It's incredibly bad at supporting coherent meta-conversations, and that's not just a problem of the technology.

The model that ActivityPub implementations tend to follow for supporting group interactions, is to let groups emerge organically in an informal manner, as a natural result of the clustering of people who are similarly aligned.

That works really well for a lot of things, but not for conflict resolution *between* those organic groups, whichever they are.

Each group mostly just sees the things posted by 'their' group, and so is very likely to get only half the story of what's going on, if that; and usually through the lens of a member of their group. This means that in a conflict, both parties are operating on a wildly different view of what actually happened.

I'm not sure that you can actually fix this on a technical level, I suspect it's fundamental to this model. If you don't have explicitly defined communities "in" which something happens, there is simply no reliable way to get a complete view of all conversation around a topic within that community. There isn't even a way to convene a meeting to sort things out.

Sure, you have hashtags, and you have group accounts, but all of these are opt-in and so only make discoverable those things which are explicitly posted to them, which is usually only a small fraction of what actually happened, and not the parts that are important to understanding it.

I can't see how *anything* that's built around personal profiles/timelines foremost, would avoid this fate. It seems like a fundamental and far-reaching design error to me, something that practically guarantees unsustainable conflict, no matter how good the moderation tools.

It's also a design choice that basically every social platform since 2010 has made.