> there is no model for replacing poor governance

Do you have any model to propose? Because most democratic models you would see for country governance (to which you drew parallels) rely on some key characteristics that don't apply to open source governance, making them not really transferable.

▲

layer8 7 hours ago | parent | next [-]

Debian is an important open-source project that has had codified democratic governance for over quarter of a century now: https://www.debian.org/devel/constitution

▲

JoshTriplett 7 hours ago | parent [-]

One property Debian has, which most projects don't have, is a very clearly established electorate to enfranchise.

A few projects have a clear set of members. Most projects don't; they have contributors of code/documentation/triage/community/etc, but no clear delineation of who is or should be an enfranchised member of the project. Often, projects don't end up defining this until they need it as part of establishing some preferred form of governance.

	▲	hobofan 3 hours ago \| parent [-]
		> A few projects have a clear set of members. Exactly. I'm not too versed in political theory, but one thing from my perception that provides coherence/continuity in nations is the immobility of participants (for better or worse of the individual). E.g. if you have 70/30 election outcome you still have to factor in the needs of the 30% as they may provide important economic functions and can't just leave on election loss. In contrast, in an open source project, even if you can clearly delineate membership and based on that voting rights for a democratic process there is very little preventing the "losers" from forking (it almost entirely comes down to brand). The outcome of that would just be an empty brand shell with a good chunk of their contribution activity gone.

▲

edent 8 hours ago | parent | prev [-]

Most organisation (in my country at least) have a written set of objectives and a legal structure which diffuses power.

Any co-op, limited company, charitable association, etc can provide a good model - depending on the nature of the project.

As I say, it is probably overkill for most OSS projects. But once you get to a certain size, I think it is obvious that you need a way to ensure the project's longevity.

The death or disgrace of a CEO rarely destroys a company. There's a board their to temper their behaviour, a structure to ensure succession, and (most importantly) a set of expectations upon which their community can rely.

I'm not saying that's the only way to do it. I'm not even suggesting it is the perfect way to do it. But I think it is better than hoping the BDFL doesn't implode.

	▲	hartator 7 hours ago \| parent \| next [-]
		Non-profit entity has usually more drama. And, you can still be “dictator” of a non-profit.
	▲	paulryanrogers 7 hours ago \| parent \| prev [-]
		Sadly plenty of boards are filled with CEO sycophants, so even that model is not immune.