[cyphar]

Play-democracy is a well-known problem in a lot of projects. OpenSolaris had the same issue, and it had some pretty disastrous consequences (though they modelled themselves after Apache which isn't really the best model).

Honestly, I think it comes down to the fact that democracy is definitely sexy. The idea that any individual could gain the power they want through based on public recognition, where the public has a say on who is in power, is very attractive (and is one of the reasons many governments are democratically elected -- because there would be an uprising otherwise). And obviously it's great for societies where the decisions can very much be a matter of life or death -- but for projects that are trying to get a very specific thing done you can end up in a death-by-bureaucracy system.

Also, democracy is one of the easiest ways of getting out of having to solve an issue -- just let the people decide. There were several completely different PEPs that outlined new governance models, and due to lack of consensus this model was adopted since it can be used to get any other model (and in the Preamble they state that this new governance model could be used to pick a different one).

EDIT: "The people" could be any subset of people -- not just the general public or users of software. Several democratic systems (Westminster-like and the American primary election system) work in the same manner -- where only a subset of people make decisions about the country or party.

[kibwen]

I suspect you're being downvoted because these models don't particularly resemble democracy, which would presumably involve having the users of the software vote, rather than the developers of the software.

[cyphar]

In Westminster-style democracies, party members elect their party's leader (the general public don't get to decide who the leader of a particular party is).

In America, you can only vote in primaries (that is, choose presidential and other candidates) if you are a registered voter of that party -- which means that it's also restricted to a subset of the public.

So I'm not sure I agree it is a mismatch to most democracies. Obviously there is a difference, but it's still clearly democratic in spirit (and actually matches how some real democracies work in some aspects).

[sciurus]

Nitpick, but like many things in America that varies from state to state. Many have "open primaries" where you just show up and request whichever party's ballot you want.

https://en.m.wikipedia.org/wiki/Open_primaries_in_the_United...

[kibwen]

I'd say this looks to be conflating "voting" with democracy-in-spirit. The OP mentions that about 90 people participated in this vote, which is less than a slim fraction of Python's hundreds of thousands of users. This isn't analogous to a representative democracy, because that would imply some sort of election to appoint the "representatives", which wasn't the case.

I'm not particularly interested in arguing about what democracy "is" (we'll be here all day, and then some), but I think it's clear that this model is what democracy "isn't". And I'm not saying that's a bad thing (certainly different styles of governance are suitable for different contexts).

[cyphar]

I would argue that the people most impacted by the steering committee's decisions would be the core team -- who are the people that voted, and thus the steering committee model represents the will of the core team.

Obviously users will be impacted, but in such a tangential way that I would argue that it'd be more like how other countries are impacted by the decisions of a democratic country's leadership (and you wouldn't argue that Canadians should have the right to vote in American presidential elections). Just like the Linux CoC, I don't understand why people who don't contribute to the project should be involved in how the project's development is run.

[C1sc0cat]

Even Athenian Democracy (which had a restricted electorate) didn't have every one vote on every thing.