The Open Source Pledge: peer pressure to pay maintainers

In the early days of open source, it was a struggle to get companies to accept the concept and trust its development model. Now, companies have few qualms about using it, but do tend to take open source and those who maintain it for granted. The struggle now is to find ways to compensate producers of the software, sustain the open‑source commons, and avoid burning out maintainers. The Open Source Pledge project is an effort to persuade companies to pay maintainers by making it a social norm. On October 8, the project is launching a marketing campaign to raise awareness and try to get a larger conversation started around paying maintainers.

Within the open-source community itself, there is no lack of awareness: people have been raising the alarm for years that open‑source projects are under‑funded, that maintainers are overworked and in danger of burnout. The xkcd dependency comic has appeared in so many presentations, reports, and blog posts about the topic that "random person in Nebraska" is part of the lexicon. And just in case anyone had missed the memo, the XZ utils backdoor put a powerful spotlight on the precarious nature of an ecosystem built on volunteer labor and the dangers of maintainer burnout.

As the saying goes, admitting that there is a problem is the first step toward recovery. Most agree that maintainers deserve to be paid for their work, and that things would be better if the maintainers were able to make a living working on open source. Unfortunately, there's little consensus about who should pay and how.

How it started

Planning for the project kicked off in April 2024, with inspiration from projects like Pledge 1% and The Climate Pledge. The project soft-launched in June at OSSPledge.com with Sentry as its sponsor and its head of open source, Chad Whitacre, leading the effort. The project later moved to reduce the focus on Sentry's sponsorship and make it a more vendor-neutral project.

Whitacre has been trying to find solutions to maintainer funding for well more than a decade now. He founded the (now-defunct) maintainer-payment platform Gratipay in 2012, and in 2016 he co-founded a group called Sustain that puts on events and hosts working groups around making open source sustainable. Now he, and other Open Source Pledge contributors, think that the answer, or at least an answer, is to establish a new norm where companies voluntarily pay maintainers, based roughly on their usage, and then talk about it. The hope is that social pressure will encourage more and more companies to participate.

How it works

To join the pledge, companies would pay $2,000 per year (or more) for every "full-time equivalent developer on staff" directly to maintainers or foundations that sponsor open-source maintainers. I spoke to Whitacre on October 2. He said that the rationale for focusing on the number of developers, rather than a company's revenue or another metric, is that developers are the ones whose productivity benefits the most from open‑source software. The $2,000 figure is something that Whitacre came up with many years ago, based on some back-of-the-napkin math (and rounded down a bit):

Companies benefit from open source most directly in their own technical employees' boosted productivity. There are 21M employed programmers in the world. To move 25,000 of them into full-time open source work (at a 50% cost savings) and also account for non-technical contributors, each company with their own technical employees should pay the open source community $2,143 per year per technical employee at the company.

So, if a company has 10 full-time developers on staff, its minimum contribution should be $20,000 per year. That is a big increase from $0, but a bargain compared to proprietary software licensing and subscriptions.

The project does not handle, nor does it want to handle, any of the money itself. Companies pay their dues directly to maintainers, using platforms like the Open Source Collective, or give funds to foundations that pay maintainers. Whitacre said that the PHP Foundation was "the farthest along" in terms of directly funding developers to work on an open‑source project. He also said that the project was not in the business of providing resources on how to take payments or work with maintainers to fund them: "Foundations and platforms are the answer to that [...] let the foundations become stewards".

Once a company has paid, it should publish a blog post that details the payments that it has made in the past year. The project points to blog posts by Sentry and Astral as examples. The post should include an itemized list of the amounts paid and to which foundation and/or maintainers. Finally, the company needs to create a JSON file with details about its commitments, submit a pull request to the project's GitHub repository, and supply its branding materials. Once that is complete (and the pull-request is accepted), the company will show up as a member.

What counts

Some might have questions about what counts as a payment to maintainers. For example, does a payment to an outside developer working on an open‑source library to improve it for a company's use count toward the Open Source Pledge? Does it count to pay Red Hat or SUSE for subscriptions to open‑source products? The project site does not get into specifics, currently, but the topic is under discussion in an issue on GitHub.

The general consensus is that "no strings attached" donations that provide little direct benefit to the donor are acceptable, while payments for services rendered do not count toward the pledge. A donation to, say, the OpenBSD Foundation would count—even if the company gets an indirect benefit like public recognition for its donation. Targeted donations to a foundation or group that result in specific benefits for a company would not count. For example, I asked specifically about the Linux Foundation and whether a company's membership in that foundation would count. He said that the goal was to pay maintainers, and that while the Linux Foundation does pay some kernel maintainers, its budget is primarily focused on marketing and events so that would not count towards the pledge. Nor does having open‑source developers on staff. Whitacre said that it's valuable and good when companies hire people to work on open source, but the goal of the pledge was to "add a zero to every [open-source] foundation's budget" to allow the foundations to pay developers.

During my conversation with Whitacre, he acknowledged that there were many what-ifs and edge cases, but generally if money was exchanged with expectations of specific work then it's essentially a company buying a service or product and that does not count. However, he did welcome a debate. "If we can get everybody arguing about what it means [to participate in the pledge], then we win."

Launch

On October 8, the project is planning a promotional campaign to raise awareness and bring more companies into the fold. This includes, Whitacre said, "three of the most expensive billboards in the world" on the Bay Bridge in San Francisco, and running an ad on the Nasdaq MarketSite in New York City. Companies that join the pledge before the launch will be recognized with an "innovator" badge to denote their early involvement.

More than 20 companies have signed up for the pledge so far. Sentry has the largest number of developers, 135, but most of the companies that have signed up so far have fewer than ten developers. The amount pledged per developer varies widely, with one company (Frontend Masters) paying $10,000 per developer for six developers. Whitacre said he would be thrilled to have the first 200-developer or 1,000-developer company join the effort but knows that would take time. He said that he had spoken to people at a few larger organizations but knew that it would take much longer to be able to land those companies.

Can it work?

The project is banking on what Whitacre calls "social validation" to nudge companies into ponying up for open‑source development. Essentially, the group is pinning its hopes on societal pressure to persuade a critical mass of companies into participating and paying into the open‑source commons through foundations or other means.

It is unlikely to be the solution to paying maintainers, but it might be a significant piece of a solution if the group gets traction. It will be interesting to see if this approach works, and if companies are susceptible to social pressure where they haven't responded to other efforts to help sustain open source.