I'm not just listing theoretical features, these are things that happen in the network right now. Is there anything I mentioned in my comment that I forgot to give an example for?

Doesn’t matter whether migration is technically possible under ideal conditions because if you’ll need it they won’t be ideal.

I don't see why so many people say migration is only "technically" possible, migration can be done today. If there is more demand for third party servers, say, if Bluesky starts fucking up with moderation more, more third party servers will pop up, because right now the user concentration isn't a technical problem or fault of the protocol. I don't disagree that it's a problem.

And ATProto’s architecture, particularly the cost and complexity of running the more demanding components that need to have a global view of the network, structurally favors concentration at those layers.

It's not necessary to have a global view of the network to participate in the network.
It is possible to have a global view of the network without a relay using constellation, constellation instances are very cheap to run, and work by indexing backlinks. It's what powers reddwarf and recently wafrn (wafrn optionally supports relays as well).

Atproto isn't significantly more complicated than AP, it's just different.

[–] 73ms@sopuli.xyz 1 points 3 hours ago* (last edited 3 hours ago) (1 children)

The features are theoretical in the sense that there is no real guarantee they'd be possible after BSky corp changes their behavior and that they are in use only in the least significant way possible, for tiny and irrelevant numbers of users. But of course this is just restating the obvious again. For a network truly to be shielded against this sort of thing it should be decentralized already before.

See this for how constellation makes no difference.

[–] ricci@discuss.systems 1 points 3 hours ago (1 children)

@73ms @irelephant

I also wrote a more complete thing about constellation here: https://discuss.systems/@ricci/116132447206279469

[–] ricci@discuss.systems 1 points 3 hours ago (1 children)

@73ms @irelephant

In terms of how prepared atproto users are to use adversarial migration, I did a point-in-time survey a couple months ago: https://rob.leaflet.pub/3m7isflo7ls23

I suspect the situation has slightly improved since then, as some or all of the migration tools offer to set up rotation keys for you, but I haven't had time time to do a followup.

[–] 73ms@sopuli.xyz 1 points 1 hour ago* (last edited 1 hour ago)

@ricci@discuss.systems Thanks for the link to the explanation, that thread in general and the survey! All were quite insightful.