Ok so basically Twitter, Myspace, Instagram, and Tiktok mostly rely on "follow sets" which is inclusive of everyone you tap "follow" on. Those are centralized systems with their own proprietary stack of databases and indexing so users can think less about the backend and more about the meaningful interactions they participate in!

Pubky instead relies on homeserver architecture allowing for any method of handling semantic user sets, like "friend lists" or aggregate stastics like follower counts. To my knowledge, the Pubky ecosystem is FOSS, allowing anyone to create their own format for expressing semantic web data (like a nostr event for example!) without relying on simple relay architecture.

Because the Nostr network is an event-driven system of relay infrastructure, it can define schemas for any type of list. However since relays only support "read or write", data cannot be abstracted over-the-wire, meaning every schema must be predefined and interop with relevant schemas.

Cornychat attempts to create semantic relationships between users and rooms using a specific replaceable event kind 30000, which I don't remember what it means, but basically you can think of it as JSON trapped in a giant bowl of spaghetti.

Thank you for coming to my PubTalk.