Sven Slootweg (soft-deprecated) @joepie91@pixie.town

@kim An additional issue with that approach is also that it'll disadvantage any late additions, because there will be less people who can vote for them.

I think it's also worth asking what purpose a vote is meant to serve here to begin with. What question is it supposed to answer, if your goal is to find some outcome that everybody can be happy with?

Like, say that 80% votes yes and 20% votes no. That's a majority "yes", but that doesn't tell you anything about *why* people voted "no", or whether those are addressable concerns.

And you can *ask* people what their concerns are, but if you do, then what is the value of knowing the 80/20%? Does it make any difference from 90/10% or 99/1%? There's still people unhappy and concerns unaddressed either way.

Voting systems are extremely low-fidelity and ultimately are designed to serve "oppression of the majority" type models, where the options are predetermined - that's the only context in which these numbers can make sense, as far as I can tell. Do we actually have any purpose for that here?

Or to put it differently: if there *wasn't* a vote, and *only* a consensus building process... what information would we really be missing? What problem would that really cause, that justifies introducing voting?

@kim My experience is that as soon as you offer anything that looks like voting options, people assume "oh the talking stage is over" and it's almost impossible to get a good, in-depth engaged discussion on the matter

@kim But yeah the crucial difference between a vote and consensus building is that the former pre-defines the available options regardless of whether any of them are good, whereas the latter is a process to find the optimal solution without assuming that it's already in the list.

It's a far more inclusive process, that doesn't run the same risk of marginalized needs and perspectives getting overlooked.

@kim (Obligatory "bigots can get fucked" of course, consensus building is a process that only includes those who act in good faith, even if there may be a disagreement between them)

@kim Short version: a process of working towards a general universal agreement of *some* sort by talking through everybody's needs and concerns (and strictly expecting constructive discussion patterns around that only), to figure out some solution that can work for everybody.

That does not *necessarily* mean a compromise - sometimes the answer is a compromise, sometimes the answer is another new third solution, sometimes it's just a matter of correcting some misconceptions and concerns. Consensus building is the process of figuring that out collaboratively.

@kim Consensus-building is probably gonna be a better approach there than voting tbh

@eevee iirc there's some way to disrupt their 'smell' and lay a false trail that leads them back outside, but I forgot what it was

@zens@merveilles.town Jinja was just one example of many - almost every templater I've evaluated has similar design issues :/

(Mustache partials *do* actually have fairly clean scoping rules IIRC, unlike most templaters, but I dislike using it for other reasons)

@zens@merveilles.town I answered that in the toot you're replying to...

@zens@merveilles.town I do use JSX. And of course it's not HTML - just like all the template languages for Ninja and such aren't. It's a meta-format, that *produces* HTML (of whatever specified shape!), which is what ultimately matters.

I don't use (state) hooks server-side - there is no purpose to them, because there is no interactive state management.

I *do* sometimes use contexts, but they are not globals - they are scoped to subtrees, so more closely analogous to lexical scope.

@eevee and we didn't even get the attempts