"I'm going to assume you understand atproto pretty well. If not, there's plenty of good reading out there already;
- [ATProto for distributed systems engineers](https://atproto.com/articles/atproto-for-distsys-engineers) - atproto.com
- [How Does Bluesky Work](https://steveklabnik.com/writing/how-does-bluesky-work/) - [@steveklabnik.com](https://bsky.app/profile/did:plc:3danwc67lo7obz2fmdg6jxcr)
- [Nostr and ATProto](https://shreyanjain.net/2024/07/05/nostr-and-atproto.html) - [@shreyanjain.net](https://bsky.app/profile/did:plc:bnqkww7bjxaacajzvu5gswdf)
Having read these, or pretended to, you can recite how data flows from a client to a PDS through a relay to an AppView. After that, though, how does all this *actually* happen? If you were a living, breathing Bluesky AppView, what specific HTTP endpoints would you need to frantically `curl` and what functions would you need to implement? To the best of my knowledge, this information is decentralized across a number of code comments, blog posts, Bluesky posts, and Discord messages, so I'm going ot try "