Open Source Developers Are Exhausted, Unpaid, and Ready to Walk Away
75 points by thang
75 points by thang
Predictable income through decentralized funding that preserves maintainer autonomy.
I really wish a better mechanism existed for deciding who gets this funding, rather than "whoever writes the best grants". It is certainly does the job on some levels, but I spent 5 years in academia watching the grant-writing treadmill and the toxic culture that creates.
I wish GitHub Sponsors would get more popular. This sort of patronage is a good way to fund projects in a way that doesn't depend on the artificial scarcity of the final product. Sadly only a relative minority of developers are popular enough to be funded.
The problem keeping sponsorships from being a sustainable solution is that most sponsorships come from other devs. If you're lucky you'll get some people who succeeded at scaling a startup that pay it forward, but most of the time it's devs moving money around in a near-circle while the platform gets a cut.
That's an interesting point. When you put it like that, it sounds like the intermediary is the only one actually winning here...
I don't know if more popularity will be enough; the amounts are absolutely paltry. If tokio, one of the most relied-upon library teams in the Rust ecosystem, only gets $2340 per month then we're 2 orders of magnitude short of paying developers what they're worth. The bloke who built my former RSS reader of choice has 7 sponsors. By now, "people will pay for stuff if they don't have to" seems to be a failed experiment.
By now, "people will pay for stuff if they don't have to" seems to be a failed experiment.
That’s another win for utility maximization, unfortunately. Perhaps it’s time for some game theory.
Patronage is great and I agree we should have more of that, but I do not want GitHub in between me and the people I'm trying to fund.
If you haven't tried it already, I can personally commend sending e-mails like "I really appreciate your work on $project. What's the easiest way for me to send you the cost of a nice lunch for two, as thanks? I'm in $country, in case that makes a difference." If their GitHub or other forge profile page doesn't list their address, you might find it in Git commit metadata.
Sometimes the answer is that it's just plain easier to use a platform they've already set up, even if it charges commission. I try not to complain. After all, I'm trying to thank them for work they've already done, not create a new chore for them to do that isn't really related to software.
Are there alternatives? GitHub Sponsors is great because it integrates into the repository itself.
Yes! Liberapay, OpenCollective, Thanks.dev, Ko-fi
You can set-up funding.yml that will show additional ways of funding project/people/org
Voluntary donations are only for feeling good, they can never be the actual solution. You somehow need to force companies to pay, because if you don't force my company to pay you, I can't get them to donate to you without spending undue political capital, regardless of how much I personally like and support your project.
Would it be possible for the grant writing process to be better if people who are reading the grants are closer to the action so to speak?
To be honest I believe academia people who talk about the grant process being very very very awful but I have a hard time coming up with an alternative that doesn't have even nastier incentives. The grant stuff seems to be "convince somebody your work is important" and while that means that the somebody there is important, it feels a bit better than things like "be friends with somebody who makes the decision" (though maybe those dynamics are still at play with grants!)
The fix isn't complicated. Treat maintainers like the humans they are, not free infrastructure. Companies profiting from open source need to contribute financially (at the very least).
Unfortunately this dehumanization is a cultural problem within open source, and it's clear from the language used around funding
Software doesn't need financial contributions, software doesn't have "maintenance costs" - people writing software need money to be able to pay their rent, to be able to eat, and to be able to seek medical care, but there is a cultural taboo around stating this explicitly which is why the language has centralized around this doublespeak
Pay OSS developers reliably. Not donations or tips. Predictable income through decentralized funding that preserves maintainer autonomy. Foster recognition and respect too.
I have written before about how I've largely given up on sponsorship, donations and tips, so I won't rehash that. If people want to look at numbers, I have done yearly financial retrospectives for 2023 and 2024, and I will be doing another one for 2025 at the end of this year
2025 was the year that I introduced dual licensing for commercial use (the primary license only permits personal use), and it took only a few months for the number of active licenses to surpass the number of sponsors
From surveying people who previously did not sponsor but are now individual commercial use license holders, it's clear that being able to submit a license receipt for reimbursement to their employer makes a big difference, and for me this is the desired goal - workers supporting workers in the redistribution of wealth hoarded by corporations
people writing software need money to be able to pay their rent, to be able to eat, and to be able to seek medical care, but there is a cultural taboo around stating this explicitly
It's especially bizzare that the taboo persists in a world where even in developed countries almost no one has any financial security anymore, but lots of people still feel a need to pretend that they are closer to being independently wealthy than to being homeless, and people who control the funding are certainly happy to exploit that, I suppose.
From surveying people who previously did not sponsor but are now individual commercial use license holders, it's clear that being able to submit a license receipt
I also learnt it the hard way when I started VyOS that corporate sponsorships just aren't going to happen. All companies have processes for buying things, but most of them don't even have any process for donations or sponsorships, even if there are people in the company who want to do that. In many companies it's also relatively easy to convince higher-ups to buy licenses but impossible to convince them to give any money for something available for free. When we restricted binary LTS images to paying customers and on, the peanut gallery kept shouting that it would kill the project, but the opposite it true — VyOS would be dead now if we expected people to start paying for support or donating when they could get everything at no cost.
All companies have processes for buying things, but most of them don't even have any process for donations or sponsorships, even if there are people in the company who want to do that.
The process is the big thing. I once spent ~50 hours at a big company donating ~$1k to open source. And nobody pushed back or resisted in any way, it just took that long to navigate the bureaucracy. If I had been able to buy something, like a support license, it would have taken less than an hour.
One thing I would really like to see is support bundles for the major libraries in an ecosystem that distributed the money to the developers somehow. These wouldn't even have to do much meaningfully different than allow companies to post issues on github – the point is that companies are already used to paying for support.
All companies have processes for buying things, but most of them don't even have any process for donations or sponsorships, even if there are people in the company who want to do that.
I think at some point Vim had sponsorships and paid licenses. Paid licenses reaffirmed the rights from the opensource license and then added the perks that sponsors got (same amount of time for perks, same price). Apparently in some companies even this was good enough to expense the license, but sponsorship wouldn't be allowed.
2025 was the year that I introduced dual licensing for commercial use (the primary license only permits personal use)
What license do you use to restrict to personal use?
The PolyForm project is a great starting point (previous discussion) - the "Strict" variant is pretty easy to modify to be personal use only without other safe harbors (which is what I did)
Perhaps they should join together into some kind of organization so that they can collectively negotiate for what they want. Some kind of united social structure, so that they can threaten to all stop working at the same time, while making specific demands in order to continue providing maintenance services. If only such a thing were possible... has anyone ever done something like this before??
You can't 'threaten to stop work' if you don't have two parties engaged in some kind of agreed relationship where both benefit. If you literally have zero usable leverage, that's your own fault. What would creating the leverage look like, is a more interesting question to me.
Sounds a little like industry licensing coalitions like MPEG-LA, where they create and publish things like video codecs and then non-profit entities can use them for free (or at least aren't worth litigating against) while for-profit entities have to cough up for licensing costs or get sued. (Less evil industry coalitions like Khronos seem similar but different, their job is less to make money for the creators and more to maintain a collaborative commons.)
Unfortunately those kind of coalitions tend to operate to enrich shareholders and lawyers, rather than creators. But perhaps we can learn something from the collective negotiating model? Alas now that I think of it this was kinda what the FSF was supposed to be for...
Definitely agree that looking at all kinds of successful models would be helpful. I think even if we wouldn't necessarily agree with some particular outcome (e.g. shareholders getting rich), there's something to be gained from understanding how systems get set up in a way that says 'For some thing to come out of this end, this is what has to go in at the other end. If there isn't enough stuff going in, the thing will stop coming out. Right now, here's what the 'stuff in' > 'stuff out' relationship looks like and where there are gaps or surplusses. If there are gaps in what's going in and you care about the stuff coming out, either help put more stuff in or expect the stuff to stop coming out'. Capitalist businesses would say 'if there isn't a way to set this up so the whole thing works, none of it will even begin to exist'. imo open source projects don't have a very clear understanding of their own 'stuff in > stuff out' relationships because they're so messy and meat based and probably include things like 'if my partner keeps taking care of our kids, then I have the spare time to keep working on this project' and so what some open source project actually needs is for the maintainer's partner to be supported, or for them to be able to afford childcare or something, and a huge part of the 'in' for many developers to begin with is enjoyment, reward, challenge, etc. imo2 there's no shame and I would call it a requirement or perhaps even responsibility to be able to understand your own 'in > out' mapping and proactively communicate that so that people can at least have some rough idea of how you need support. imo3 the combination of not doing that and delivering something for free makes both ends of the in > out pipe really confusing and impossible to manage for everyone involved. capitalism is, unfortunately, really good at this exposing and quantifying the relationship between the in and out. (sorry this is an unedited dump/verbalisation of thoughts I've had on this subject for a long time)
Maintainers can yank versions, poison them. You name it.
That kills trust, which is essential for open-source.
A group simply threatening to walk away if their needs aren't met sends a more powerful signal and threatening to poison the well to spite a few members of the community.
Difficulty Getting Paid: Sixty percent of OSS maintainers receive no payment whatsoever (according to the Tidelift survey). They work full-time jobs, then maintain critical infrastructure for free.
This is obviously true. But it needs to be remembered that a huge number of FOSS devs don't want to be paid and treat the work as a hobby. Many have never even attempted to be paid.
It also does start to become difficult once you have more than one maintainer and a lot of outside contributions. At least, that's always how I felt for the one moderately successful open source project I was one of the founders for.
At what point do I cut in other contributors, how about people who contributed in the past?
There are of course solutions for that and it highly differs per project. But it certainly can factor into it all.
I sometimes wonder if it was a mistake to push for open source licenses that so freely allow commercial use.
On the one hand it is most likely one of the major reasons for the widespread adoption of open source, but on the other hand, the motivations for the giants who profit on this, have nothing other than kindness to motivate them to pay the developers.
A recent article about Google’s conduct on GitHub regarding security reports comes to mind.
The open source developer is essentially forced to provide free labor for big corporations that profit off of this.
It doesn’t feel right, fair, or sustainable.
"forced" seems strong. What if they simply do not?
Then they get their projects taken away from them. Remember left-pad? He went away because they took his username. He unpublished his projects.
That in turn broke many big projects that used his small silly oneliner lib. That in turn killed many commercial projects.
So npm (I don't remember if they were already owned by Microsoft back then) just re-published his code "for the good of the ecosystem" or some such.
You either play by the rules the few big centralised players set, or they take your shit away.
What exactly got taken away in that story? Is NPM forcing him to do work for free by re publishing his lib themselves?
they took his username
It sounds like they didn't just republish a compatible version under a new artifact identifier.
I guess it's this article: FFmpeg to Google: Fund Us or Stop Sending Bugs. It's kind of surprising to see so many dismissive comments on Lobsters on that article though.
I resent the blanket term of "Open Source Developers" meaning people who want to get paid for developing open source code. I think that implicitly presupposes that open source is made to be commercially valuable, which is not true! I'm not a big open source developer, but the few times I did open source something it was because I wrote some code that I wanted to share with the world. I don't want to get paid for it, even if a major company decided to use it.
I think a better way to handle this is to change the social contract to treat Open Source as a gift. If you don't like anything for whatever reason, you are free to not use it. Maintainers have no obligation to fix anything, even if they are critical security vulnerabilities. So if you use an open source project commercially, and you run into an issue, remember that and either fix it yourself or pay someone to fix it for you. If you pay the person who created it, that's great! But I don't want the core value of open source to be that you build something and give it away with the expectation of support - both you supporting users and users supporting you.
Yeah a lot of this is a mood. Sometimes I'm not sure if I regret making Anubis open source. People get super entitled and bitchy in ways that I just don't want to deal with.
The whole idea that anything can be so "shared" as to have no value in itself is not a problem if the rest of the world ensures that nobody _is_ starving or needing money. For young people who have parents who pay for them or student grants or loans and basically have yet to figure out that it costs a hell of a lot of money to live in a highly advanced society, this is not such a bad idea. Grow up, graduate, marry, start a family, buy a house, have an accident, get seriously ill for a while, or a number of other very expensive things people actually do all the time, and the value of your work starts to get very real and concrete to you, at which point giving away things to be "nice" to some "community" which turns out not to be "nice" _enough_ in return that you will actually stay alive, is no longer an option.
"Fix it, fork it, f*ck off" becomes the phrase of choice
This should be the starting point. If you aren't paying me, I'll do whatever work I feel like doing. You can do the work to make things better for your use case.
I remember when open source was a past time hobby.
I remember when it was about ‘scratching an itch’. You write software that solves a problem for you. The value you get from it is that now there exists some software that solves your problem. You open source it and maybe the next person with similar problems doesn’t have to start from scratch. And maybe, if they use yours as a base, they might fix some bugs or add a feature you want but didn’t have time for.
And maybe if a bunch of you have similar problems, you might do a bit of organising to avoid duplicating effort.
And that’s pretty much how it still is for project I’ve created or joined. Some have companies involved, but they understand the social contract embodied by the ‘no warranty’ bit of the license: if they want something that isn’t there, they can add it, pay someone (sometime me) to add it, or fork it and add it themselves. Most of what I’ve done is permissively licensed, if they want to do an in-house proprietary fork, they can: I don’t lose anything (and I don’t have to deal with any of their issues).
Was this article written by AI? I feel like it's dramatic flourishes before the end of each paragraph and overuse of bold reads a lot like ChatGPT.
Edit: I think maybe ChatGPT was just trained a whole lot on this writing style. The author has been doing something like it since 2021. I guess I just don't read this type of writing a lot.
Some good suggestions there.
Pay OSS developers reliably. The question is where the fund comes from
I don't think it is entirely funding that is the issue here when we are specifically talking about burnout. To be very clear, I do think OSS maintainers deserve to be paid for the countless hours they invest.
But, a lot of the wear on people I am seeing would still be there due to a variety of reasons. Two main reasons boil down to it often feeling like thankless, somewhat lonely work. In my experience there is always a shortage of maintainers compared to the work. Even if you are getting paid you still only have a finite amount of time to spend so you need to make choices. A lot of OSS projects have a hard time getting people to contribute and stick around. In fact, a majority of open source project is probably just one maybe two people outside incidental contributors.
Then there is the demanding attitude from parts of the audience. :Which in combination to the previous point adds up quickly. People/companies can be incredibly entitled in their demands, to ridiculous degrees. Dealing with that can be incredibly draining. In fact, I feel like getting paid can even make this one bigger (I still think OSS maintainers should be paid). As and now you are getting paid so they should be accommodated, even though they aren't necessary the one paying the maintainer.
Typing this out loud, they probably do relate to why most OSS work doesn't get funded nearly enough. Very few people think to contribute, and one group even throws in demands in there as well.
With capitalism being the way of life, opensource makes zero sense, especially for very large projects used in production. It’s self-inflicted though and very easy to fix.