jb55 on Nostr: because notes are so easy to replicate across the network, we will eventually have ...

because notes are so easy to replicate across the network, we will eventually have lots of different tech to retrieve the note even if the relay is down.

since all damus clients and others cache notes locally, there's no reason we couldn't have a dht+p2p network for gossiping+negentropy syncing notes between clients.

this would be a nostr p2p network that enables eventual consistency. this would also greatly enhance censorship resistance. imagine if all a large chunk of relays went down and then clients switched into p2p mode.

This is not a replacement for mechanisms like outbox, but simply as a fallback for when you *really* can't find the note, having it as a fallback would reduce gossip spam as well.

or maybe this just becomes an alternative, experimental transport for fetching notes. it could also work over local p2p links like bluetooth and multicast.

there are lots of fun ways we could make nostr more robust to network failures and censorship. we just need to try it!

quoting
note15pe…3mpk

i like the idea of dht + gossip as a layer for nodes that are trying to find specific ids after the source relay is down. p2p for eventual consistency...