> 82 percent of Maven Central demand comes from fewer than 1 percent of IPs
This is interesting, and the author goes on to say that 80% comes from the largest cloud providers. But I wonder how much of that is coming from CI pipelines and how much is actually the cloud provider's usage?
When looking at the OpenBSD Foundation list of donors, I can see Microsoft has giving a decent amount to OpenBSD over the years. I assume for use of OpenSSH. I have to give MS credit for this.
Now for a different company, IBM is completely missing and they use OpenSSH everywhere, even on AIX. And now it is even more glaring since they own Red Hat.
Yes this is a small "survey", but it shows not all large companies donates to projects that are really critical to their products.
So yes, at the very least, if your company depends upon a Open Source Project, you should throw a bit over the wall :)
> We must realign how businesses work with open source so that payment is no longer an optional charitable gift but a cost of doing business. To do that, we need an organization to create a viable, supportable path from big business to individual programmer. It's time for someone to step up and make this happen. Businesses, open source software, and maintainers will all be better off for it.
Congratulations, you rediscovered commercial software - where you are legally obligated pay to use software.
Maybe providers of these services should start randomly return 402 return code. At least for those request which do not have sufficient authentication linked to existing payments.
Paying for hosting costs seems straightforward. Sonatype has decided to host Maven Central and treat it effectively as a marketing expense, and they are free to change that if they want. Same for the hosts of PyPi, RubyGems, etc.
Developer labor is a separate issue and what most people (including the article author) seem most confused about. Open source developers generally fall into one of a few different categories: Hobbyists looking for enjoyment, aspiring professionals looking for experience, startups looking for exposure and adoption, and corporate employees maintaining software their business relies on.
Contributors to vast majority of usable software fall into the last category, yet all of the focus is on the subset of people in the first category that attract a userbase that is meaningful enough to theoretically warrant some support, which is probably the smallest faction of open source developers by a large margin.
I do not understand the angst over this. If a hobbyist gets tired of their hobby, they are free to move on. If they feel exploited by some big corporation, change the license going forward, make the repo private, push harder for compensation until they feel properly compensated, or take whatever other action resolves their internal issue.
All the other categories of developer are fine as is, and the dissatisfied category could be as well if they took more agency over the situation.
The tip jar is fine, the problem is that most corporations have no process to drop anything in the tip jar without purchase orders signed in triplicate, sent in, sent back, queried, lost, found, subjected to public inquiry, lost again, and finally buried in soft peat for three months and recycled as firelighters.
The same process is why open source is such a hit among the developers that actually accomplish real work in such corporations.
> I mean, after all, they've made billions from this code.
As someone whose dream job is to just build open source software and have a comfortable life, I'm highly sympathetic to open source sustainability and I do hope we continue to seek for solutions.
But this type of statement is ridiculous. There is a hell of a lot more to business than just the code, despite what many of us software engineers want to think. It's also quite rare for a commercial company to airlift an open source project and make billions on it. There is also a massive spectrum/range of open source, from tiny nearly throwaway libraries up to massive applications.
Turning open source into commercial software is NOT the solution. Commercial software has existed forever, and continues to (try doing something non-trivial with PDFs if you want a modern painful example). If open source becomes commercial software, we'd be losing out on a mountain of greatness. Imagine if downloading a Linux distro required paying for and receiving a license? And if you want to make a distro, be prepared to buy licenses for OpenSSL and every little thing that makes up that system, and set up your accounting books and what not so you can properly distribute all your revenues to the thousands (or tens of thousands or hundreds of thousands) of sub-projects. And don't forget that you also have to have technical/legal apparati capable of enforcing, maybe auditing, etc.
Nothing is stopping you from spinning your open source project into a commercial operation right now. Plenty of people do it (it's usually called "source available" because you largely have to, by necessity, restrict redistribution, which makes it no longer "open source" according to most definitions of said term). The great thing about freedom and choice is that you can go whichever direction you want.
Copyright is a bundle of rights and copyright licenses don't have to license all of the rights in that bundle under the same terms.
You could make a license that makes copying and distribution not cost any money but charges for use. People who make distributions would then not have to pay anything for including the software in a distribution.
That seems like it would leave a lot of projects without any revenue, which doesn't seem fair in this hypothetical new world of charging for all your open source code. How many people are actually compiling and installing systemd themselves rather than having it come with their distro?
I deeply dislike this framing of mega corps vs volunteer devs. It conveniently ignores the huge amount of open source being developed by regular salaried devs as part of their job. Imho that is what we need more of instead of trying to redirect some money to these individuals which to me seems inherently unstable approach.
To make that happen I think companies should be more willing to develop and publish their own patches instead of relying on upstream for anything. Overall I think in the open source world the idea of (centralized upstream) "project" has gotten way overinflated.
I've long held that open source is one of the world's biggest anarchist experiments. Anarchism, as understood by the likes of Kropotkin, largely believed that we can self organize towards working for the wellbeing of all, that s self organized groups will genuinely build useful and high quality tools.
Rather than turning open source into just another commercial effort, I'd love to explore going the other way. Why do we need to pay open source developers? They need housing, food, etc. Maybe the better answer is to figure out how to make those things freely available to open source developers.
It's possible to imagine a world where everything works like open source -- share what you have in excess, take what you need, work on something you enjoy for the betterment of all.
I've sometimes wondered what it would be like now if some of the big names early in open source and free software had decided to take a detour of a couple decades to work in commercial software with the goal of making as much money as possible, and then retire and use that money to endow a foundation to promote free/open source software development by paying developers.
Imagine if Stallman had gotten the IBM work instead of Gates, and now it was the Free Software Foundation with billions instead of the the Gates Foundation. With those resources their endowment would be generating enough income that they would be able to pay around 10000 programmers a year $100k plus full benefits to do free software work.
That would have meant less free software for those decades while there amassing their fortunes, but then an explosion of free software afterwards.
"I want to be the selfless craftsperson giving away work for free to anyone, but I'll also pressure profit-maximizing evil mega-corporations to give me money from the good of their heart, despite the fact that I've explicitly stated in the license they don't have to" is just not a smart position to hold.
If you want evil corporations to have to pay money in exchange for using your software, add that as a condition in the license. Ah, but then it's not "free software", sorry.
There's so much unexplored space in licenses that achieve better outcomes for both the developers and their non-giant-evil-conglomerate users, but nobody is willing to touch that subject, because then they're not writing "real free software" and the "FOSS community" will not use it.
I like Cryptomator's solution: donate to get a pretty banner.
Also, it didn't work -- Mountain Duck is closed source.
Personally I donate €50 every now and then when the average of the donation goes below a certain value (varies by project) but it requires tracking in a Spreadsheet.
Something like " If you require customizations or enhancements we bill at 1000$ an hour, 8 hour minimum."
I don't particularly care if someone working at Microsoft or whatever sees my open source project and decides to use it. That's fantastic. But I'm absolutely not going to work for Microsoft for free, so they need something for one of their use cases they need to pay up.
Open source will be bigger than ever, but not necessarily for the right reasons.
IMHO, as coding agents mature, we will see a surge in open-source projects. Most will remain obscure due to poor discoverability and growing competition for users' attention. But the same forces driving that growth in volume will likely erode quality and it most certainly raising ethical concerns. A company can now perform a clean-room reimplementation of an open-source project and potentially maintain it going forward with more resources and polish than the original.
For the most successful, high-impact projects, this creates a real problem that they risk losing momentum to well-funded clones.
At that point, what is the incentive to keep building in the open?
True open source is dead. Ai and big tech will eat it instantly. Dual licensure is ok (AGPL+paid). You can disagree but then don’t whine when someone instantly repackages and markets your tool and gets paid when you don’t.
> Anyone building modern software depends on language registries such as Maven Central, PyPI, npm, crates.io, and others
So, who's going to tell Linus Torvalds (among many others) he's not writing modern software? I have a feeling I know exactly what the author considers "modern" software...
This is fine as long as everything actually stays FOSS, rather than becoming fauxpen source. The best way to do this is to license everything under AGPLv3+, which the megacorps are all allergic to, and then sell exceptions to them.
FLOSS is a tool for efficiency, for sharing automation and knowledge; just like knowledge itself, it cannot be commercial, it cannot have paywalls. Simply put, everyone must develop what they need, publishing the code from day zero, from the very first commit, this is what creates the community on a larger scale.
The problem isn't Open Source, it's the commercial organisation of knowledge which is simply unsustainable. It is society that needs to change.
29 comments
[ 3.2 ms ] story [ 43.8 ms ] threadThis is interesting, and the author goes on to say that 80% comes from the largest cloud providers. But I wonder how much of that is coming from CI pipelines and how much is actually the cloud provider's usage?
Now for a different company, IBM is completely missing and they use OpenSSH everywhere, even on AIX. And now it is even more glaring since they own Red Hat.
Yes this is a small "survey", but it shows not all large companies donates to projects that are really critical to their products.
So yes, at the very least, if your company depends upon a Open Source Project, you should throw a bit over the wall :)
Congratulations, you rediscovered commercial software - where you are legally obligated pay to use software.
I view it as a type of charity. I know not everybody can afford to use their time without compensation. that's ok!
but I will personally never charge, and I oppose this commercial mindset
Paying for hosting costs seems straightforward. Sonatype has decided to host Maven Central and treat it effectively as a marketing expense, and they are free to change that if they want. Same for the hosts of PyPi, RubyGems, etc.
Developer labor is a separate issue and what most people (including the article author) seem most confused about. Open source developers generally fall into one of a few different categories: Hobbyists looking for enjoyment, aspiring professionals looking for experience, startups looking for exposure and adoption, and corporate employees maintaining software their business relies on.
Contributors to vast majority of usable software fall into the last category, yet all of the focus is on the subset of people in the first category that attract a userbase that is meaningful enough to theoretically warrant some support, which is probably the smallest faction of open source developers by a large margin.
I do not understand the angst over this. If a hobbyist gets tired of their hobby, they are free to move on. If they feel exploited by some big corporation, change the license going forward, make the repo private, push harder for compensation until they feel properly compensated, or take whatever other action resolves their internal issue.
All the other categories of developer are fine as is, and the dissatisfied category could be as well if they took more agency over the situation.
The same process is why open source is such a hit among the developers that actually accomplish real work in such corporations.
As someone whose dream job is to just build open source software and have a comfortable life, I'm highly sympathetic to open source sustainability and I do hope we continue to seek for solutions.
But this type of statement is ridiculous. There is a hell of a lot more to business than just the code, despite what many of us software engineers want to think. It's also quite rare for a commercial company to airlift an open source project and make billions on it. There is also a massive spectrum/range of open source, from tiny nearly throwaway libraries up to massive applications.
Turning open source into commercial software is NOT the solution. Commercial software has existed forever, and continues to (try doing something non-trivial with PDFs if you want a modern painful example). If open source becomes commercial software, we'd be losing out on a mountain of greatness. Imagine if downloading a Linux distro required paying for and receiving a license? And if you want to make a distro, be prepared to buy licenses for OpenSSL and every little thing that makes up that system, and set up your accounting books and what not so you can properly distribute all your revenues to the thousands (or tens of thousands or hundreds of thousands) of sub-projects. And don't forget that you also have to have technical/legal apparati capable of enforcing, maybe auditing, etc.
Nothing is stopping you from spinning your open source project into a commercial operation right now. Plenty of people do it (it's usually called "source available" because you largely have to, by necessity, restrict redistribution, which makes it no longer "open source" according to most definitions of said term). The great thing about freedom and choice is that you can go whichever direction you want.
You could make a license that makes copying and distribution not cost any money but charges for use. People who make distributions would then not have to pay anything for including the software in a distribution.
To make that happen I think companies should be more willing to develop and publish their own patches instead of relying on upstream for anything. Overall I think in the open source world the idea of (centralized upstream) "project" has gotten way overinflated.
If I wanted to get paid for the software I make in my free time, I would put a price on it.
If someone likes what I do personally, they can donate on my Patreon or kofi or whatever.
If I want my project to only be used for other free software, then I make it GPL or AGPL. That's it.
If someone uses my software and works for a company and needs support, we can talk about a support contract.
Rather than turning open source into just another commercial effort, I'd love to explore going the other way. Why do we need to pay open source developers? They need housing, food, etc. Maybe the better answer is to figure out how to make those things freely available to open source developers.
It's possible to imagine a world where everything works like open source -- share what you have in excess, take what you need, work on something you enjoy for the betterment of all.
Imagine if Stallman had gotten the IBM work instead of Gates, and now it was the Free Software Foundation with billions instead of the the Gates Foundation. With those resources their endowment would be generating enough income that they would be able to pay around 10000 programmers a year $100k plus full benefits to do free software work.
That would have meant less free software for those decades while there amassing their fortunes, but then an explosion of free software afterwards.
"I want to be the selfless craftsperson giving away work for free to anyone, but I'll also pressure profit-maximizing evil mega-corporations to give me money from the good of their heart, despite the fact that I've explicitly stated in the license they don't have to" is just not a smart position to hold.
If you want evil corporations to have to pay money in exchange for using your software, add that as a condition in the license. Ah, but then it's not "free software", sorry.
There's so much unexplored space in licenses that achieve better outcomes for both the developers and their non-giant-evil-conglomerate users, but nobody is willing to touch that subject, because then they're not writing "real free software" and the "FOSS community" will not use it.
Also, it didn't work -- Mountain Duck is closed source.
Personally I donate €50 every now and then when the average of the donation goes below a certain value (varies by project) but it requires tracking in a Spreadsheet.
Something like " If you require customizations or enhancements we bill at 1000$ an hour, 8 hour minimum."
I don't particularly care if someone working at Microsoft or whatever sees my open source project and decides to use it. That's fantastic. But I'm absolutely not going to work for Microsoft for free, so they need something for one of their use cases they need to pay up.
IMHO, as coding agents mature, we will see a surge in open-source projects. Most will remain obscure due to poor discoverability and growing competition for users' attention. But the same forces driving that growth in volume will likely erode quality and it most certainly raising ethical concerns. A company can now perform a clean-room reimplementation of an open-source project and potentially maintain it going forward with more resources and polish than the original.
For the most successful, high-impact projects, this creates a real problem that they risk losing momentum to well-funded clones.
At that point, what is the incentive to keep building in the open?
So, who's going to tell Linus Torvalds (among many others) he's not writing modern software? I have a feeling I know exactly what the author considers "modern" software...
The problem isn't Open Source, it's the commercial organisation of knowledge which is simply unsustainable. It is society that needs to change.