Last time I checked(years ago), it was like Mastodon but there are only dozen servers, and posting once in standard clients post it on all servers. Everything is a tweet, and servers serve you contents based on what it remembers about you. The core principle is that bans and refusals by server operators are at best customary since users can receive the same information elsewhere, even transparently.

Though, they did happily filter Macau casino spams flooding the system, so I wouldn't be so sure. Workload of posting to and receiving from dozen servers was also not trivial when I was trying it out, and architecture changes to reduce duplicates and/or syncing databases across hosts were actively discussed. I guess those works were never completed judging by comments here?

It's very unlike mastodon in that server operators have minimal control over user identities and content. Spam control is still a work in progress, but has come a long way through web of trust and more responsible relay operation. I invite you to give it another try!