Is there a good overview of the federated model that explains things like: What stops the nazis from making their own instances and making the federated timeline unusable? How can people compare instances for best fit? What happens to your account information when an instance goes down or dies?
Or are we making this all up as we go? Are there examples this is mimicking?
@tjvr Agree that helps me think about it, but no federated email timeline or all email from your domain timeline, which is quite different.
@nancybaym In re: Nazis — I believe other instances can block/blacklist any given instance. Granted, that means there would have to be a concerted effort amongst “good” instances to blacklist “bad” instances, but I imagine that would not be an issue, since the “bad” instance would make itself pretty well-known enough to make easy any such effort.
@nancybaym My understanding is that your own instance only shows instances that it's federated to on the federated timeline.
So if your instance doesn't federate with the nazi instances, you'll never get to see 'em.
@nancybaym Nazis are free to make their own instance. But every other instance is also free to blacklist their instance and so ban them from the federated timeline.
And even that becomes only necessary if someone from the non nazi instance follows someone from the nazi instance.
@nancybaym this is a good overview on how gnu social works
https://robek.world/featured/what-is-gnu-social-and-is-mastodon-social-a-twitter-clone/
this is discussion among the admins of instances about potential for pollution in federated timeline and solutions for it too, a blacklist is being used already. but other thoughts are happening
@nancybaym My facile understanding is that it's similar to TPB proxies: front ends running the same protocols allowing end users to share the same ad hoc network.
@nancybaym email is the closest example I think