-hosting. With the latter group, there's a steeper learning curve (that would also be there for headscale or most other self-hosted mesh VPN solution fwiw)
In order to safeguard the long-term innovation, sustainability, and collaborative spirit of NetBird, we are switching to the AGPLv3 license - ensuring it remains a powerful, community-driven resource for decades to come.
Looks like AGPL is a new norm? Redis switched to AGPL too. SSPL is also common on the server side. Curious of how you view AGPL vs. SSPL and choose the former.
SSPL is appealing for business but it is not open source. That is a deal breaker for us. We want to remain open source under a license that is recognized as open source.
SSPL is open source by every definition. The OSI rejected it, but their explanation for why boils down to "neener neener" (it's not backed by any facts). Most other organizations haven't bothered to take a position because no notable software uses it and it's not worth the hassle to evaluate (mongodb and redis have better alternatives so nobody cares about them).
late addition: the OSI is a consortium of software and cloud companies; the ones whose business model is ruined by SSPL. They aren't neutral and we probably shouldn't let them be the arbiters of what counts as open source.
Foundations run in a more non-profit, community-oriented way include the FSF, EFF, and Debian, none of which made any significant comment. Debian has excluded SSPL software, but their criteria for inclusion are stricter than simply "is it open source?" and they announced it was simpler to replace them with their superior non-SSPL equivalents than to actually tackle the question.
They're switching because they saw the failure of doing the source-available rugpull and causing other, sometimes even more successful, forks to show up, like Redis and Valkey. SSPL is not open source so it's not something I'd ever choose.
>The BSD-3 license, under which NetBird has operated until now, is a permissive license. It was instrumental in our early growth, offering maximum flexibility and encouraging wide adoption. However, this permissiveness also presents a significant long-term challenge with an imbalance where the value created by a community can be captured and privatized, ultimately undermining the sustainability of the open-source project itself. Well, AGPLv3 addresses this imbalance.
How is this logic not literally Embrace, Extend, Extinguish?
>How is this logic not literally Embrace, Extend, Extinguish?
It is exactly that. We need more free software which is actually free for everyone and every use case in all the senses of free. We don't need more "free software" except there are owners who get to control who uses it, how they use it, and how they can make money with it.
There is SO MUCH WASTE that could be eliminated by a few developers getting paid decent salaries to put their work into the public domain (by this I mean BSD style very permissive licenses).
Imagine a grant giving organization that companies were encouraged to give a hundredth of a percent of their revenue to which focused on paying full time developers to build and maintain fully featured tools which are the most useful to society as a whole.
I have said a lot of times that I feel like people keep trying to reinvent a state and taxes to pay for shared infrastructure with open source maintenance. I don’t know how to use the state to solve the problem without severely degrading the quality of what gets built though.
I don't want the state to do it. I would prefer a future where the state becomes less necessary because folks just do stuff themselves. The state already gives tax incentives for donations to nonprofits, so it's not inventing something new. Open source software where people contribute their work to share with nothing expected in return is an early signal of something I'm hoping will become a major feature of a post-scarcity economy. People "working" for the benefit of everyone on their own without expectation of returns (and receiving the benefits of others who do the same). It's not clear that it will be possible for the whole economy to work this way, but there are plenty of things where people want to do work to share because they choose to and it just eliminates a whole bunch of unnecessary work when "for hire" things can be replaced by "for free".
That's a weird take. If another $company wants to continue development with BSD-3 license, they can do so starting today and nothing of value would be lost.
The change is NetBird company saying, the improvements from now on are AGPLv3 licensed, but that doesnt stop from anyone to fork today and continue with BSD-3 license.
More and more firms are realising that AGPL coupled with dual licensing allows them to be open source whilst protecting their commercial interests.
AGPL doesn't have the critical "no commercialization" clause that prevents a direct competitor. It just prevents hyperscalers.
AGPL will stop Amazon. It won't stop WP Engine.
There needs to be a license that enables your customers to use you freely, but not your competitors from reselling your hard work.
Your customers, potential competitors, and even hyperscalers are all possibly the same people at different points. AGPL works because corporate lawyers are horrified by it, not because it actually discriminates against hyperscalers. If you find something that even sane lawyers wouldn't tolerate, then you have greatly limited the benefits of being open source in the first place. If you make a license that actually discriminates on user or use case, then it's not open source.
If you want to protect your project from being resold by potential competitors, do not release it as open source.
I think this problem might solve itself, though. Slowly but surely, companies and power users have become very wary of VC funded companies making big promises and big open source releases, with the knowledge that there is rarely a plan for sustainability and that there is a good chance if they stand on that rug it could be pulled later. Soon, if trends continue, the advantages that you once got from announcing something as open source will start to evaporate and turn into a liability as people start seeing ahead to the eventual "but of course we have to be able to monetize this eventually" stage.
The way I see it, a project can always be open sourced later on once there's a way to do it and ensure the company can remain sustainable. For the flagship product of a company, especially a VC-funded company, not starting open source is the ethical thing to do.
> For the flagship product of a company, especially a VC-funded company, not starting open source is the ethical thing to do.
I can't even..
Seriously, I don't understand where your argument is coming from. Because, if you look at it from the greater good perspective, commercial open-source is one of the only venues to build high quality software that can be freely self-hosted, modified and built-upon. Yet, you basically push for people building closed source software, due to what I understand is dogmatic believe of what open-source should mean.
> If you make a license that actually discriminates on user or use case, then it's not open source.
Based on a definition, bought and sponsored by the hyperscaler lobbies. Why the hell would discriminating against hyperscaller selling the product makes it non "open-source". 99.9999% users of open-source are not hyperscaler wanting to host and sell the product and will get value from the project.
Firstly, no it wasn't. The history of the open source definition and OSI are interesting, including hijacking a term that was already in use, but the hyperscalers had nothing to do with it. I'll avoid restating Wikipedia, it's enough to say that's not right.
Secondly, you can call it whatever you want but "open source" licenses that discriminate against user and use case are useless. Even if I just want to use something like a data structure implementation from your "open source" release in some unrelated project, I now have to inherit all of this baggage about your competitors. That just doesn't make any sense. The endgame of that is an ecosystem of open source that ultimately serves absolutely nobody except for maybe startup PR needs.
> The endgame of that is an ecosystem of open source that ultimately serves absolutely nobody except for maybe startup PR needs.
We self-host n8n, which by definition not open-source, and love it. It serves us. Not just n8n PR. That is the case for almost all self served products for non-commercial reasons
This is extremely confusing. I'm just talking about how it would harm the open source ecosystem if we allowed things that were not open source to be advertised as open source (and how it's bad when something is unopened after people start relying on it being open). I am not claiming that only open source software is good for society. Like, I don't really think it's a coincidence that GNU, DSFG and OSI all remain in relative agreement for decades over what makes something truly free or open source. They weren't all bought and sold by hyperscalers. There's reasons for the freedoms that are required to meet the definitions.
On the other hand, I think you can release shared source and closed source software that is still plenty useful and beneficial. For example, I am a very big fan of how Unreal Engine is licensed. Yes, it isn't "open source" and it isn't marketed as such. You can't take Unreal Engine components and go use them elsewhere even if you're not competing with Epic Games. Still, they provide some extremely powerful and useful software free of charge to basically all independent game developers. I think that's fantastic.
But that's all aside from what open source is. Unreal Engine isn't open source and it doesn't do anything for the open source ecosystem. Which is fine, because the entire world doesn't revolve around open source.
"fair source" is one of the new nomenclatures for software with source available that is anti-hyperscaler.
There is clearly a tension here where a bunch of people want to call their software open source without actually believing in what open source actually means. And this is absurd because it is betraying the fact that it's a marketing term. It seems like a bunch of startups realize that if they're trying to market to certain kinds of developers, if the software isn't open source, they won't touch it.
You seem to believe that we shouldn't create closed source software but at the same time directly advocating for it—that's what it is that's what you're describing. If you want a more practical and less idealistic reason why it sucks when software companies prevent people like AWS from using their software, it's because the actual users of your software, the customers of AWS, wish to pay AWS to host it for them. They would like to hire the lovely folks who work at AWS to host and manage the software for them and the license prevents that. You are preventing your users from doing what they want with the software. And that's the rub, that's why it's important to not discriminate based off of use.
Open source means Amazon pillages you and takes your margin. Period.
That just isn't going to work anymore.
The hyperscalers know this and they want you to keep building software this way. They'll write a managed version of your thing and collect all your money.
Fair source would be "customers can use this in an unlimited way as long as they don't sell a managed version".
That's what we need to do. Carve out the ability to make money for the originators doing the work.
Open source has turned us all into serfs working on giant kingdoms we don't own.
NOTE ALSO, the OSI has been so corporate captured that their definition of open source AI is not at all open. You shouldn't trust them at all.
I find it odd that lots of people talk about the OSI being corporate captured but I don't defer to them as to what 'counts'. I couldn't care less what the OSI does or approves of. Any software that grants users the four freedoms is Open Source—it's the same criteria to be Free Software which is decidedly not at all corporate captured and hasn't ever changed. The very first freedom of Free Software the freedom to run it for any purpose. It's not some small detail.
>> There is clearly a tension here where a bunch of people want to call their software open source without actually believing in what open source actually means.
So much this. If Open Source is a marketing term then companies are free to call whatever they do "Open Source" and the term becomes meaningless. You may as well call the software "Free Range".
On the other hand if Open Source is a defined thing, built around a specific definition, and the 4 freedoms, and so on, (hint: it is) then the marketing term could lead you places you don't want to go.
So there's a generation of programmers who want to "redefine" Open Source to suit their preferences. Dilute the definition until it's meaningless. I'm strongly against this.
To be clear everyone is welcome to use any license, with any restriction they like. But if it doesn't conform to the OSI definition don't call it Open Source.
Stop trying to turn a technical specification into a generic marketing term.
The term Open Source was coned because ESR, Bruce Perens and co. didn't like the term "Free Software" - they were afraid it will scare the software business away. At that epoch it was logical as software was mostly sold as a product, even a physical one. One of the arguments of OSI is you can build your services around OSS, just like Red Hat did.
Fast forward a few decades and it turns out you no longer can because any software you build in this way will be used by Amazon and they will earn the money, not you. From this PoV, pure OSS is almost hurting web-based software ecosystem development.
>> The term Open Source was coned because ESR, Bruce Perens and co. didn't like the term "Free Software" - they were afraid it will scare the software business away.
Sure. They didn't completely align with Free Software, so they coined a new term for a new kind of license.
Current companies are free to do the same thing. If they don't like the Open Source or Free software licenses they are free to coin their own term, write their own license. Nobody has any problem with this, and every company is free to write whatever license they like.
What they don't get to do is co-opt the term "Open Source" and then change the meaning of it to something quite different. That's not ok.
I completely agree that Open Source is not necessarily a viable business strategy for most businesses. The solution to that is to not be Open Source.
For some reason, some companies though think this is not ok. They want to use the marketing term "Open Source", as if it means whatever they want it to mean. They want to build on the goodwill generated by decades of Open Source developers, and then bait-and-switch that goodwill at some point. That's no ok.
>> it turns out you no longer can because any software you build in this way will be used by Amazon and they will earn the money, not you.
You make this sound like this is something new. Whereas Open Source has always been used by big companies to make money. MacOS is based on NetBSD. That's over 25 years old now.
So yes, if you want to make a software business, and you plan to sell your product, then releasing it for free and allowing others to use it commercially is probably not a great business strategy. So if releasing it as Open Source is contrary to your business goals then, you know, don't release it as Open Source.
Framing it this way implies that open source is "flawed" specifically because it was trying to eschew the ideological aspects of the free software movement and make it more appealing to businesses and underscore its technical advantages rather than ethical. However this oddly common viewpoint ignores that the free software movement doesn't really define "free software" much differently than the open source software movement defines "open source software"; free software licenses still can't discriminate against different users or use cases.
https://www.gnu.org/philosophy/free-sw.en.html
> Free software can be commercial
> “Free software” does not mean “noncommercial.” On the contrary, a free program must be available for commercial use, commercial development, and commercial distribution. This policy is of fundamental importance—without this, free software could not achieve its aims.
What Amazon has done is of course bad for a lot of kinds of businesses like Elastic, but neither open source nor free software is about business models or how software development can be sustainable. For obvious reasons the perspective of VC-funded startups is overrepresented and overrated here of all places, but outside this space nobody cares. You license out your software to a community in hopes that it will help your business, and it might; you can get your software into package managers and Linux distributions easier if it is OSI/DFSG approved. The fact is, it's not a bug that this allows Amazon to go and use it and monetize it. Amazon uses and monetizes Linux and associated projects every single day, and it is ultimately beneficial to Linux and many open source projects. Best I can say is maybe you could argue Amazon can abuse their market position to compete unfairly with others, but that is truly irrespective of software licensing.
So what I make of this is simple: you shouldn't open source your flagship product unless your business model has a robust way to remain sustainable. In doing so, you are making a gamble that nobody else can monetize it better than you. Using an unfair CLA to trick people into contributing to a product that you already know with near certainty will be eventually closed source is just a dick move, and not knowing any better than to avoid this circumstance is negligent. Yes, as a proponent of the benefits of open source software, it is a bummer that we can't just have nice things. However, the last few years of CLA-based rugpulls have done great harm to people's trust in open source projects and startups based on open source, and I ultimately don't think we're better off for it.
> Stop trying to turn a technical specification into a generic marketing term.
It's already gone.
OpenAI, OpenArt, OpenWeb, OpenSea, OpenGov
The OSI is fucked too. The OSI definition of "open source AI" is horrible as you can't replicate or easily modify the systems without the data and tooling that the OSI definition doesn't require.
If AI eats software and the OSI definition of "open" AI wins, then we've already been captured by big business marketing.
> If you make a license that actually discriminates on user or use case, then it's not open source.
No true scotsman.
I would love to hear your explanation for why you think that statement is an invocation of "no true Scotsman" of all things. Are you sure you know what that means?
Because I anticipate the following:
A: If you make a license that actually discriminates on user or use case, then it's not open source.
B: But here are n open source projects that discriminate on user or use case, e.g. everything GPL, everything with dual license, etc.
A: yeeeeeeaaaah, but those aren’t TRUE open source projects!!! A true open source project would never do that!
> There needs to be a license that enables your customers to use you freely, but not your competitors from reselling your hard work.
There are such licenses. They are just not open source.
> that enables your customers to use you freely [...] There are such licenses
There are such licenses only if you change the definition of "freely" to fit the narrative. Historically, "freely" (as in "free software") means granting end-user four essential software freedoms:
- The freedom to run the program as you wish, for any purpose (freedom 0).
- The freedom to study how the program works, and change it to make it do what you wish (freedom 1). Access to the source code is a precondition for this.
- The freedom to redistribute copies so you can help others (freedom 2).
- The freedom to distribute copies of your modified versions to others (freedom 3). By doing this you can give the whole community a chance to benefit from your changes. Access to the source code is a precondition for this.
If I can't redistribute ("resell") software, or can't run it and let others access it for a fee - it's not "use freely" anymore.
Thanks. That's my definition of "use freely." The argument about what's open source mostly boils down to whether you accept these terms.
This one is great, but unfortunately not open source: https://github.com/n8n-io/n8n/blob/master/LICENSE.md
Not sure what you mean by "unfortunately". If you're looking for a license that's not open source, you'll find one that isn't - obviously.
It is also an example of why I, as an end user, prefer FOSS licenses. The possibility of forking. The redistribution clauses there do not prevent forks, but definitely make them less likely, so you remain dependent on a single supplier.
This is great.
"Open" is too binary. You can't exclude companies that would harm you.
We need "fair", "equitable", or "sustainable" (as this one is termed) licenses.
You should be able to give your customers 100% the ability to use, modify, and redistribute so long as they don't resell you.
You should be able to cut off Amazon and Google from using you.
You should be able to prevent certain uses, like managed versions of your software.
The problem is that it creates a lot of uncertainty. For example, with this license, what usage is commercial? Can you distribute it to customers who buy other things from you? Can you offer services related to it - e.g. can you install it for a customer?
If you modify the software under what terms are your modifications distributed? Do they have to be distributed under the same terms? There is not transfer of copyright so the original authors cannot distribute the modified version commercially as part of their enterprise edition, that means that every modified version is a fork that cannot be merged back in?
As I said in another comment discouraging forks is a disadvantage for users as it leaves you dependant on one supplier.
The AGPL achieves the aim of deterring the likes of Amazon from providing the software as a service (at least without contributing their changes back, in practice the will not use it at all) without these sorts of problems.
Fair point, but it won't be considered open source. NetBird wants to stay open source. AGPL gives certain protection, but is not 100% bullet proof.
> AGPL will stop Amazon. It won't stop WP Engine.
WP Engine is not a problem, Amazon is.
The company doing 90% of the effort should see 90% of the upside.
WP Engine, as I understand it, contributes next to nothing to the codebase and takes a huge bite of the overall market.
"no commercializaton" is not free software.
The users should not be restricted. The responsibility comes from redistribution.
Maximum freedom is attained when users are restricted from taking away the freedom of other users.
I meant users literally.
rms told me - you can USE the software for any purpose.
Yep, and if your use involves giving the software to other people, but you don't want to give them the source code, you're restricted from that use.
There is no legal definition where "use" is somehow equal to "distribute" as you're using it.
There's no legal definition of any of this, it's fully specified by the license.
I think that's perfectly fair. The community is quick to put whoever builds commercial OSS software on a cross the moment they change their license to ensure they still have a competitive advantage. Instead, we should encourage commercial OSS companies. COSS companiesare one of the only venues for creating high-quality OSS projects that you can self-host.
I personally think the definition of open-source is problematic (and clearly biased by the lobbies of hyperscalers). Why aren't n8n or MongoDB considered open-source? (https://github.com/n8n-io/n8n/blob/master/LICENSE.md, https://www.mongodb.com/legal/licensing/community-edition) Why does requesting that others not sell your product make the project not open-source?
> I think that's perfectly fair. The community is quick to put whoever builds commercial OSS software on a cross the moment they change their license to ensure they still have a competitive advantage. Instead, we should encourage commercial OSS companies.
Complete agreement there. I'd like to laud NetBird for using AGPL rather than one of the recent VC-fueled proprietary-with-source-available licenses.
> I personally think the definition of open-source is problematic (and clearly biased by the lobbies of hyperscalers).
Open Source has existed since before "hyperscaler" was a concept that existed, and before Software as a Service was a going concern. Its definition has not in any way been affected by the lobby of an industry that didn't exist when it was defined.
One rationale for not changing the definition of Open Source is an issue of Schelling points / focal points ( https://en.wikipedia.org/wiki/Focal_point_(game_theory) ). Right now, we have a common definition of Open Source; if everyone could put their pet restriction in ("no military", "no SaaS", "no AI", "no nuclear power"), we'd end up with a hundred variants and no ability to collaborate and share code across projects.
> Why aren't n8n or MongoDB considered open-source?
In the case of MongoDB, it's because the SSPL requires that all the software used to offer the network service is also licensed under the SSPL. That prevents it from being used to write Free Software by mixing free programs and libraries that use a different license, even if they are free.
So, for example, if your network service supports managing MongoDB instances, and it includes Caddy or Nginx, then you're not complying with the license, as Caddy and Nginx aren't released under the SSPL and you cannot relicense them.
> Why does requesting that others not sell your product make the project not open-source?
Because requesting them to not do that makes your program proprietary, and thus non-free by definition.
The OSI definition doesn't allow that kind of restriction, mainly because it's all about keeping software as free and open as possible.
But the thing is, commercial open source companies play a huge role in making great open source tools, especially ones you can self-host. Without them, a lot of the software we rely on wouldn't even exist. People often push back when these companies change their licenses, but they forget the reality. Big cloud providers can make tons of money off open source projects without giving anything back. That's a tough spot for the folks.
I'm sure that in the nearest future we will have some COSS licenses :) Well, as an open source contributor I hope so
Making money isn't the problem. Restricting it is.
> The community is quick to put whoever builds commercial OSS software on a cross the moment they change their license to ensure they still have a competitive advantage.
Yes, people object to pulling a bait and switch by taking something that was open source and then making it not open source.
> Instead, we should encourage commercial OSS companies. COSS companiesare one of the only venues for creating high-quality OSS projects that you can self-host.
We do encourage that. We also discourage commercial fake-OSS companies.
> I personally think the definition of open-source is problematic (and clearly biased by the lobbies of hyperscalers). Why aren't n8n or MongoDB considered open-source? (https://github.com/n8n-io/n8n/blob/master/LICENSE.md, https://www.mongodb.com/legal/licensing/community-edition) Why does requesting that others not sell your product make the project not open-source?
Part of the point of Open Source is that the software isn't completely tied to a single company. If software is under one of those more restrictive licenses and the company goes under, the software is dead. If software under an open source license is developed by a company that goes under, one or even many other companies can continue working on it. This also applies while the company is alive, too; as you note, commercial companies developing open source software is a good thing, and preventing parallel development or forks is bad for the ecosystem.
I'll put it rather bluntly: present-day open source is largely free labor for SaaS companies.
SaaS, meanwhile, is the least open and least free model of software distribution, significantly less open or free (as in freedom) than closed-source commercial software you run yourself. This model, SaaS, is powered from the ground up by open source, and most SaaS gives little or nothing back. Some SaaS is not much more than a management and UI layer built around pre-existing open source standards and code.
Something is very wrong if open source exists largely to enable the least free model of software distribution. Open source as currently conceptualized is stuck in the pre-SaaS eras of the 1980s and 1990s and refuses to adapt to what "free" and "open" mean in the new landscape.
It doesn't help that the OSI is fully captured by companies like Google, Amazon, Microsoft, and Meta with a vested interest in promoting the SaaS and cloud-first model. If local-first ever gained traction it would be a threat to not just their SaaS products but their incredibly lucrative cloud businesses.
>> I'll put it rather bluntly: present-day open source is largely free labor for SaaS companies.
I've used Free Software and Open Source over 20 years and have never paid a SaaS company a dime.
OTOH the most valuable software that I regularly use is Free Software as opposed to Open Source Software. So maybe the OSS really is primarily free labor for SaaS?
I'm an OG open source user too. I installed Slackware Linux with kernel 0.99.15 from a stack of floppies as a teenager.
You're part of a very, very small minority. I'm talking about the majority of developers and what the majority of users experience.
What most users are experiencing today is an aggressively non-free non-open zero-privacy rent-seeking software environment that is enabled by open source under the hood. This seems contrary to the stated goals of free software.
>> What most users are experiencing today is an aggressively non-free non-open zero-privacy rent-seeking software environment that is enabled by open source under the hood. This seems contrary to the stated goals of free software.
That's why GPL is the preferred license for user freedom. HN is by its nature full of people who want to develop stuff for profit. Maybe it's time to stop hanging around here.
> SaaS, meanwhile, is the least open and least free model of software distribution.
Less free and less open than Oracle DBMS? That seems like a stretch.
Very much so. When you run Oracle DBMS you have your own data in your custody. You also have the software and the ability to spin it up in the future. It’s a binary blob but there’s VMs and longer term emulation for that. I can run MS-DOS software for 16-bit x86 just fine on my ARM64 Mac. This includes unmodified commercial closed source software from decades ago.
With SaaS you usually don’t have your data let alone the code. The app cannot be run by anyone but the ones running it. Old versions can’t be run. If the company goes away it’s gone forever along with the data.
A lot of SaaS is tied to a specific cloud with its specific managed services, so it can’t even be moved between cloud providers without significant effort.
Honestly I think it's fair. Really the only people affected by this change are people creating proprietary forks, everyone else benefits from this change.
I just wish there was a way to ensure that the company itself doesn't do a proprietary fork.
Simple. Don't have contribution agreements. Contributors maintain their copyright so as they can prevent relicensing. Mind, FSF requires copyright for their submissions (I believe), but they're, arguably, a "good actor" in this space.
But, if the code base becomes a patchwork of contributors, it can become difficult to relicense.
FSF stopped requiring contributor agreements years ago.
It’s not fair. The AGPL violates freedom 0 and is nonfree.
https://sneak.berlin/20250720/the-agpl-is-nonfree/
Running a SaaS with in-house modifications is a protected use case for free software. The AGPL is a EULA masquerading as a license.
Definition of freedom has changed in cloud era. The idea behind GPL was, if you modify your code and distribute it, you need to contribute back the changes, with the wider goal of increasing the commons. All the legalese was just to make sure that the idea worked.
In the era of cloud, distribution needs to include distribution over wire, because so many apps are now run in the cloud. And that's why AGPL. It preserves the spirit of GPL in modern times.
Yes, this idea that you must “contribute back” is itself an infringement on privacy, and is why copyleft licenses are falling out of favor of things like MIT.
The spirit of the GPL was never a good idea.
Not the OP, but here’s how I see it: freedom doesn’t exist in a vacuum. Whether it’s in society or in software, real freedom comes with some responsibility. You can’t just take, take, take and expect the ecosystem to stay healthy. That’s not how any of this works.
Same thing applies to free software. If we actually care about keeping that world alive, there’s gotta be some duty baked in—some expectation to contribute back, or at least not strip-mine the commons and bounce.
MIT’s blown up mostly because the big players don’t need the broader free software community anymore. They’ve got the scale, the headcount, and the cash to build and maintain their own internal ecosystems. So from their POV, permissive licenses like MIT are perfect—no obligations, no copyleft, no friction.
But let’s not pretend that’s “freedom” in the idealistic sense. It’s freedom to extract, sure. Freedom to integrate and forget where it came from. For a lot of people on HN, that’s fine. But if you care about the sustainability of the broader ecosystem? Then yeah, we’ve got to talk about duty, not just rights.
If you are distributing an AGPL software as a derivative work, you must also distribute the source. You can make as many modifications for internal use and not distribute them, if you're not making a derivative work.
> They see this as bad, because they think you should be forced to publish your modifications to the software you use internally, even if you don’t want to.
This is incorrect. You are only forced to publish if you are creating derivative works. See how overleaf uses propietary git integration with AGPL overleaf.
That article makes a lot of dubious claims. The central claim is that by breaking links to the source code in development on your own machine with no one else using it, then you are breaking the license by not showing a notice offering the source to yourself.
This is typical of software developers trying to interpret law. Can you imagine someone explaining to a judge that they are suing for a breach of license terms under the circumstances. "So, you are saying he did not give himself access to the code on his laptop?"
Even if that nonsense was correct, there is a dead easy workaround. Run a server with the code on it bound to localhost and you then have your network server for all users interacting with the code (yourself!). Not needed, just an additional layer of proof the claim "it is impossible to comply" is false.
Edit: to add, I am also not impressed by the author's other blog posts, such as a moan about not having PRs for FOSS projects accepted for good reasons (if you dig down into it). Lots of other complaining and nonsense too.
The best part about my website is that visiting it is completely optional.
When given the choice, there are very few cases where I don't pick the open source product. The peace of mind it won't be shutdown on you, the confidence the company has in their code quality, or the option to ignore migrations and stay on previous versions is worth way more than featureset deltas.
Licensing talk is confusing to someone not steeped in it. I talked to Claude about it 1-2 weeks ago when it was first announced and it was framed as a reinforcement of FOSS ideals.
I actually made the jump from tailscale -> netbird last month. Definitely more work and learning, but much more aligned w/ my perspective of self-hosting and open-source software. (Yes I thought about headscale but the YouTube reviews of netbird won me over).
What did you find harder to achieve with NetBird than with Tailscale? I refer to more work and learning. Or is it purely related top self-hosting?
Going from tailscale in the cloud -> self-hosting netbird required a richer mental model so I had to learn more about routing, why ZTNA is better, what is OIDC, etc
I do think that netbird's documentation is easier to read than tailscale's, but the tone does still assume a solid foundational networking background in places.
My uneducated guess is that the product is appealing to networking professionals, but a growing number of current/former tailscale users that are otherwise new to networking, but familiar w/ self-hosting. With the latter group, there's a steeper learning curve (that would also be there for headscale or most other self-hosted mesh VPN solution fwiw)
The future continues to be AGPL:
https://vadosware.io/post/the-future-of-free-and-open-source...
In order to safeguard the long-term innovation, sustainability, and collaborative spirit of NetBird, we are switching to the AGPLv3 license - ensuring it remains a powerful, community-driven resource for decades to come.
Looks like AGPL is a new norm? Redis switched to AGPL too. SSPL is also common on the server side. Curious of how you view AGPL vs. SSPL and choose the former.
SSPL is appealing for business but it is not open source. That is a deal breaker for us. We want to remain open source under a license that is recognized as open source.
SSPL is open source by every definition. The OSI rejected it, but their explanation for why boils down to "neener neener" (it's not backed by any facts). Most other organizations haven't bothered to take a position because no notable software uses it and it's not worth the hassle to evaluate (mongodb and redis have better alternatives so nobody cares about them).
late addition: the OSI is a consortium of software and cloud companies; the ones whose business model is ruined by SSPL. They aren't neutral and we probably shouldn't let them be the arbiters of what counts as open source.
Foundations run in a more non-profit, community-oriented way include the FSF, EFF, and Debian, none of which made any significant comment. Debian has excluded SSPL software, but their criteria for inclusion are stricter than simply "is it open source?" and they announced it was simpler to replace them with their superior non-SSPL equivalents than to actually tackle the question.
They're switching because they saw the failure of doing the source-available rugpull and causing other, sometimes even more successful, forks to show up, like Redis and Valkey. SSPL is not open source so it's not something I'd ever choose.
+1
>The BSD-3 license, under which NetBird has operated until now, is a permissive license. It was instrumental in our early growth, offering maximum flexibility and encouraging wide adoption. However, this permissiveness also presents a significant long-term challenge with an imbalance where the value created by a community can be captured and privatized, ultimately undermining the sustainability of the open-source project itself. Well, AGPLv3 addresses this imbalance.
How is this logic not literally Embrace, Extend, Extinguish?
They are literally changing license to protect from E³. The only logic that makes what they do E³ is the slavery is freedom logic.
>How is this logic not literally Embrace, Extend, Extinguish?
It is exactly that. We need more free software which is actually free for everyone and every use case in all the senses of free. We don't need more "free software" except there are owners who get to control who uses it, how they use it, and how they can make money with it.
There is SO MUCH WASTE that could be eliminated by a few developers getting paid decent salaries to put their work into the public domain (by this I mean BSD style very permissive licenses).
Imagine a grant giving organization that companies were encouraged to give a hundredth of a percent of their revenue to which focused on paying full time developers to build and maintain fully featured tools which are the most useful to society as a whole.
I have said a lot of times that I feel like people keep trying to reinvent a state and taxes to pay for shared infrastructure with open source maintenance. I don’t know how to use the state to solve the problem without severely degrading the quality of what gets built though.
I don't want the state to do it. I would prefer a future where the state becomes less necessary because folks just do stuff themselves. The state already gives tax incentives for donations to nonprofits, so it's not inventing something new. Open source software where people contribute their work to share with nothing expected in return is an early signal of something I'm hoping will become a major feature of a post-scarcity economy. People "working" for the benefit of everyone on their own without expectation of returns (and receiving the benefits of others who do the same). It's not clear that it will be possible for the whole economy to work this way, but there are plenty of things where people want to do work to share because they choose to and it just eliminates a whole bunch of unnecessary work when "for hire" things can be replaced by "for free".
Who does this extinguish other than people who didn't want to share with the rest of the world---ie. hoard?
That's a weird take. If another $company wants to continue development with BSD-3 license, they can do so starting today and nothing of value would be lost.
The change is NetBird company saying, the improvements from now on are AGPLv3 licensed, but that doesnt stop from anyone to fork today and continue with BSD-3 license.