Challenges to funding open source
I’ve had the good fortune to get paid to write open source as part of my job several times. For more than 15 years, I’ve also done a lot of open source development in my free time as a volunteer. Along the way, there’s been a fairly constant refrain that it’d be better if more open source maintainers were paid to maintain their projects. And I’ve seen a lot of ideas for how that could happen. While some of these ideas were successful, many more were not. I’m writing this post to explain three reasons that I’ve seen open source funding concepts fail. These aren’t the only reasons, but they’re reasons that I think are often ignored in the discourse.
This is not a normative piece, I’m not expressing an opinion on if anything here is good or bad. This is entirely descriptive of the things I have seen.
You can’t articulate what it is people are funding
Specifically, you can’t articulate how the world with funding would be different than the world without it. For many maintainers, the ideal form of open source funding is something like: instead of working on my project in my spare time, I get paid to do it and I can do it during the work day instead of stealing time from my other hobbies and personal obligations. Unfortunately, if you’re asking someone for money, this doesn’t tell them anything they’re getting for their money.
The underlying premise here is that if you can maintain your project without stealing time from other activities that the project will be more sustainable. This is a better premise, but still falls short, because sustainability is not about anything being different, it’s in fact about things being able to stay the same more predictably. Rationally, this sort of risk mitigation may make sense to invest in, but practically its a very difficult to assess risk (What are the odds that the existing maintainers will stop doing so as volunteers? Who knows! Many open source projects go years, if not decades, with exclusively volunteer maintainers). This explains why we see so little of the open-source funding that does occur is justified in this way.
If you are an open-source maintainer, be prepared to articulate: what will be different from the status quo if you have more time and money to work on your project. Is there a major feature that needs dedicated design and implementation time? A queue of bugs that need to be triaged and fixed? All of these are concrete things that can motivate people to support an open source project.
Many open source maintainers do not want to be paid
Or more accurately, they don’t want the expectations that come with taking money. If you’ve accepted money, that generally comes with some expectations (see the previous section). However, for many open source maintainers, having no expectations is the point.
For many folks, open source is enjoyable to work on precisely because you can reject features that offend your sensibilities, even if many users want them. You can take as long as you want until you find a design you like, you don’t have to ship something “good enough”. You can walk away for months if you’re bored, frustrated, or just don’t feel like working on the project.
There’s nothing wrong with this mode of maintainership (indeed, in my experience, having this approach makes my own open-source work more sustainable, even if it is entirely nights and weekends). But if this is why you love open-source, it is going to be challenging to find a way to take money for it without compromising some of what makes it enjoyable for you.
Time and money are not interchangeable
For open-source developers with a full time job, money does not translate into significantly more time for their project, unless the money is at the level of replacing a full time salary. It can offset some costs in life, and may be very appreciated. But the largest time obligation most people have is their day job, and for folks with full time jobs, a fraction of their salary doesn’t buy them extra hours in a week.
This means that for many open-source developers, small sums of money do not substantially increase their capacity to maintain their project. It doesn’t give them additional time to implement features, fix bugs, or triage issues. This means that to have an impact, open-source maintainers often need to find salary-replacement level funding for their open-source work.
This combines poorly with another simple observation: many open-source projects, including critical ones, do not make particular sense to fund at a full time level. A library for a stable compression format, for example, does not make sense to staff for 40 hours a week, that much time cannot be used efficiently. Of course, even if the format is stable there’s more than 0 hours a week of work! There’s porting to new platforms, updating for yet another backwards incompatible change in a dependency or CI platform, triaging bug reports, performance improvements, etc.
Perhaps some projects are so critical that as a society that we should fund them as a full-time endeavor even though the scope doesn’t justify it. However, one risk this carries is that developers who find themselves in such a position will need to be on guard against the temptations that more time provides, such as backwards incompatible changes for limited value and adding features of marginal value but with substantial new attack surface.
As an open-source maintainer, ask yourself how your life situation allows you to channel money effectively into work on your project, and what the scope of your project reasonably justifies in terms of hours spent on it.