Tangentially related, but I've been feeling like perhaps we've been jumping the gun with Lemmy and Kbin...
Maybe it would have made more sense, at least from an interoperability statepoint, to make an activitypub User protocol that can be selfhosted easily. So instead of making a new account for each Lemmy or Kbin instance someone has, you can instead connect your activitypub account to it.
Additionally as I know hosting has been an increasing problem around here, perhaps it would have been better to make instances single focused message boards ( i.e. just political humor, just animemes, just 196 etc.) So that the instance host can more easily manage moderation and hosting costs without ballooning things they might not necessarily care about to also be hosted by them.
This is just idle musings though, I'm not sure of how the community would be receptive to such thoughts.
Many end up think along these lines at some point. It’s not you! Truth is activity pub is best thought of as an experiment. It’s got flaws as a system, and its recent growth is, IMO, somewhat awkward given the long term game. It would have been much nicer if the fundamental system were more solid.
Many who’ve moved platforms from big social to Fedi, especially the non tech ones, would not be so happy to move again soon. They’re used to a decade or longer on Twitter and Facebook. So any new and improved protocol would split the user base and so struggle to gain traction. Same goes for the platforms probably.
This interview from a while ago now of a developer that has made a parallel universe to the fediverse all in their own it seems is pretty eye opening on hotter tire whole thing might be under engineered: https://medium.com/we-distribute/got-zot-mike-macgirvin-45287601ff19
To me, it seems pretty dated in what it was trying to achieve and shows it’s age in the current climate of privacy, safety and moderation concerns.
They’re used to a decade or longer on Twitter and Facebook. So any new and improved protocol would split the user base and so struggle to gain traction.
Could that be an inherent issue, no matter how you approach it? The issue might be that any social IT system which gains traction is already 10 years outdated. If true, it would mean we could start making something new now, which might be popular in 10 years, but then unable to effectively deal with whatever people deem essential in 10 years.
I’d imagine there are technological foundations which can grow into new demands without breaking things or starting anew. The internet seems to have done this well. I think there are good questions of activity pub in this regard. I do hope the protocol can be updated without having to start again.