The is a great piece by Rich Bartlett of #Enspiral, #Loomio, and #TheHub, talking about the advantages of federations of small groups, over networks of loosely connected individuals:
http://organizationunbound.org/expressive-change/build-a-network-of-small-groups-not-a-mass-movement-of-individuals/

It's reasons like these I encourage community-hosting more than self-hosting. Note: I'm not against self-hosting, for those who have the skills and motivation to set up and maintain it, it's just not my ideal that everyone self-hosts.

5 reasons to build a network of small groups, rather than a mass movement of individuals

Organization Unbound is an attempt to re-imagine the way we think about and engage in social change.

Organization Unbound

@strypey

Ain't no necessary contradiction between self-hosting, hosting by groups, group instances, and participation in groups.

Some of us belong to a lot of groups. We want all those groups to federate. In the meantime, we are federated with all of those groups.

@bhaugen sure, what I'm getting at is the ideal, expressed by some people (usually only online), that the end game is for every net user to self-host everything they use, and for groups to consist of clusters of self-hosters. Many pieces, loosely coupled, as it were. For the reason Rich gives in his piece, I see a network of '#CoHosting' small groups as the end goal, obviously with accommodations made for membership in multiple groups, incorporating self-hosters, and so on.

@strypey

Consider agency, in the sense of being able to make decisions and take actions.

We want both individual persons and organizations to have agency: to be agents.

We think organizations are important. (I personally don't even believe that single isolated humans exist. Even the recently-famous secret hermit stole everything he needed from his neighbors.)

But we also like Greg Cassell's idea about agreement-based organizations. https://medium.com/enspiral-tales/a-developmental-framework-for-agreement-based-organization-54298b128906

A Developmental Framework for Agreement-Based Organization

I’m introducing Agreement-Based Organization, derived from ongoing research, discussion, and collaboration with innovative groups…

@bhaugen consider this practical example: I have family photos and videos in digital form. don't want to lose them due to hardware or service failure, so I want multiple copies to exist. But I want them to be generally private, but shared with my family and maybe some close friends. The security doesn't need to military grade, just strong enough to obscure it from datafarms, and prevent casual cracking by script kiddies.
@bhaugen there are two general approaches to solving this problem. One would be for the family to have a private server running something like #NextCloud, that allows us to decide who can connect to it and share our media, and holds a remote backup. The other would be connecting us all via a more distributed system like #SyncThing, where we all serve as back-ups for each other, but there's no central repository.
@bhaugen but say every family has a NextCloud (or #ownCloud, or #Cozy, or whatever) for its photos and video. So does every social club, or community group, or whatever. Some of the photos hosted by some of them are of public events, and are publicly shared. So I might want to put my photos from a public meeting about cats on the Cat Club server, and sync the ones I'm in onto our family server. So a network of small groups is formed by that kind of interaction.
@bhaugen now you could create the same result, in theory, with a P2P network of individuals using SyncThing and similar systems. The reason I'm skeptical about this is that it requires every individual to be a power user. That doesn't fit with my observations of how people have used digital tech over the last 20 years. Most people want stuff that "just works", and look to a handful of geeks to function as supernodes, showing them what to use, how to use it, and fixing it when it breaks.
@bhaugen maybe we can achieve a level of digital literacy where everyone is a power user, and maybe not. But even if we can, getting to a point where there are 2-3 power users in every extended family/ community, capable of running #CoHosting for them, would be a pitstop on the way to full P2P, and a big improvement on the #DataFarming that dominates the internet today.
@bhaugen more importantly, coming back to Rich's article, if the development and deployment of digital tech is focused on serving federations of trusted communities, rather than being focused on building trustless P2P networks, it contributes to a rebuilding of both community and trust. I tend to think these are both things we need more of in the world. Keep challenging me though, it's helpful to think this stuff through as I try to write about it :)
@bhaugen sorry, my notifications have been a bit flooded and I didn't see your descriptions of what you're working on until after I'd posted my hypothetical example. I think this is an important discussion, but also quite complicated, and one where it's easy to talk past each other. Maybe something to work through a bit more slowly and thoroughly in a thread on the #OAE group?

@strypey

In that case, I am a bit lost about how to respond. Maybe the OAE group would be best.

@bhaugen I think so. It's fun to throw ideas and links around in the fediverse, but I find that focused discussions are best not held in the pub ;-)