How do we scale Nostr? “Unpopular opinion: Nostr CAN'T scale to be what people want it to be. It's an all-to-all broadcast, which scales like O(n²). It will collapse in its own success. Nostr relays are like sidechains and altcoins. It's attempting to solve an O(n²) scaling problem by having multiple copies of the same thing. It leads to Balkanization where you can't see content from or message some people. Blockchains of blockchains were dumb and this is dumb in exactly the same way, and I can't believe all of you failed your algorithms classes.” - Bob McElrath So this begs the question, how does Nostr scale long term? Or do we simply expect 2 of 3 dominant relays in the long term? cc @fiatjaf @jb55 @Derek Ross @npub1pgh3...rs38 @Vitor Pamplona @Sep

Replies (7)

So the idea is have Nostr relays for specific communities? Your point is that people are comparing Nostr to existing centralized system and assuming that Nostr needs to scale the same way. I guess my counterpoint would be that lots of people have been excited about small siloed networks in the past, such as mesh network, but in practice they never scaled to the masses. But maybe that’s fine? Idk what the end vision of folks are for Nostr. Nostr can be whatever people want it to be. A DLC oracle broadcaster, private messenger, X alternative. I’m just curious what people’s vision for it is.
It's not relays for specific communities, although that is also possible. It's just following people wherever they are.
The internet is Balkanized. The Balkans are Balkanized. Seeing everything, all the time is an unreasonable expectation. Nostr is a protocol not a platform.