… and small server operators? This is a gold nugget right here. This is pretty much the test suite for fediverse governance. Maybe even its Prime Directive?
Can you elaborate on that?
If they help to funnel more funding to smaller projects, and to maintenance as well as invention, and to developer work outside of coding too, all of that could be helpful.
Aral Balkan got the conversation going about questioning the term “user” AFAIK, and a lot of us have been influenced by this. I think there are legitimate use cases for “user”, eg where the “user” is not necessarily a person but could be a piece of software or a remote system. But once you get into the habit, it’s remarkably easy to replace most uses of “user(s)” with person or people, and in the context of social software, it does help to keep our eye on the ball.
Or phase shift into a new form, for new times. But yes, as the late David Graeber pointed out, many social movements achieve their goals far sooner than they expect, and collapse before they can agree on more ambitious ones. That’s definitely a risk for us at this inflection point.
I was going to argue with you, and then I realised we’re in heated agreement 
You’re right, there’s no point getting into the nitty gritty of where the plumbing goes, and which fittings to use, until you figure out which rooms need a water supply, and what for, and where in the building they’re going.
On that note, thanks for the bump @bumblefudge . I agree it could help to get more eyes on Fediverse Ideas, get a broader range of fedizens creating and commenting on issues there.
Is it useful to have evangelists there, with only the vaguest second-hand clue about implementation (ie people like me)? Or is it specifically a place for power implementers to jam on specs?
It might also be easier to pitch it to people who don’t want to install 10 different email apps because there are 10 different email protocols, and people are fragmented across the resulting jumble of email networks 
Come on Steve, that’s just BlueSky’s initials. Sure is convenient to people like me who think it is BS, in the sense you’re hinting at. In the same way it’s convenient for people who think SubStack are fascist sympathisers that their initials are SS. But just using those initials isn’t denigrating anything.
You going to check for dirt under their fingernails as well? Come on. Even Stallman is OK with having a presence on FarceBook or other Walled Gardens, if its purpose is to be punch a hole in the wall and show people how to get out.
There are those people. Also those who were preparing a bolthole, just in case, or thinking they could reserve their familar @handle, not understanding the first thing about how the fediverse works. But, that’s actually normal for social networks, and new web services in general. Account creation is always far higher than retention.
As things stand, MAU is not a better metric. Because the current implementations of NodeInfo, by which those stats are gathered, are totally whacky. As I understand it, accounts only counted in MAU if they logged in that month. Most of us leave our account logged in all the time, on all the apps we regularly use. So we’re only counted in MAU if we happen to try a new client that month, or have to log out of one of our apps and log back in for some reason.
Also, at least the fediverse has some stats collection, patchy as it is (and it’s an area that could do with some work). Are there even stats available on how many people have signed up for BlueSky or Nostr and abandoned them, or never used their accounts?
That’s true, but it’s an inherently unreliable connection. Like calling a friend on one phone, and another friend on another phone, putting one to each ear, and trying to relay messages back and forth. Far better to have a system where you can give them each others’ phone number (or @person@server.foo address, or whatever the identifier is), and they can call each other directly.
At the moment there is one bridge from the fediverse to BlueSky. One. Two to Nostr. One of which is blocked by most of the verse, and the other is also the BS bridge. It’s also a bridge to the IndieWeb.
So what would happen if a significant proportion of people in each network started talking across those bridges? I presume you’ve heard the term Single Point of Failure. Or Fail Whale, which is a specific example of one. The whole point of a federated network (or a decentralised network of any kind) is to avoid all that. Bridging is a stopgap, at best. Unless we get people running hundreds of them instead of running servers that host native accounts, in which case we’ve basically reinvented Nostr 
In the order you asked the questions; FEPs, yes, FEPs. Then as @aschrijver says, feed the output of that through to the SocialCG, and from there into a new version of the W3C standard as necessary. A formal standards process is the documentation process at the end of a period of experimentation, not the beginning.
100%. Give this man a coffee (or a joint, or a vodka, or a cigar, whichever he prefers) 
Fair point. But you seem to imagine that writing a standard is like waving a magic wand, that will cause all the implementations to start following the spec. The degree to which AP implementers follow the current AP spec suggests this is not how it works in practice, and there’s no logical reason why it would. A standards process is a rough consensus process among existing implementers or it is nothing.
The the best case scenario, this is something the SWF will help with.