Maintainer Honorarium

[Reqrefusion]

Abstract(By ChatGPT)

The text sets up a trust-based honorarium for FreeCAD’s Core Maintainers to compensate code review, release coordination, and similar work with minimal paperwork. Core Maintainers can join any day; acceptance is automatic upon FPA receipt and lasts until December 31, with a fixed honorarium of €200/month payable only if there’s at least one repo interaction that month (fixed payments for active months after acceptance may be claimed later in aggregate). Participants may also file a single Additional Honorarium Request per month for only the current or immediately preceding month—no timesheet—using €40/hour as a reference. For Core Maintainers, total monthly requests (fixed + additional) up to €1,000 are paid without a vote; amounts above €1,000 require an FPA vote. Ecosystem Maintainers can be covered by the fixed honorarium only via an annual FPA vote (renewed yearly), and all their additional requests always require voting. Requests can be submitted during the month or the next; the FPA pays via an agreed method, and notes are optional but encouraged. The program presumes unconditional trust and does not pause payments due to allegations (which are referred to the Core Maintainers Group); it is financed by an FPA fund that may be topped up or discontinued if insufficient, and it runs on a calendar-year basis with year-end review and possible adjustments.

Explanation

I have been thinking about this for a long time and have looked into practices in both open source projects and academia. The proposal below is a more mature and streamlined version of my original idea: a trust-based honorarium scheme that keeps paperwork to a minimum.

Let me start with the fixed honorarium. This is a common approach, but in many projects it drifts into something salary like, which is not sustainable. As the FPA, I do not believe we can shoulder that kind of ongoing burden. In a few examples I found more workable figures. Taking inspiration from Homebrew’s practice of paying maintainers USD 300 per month, and considering both broad accessibility and the euro to dollar exchange rate, I set the fixed amount at EUR 200. To qualify, I required only a very simple activity condition: at least one interaction within the month. That threshold is intentionally light, and even minimal contribution counts as activity.

I also reviewed how peer review is handled in academia. Token based models did not seem persuasive. By contrast, I saw debates and some implementations, especially in medical journals, around paying roughly USD 50 per hour. Building on that, I added an additional honorarium path for months when maintainers feel they contributed beyond the fixed amount. Here, EUR 40 per hour serves only as an internal reference rate for consistency. No one needs to submit hours; the requester states only the total additional amount.

Everything rests on mutual good faith. A maintainer can receive up to EUR 1,000 in total for a month without any vote; amounts above that go to an FPA vote. I have also left a door open for maintainers from the broader ecosystem of plugins and dependencies. For them, inclusion under the fixed honorarium is subject to an annual FPA vote, and every additional honorarium request, regardless of amount, also goes to a vote. In this way, the core enjoys a smooth, low friction process, while ecosystem participation proceeds with a bit more oversight.

[Reqrefusion]

To do some more brainstorming, there was a mistake in my other proposal, and it was interpreted as if a thousand euros would be given for nothing. But the other proposal was something less than development-related grants. This one is completely different. First of all, I believe we should keep the fixed amount as low as possible. This is to ensure that more people benefit. This is, of course, related to the FPA's budget, but it's also important to avoid pressure from the community. People shouldn't hesitate to think they won't be able to afford the money or to avoid pressure from the community.

Of course, this doesn't meet the actual timeframe, and some months it becomes a significant burden. For example, we're currently seeing this in the CAM workbench. I've added an optional submission system. There's a more realistic fee here, similar to what the US journal deems appropriate for reviewers. And to speed things up, there's a no-questions-asked process for up to €1,000. This isn't just about trusting the maintainers, it's about speeding things up and minimizing the effort required by the maintainers on the submission. Of course, this couldn't be unlimited, so an FPA vote above the $1,000 threshold is required. This is more about budget than trust.

[Reqrefusion]

For simplicity's sake, I've deleted the additional honoraria-related items. Other sections related to how the program works may also be deleted.