Oddbean new post about | logout
 Splitting large relay databases into many servers by event id prefix is an interesting solution.
nostr:naddr1qqxnzdejxqmnqde5xqeryvejqgsflmrj64um42nh9tu7w8nr3dffy9tjrt8xururype9ajle7alctvgrqsqqqa28gnxzlw 
 That is a cool idea. 
 Reinventing IPFS, are we? 
 I like it! Maybe also by frequency of access, so archive relays can have slower but larger storage to keep costs down while smaller, faster, purged relays can keep things speedy. 
 If Nostr succeeds, relays will need viable financial models (e.g., user fees or data monetization). As profitable businesses, relays would then be incentivized to maximize, not limit, note volume. IMO 
 As long as the fediverse offers free servers why would anyone pay to use a nostr server? 

We need a donation model.  
 There is no such thing as “free servers” once you scale, whether it’s fediverse or nostr. 
 I said donation model. And free from the users perspective. 

 
 There is no viable donation model for a network of relays serving millions of users. If nostr succeeds, relays will have to become profitable. 
 The fediverse already serves millions of users via donations. 

I don't see any reason why not. 

 
 @cloud fodder relay.tools 
 Quality is often worth more than quantity 
 No matter the solution consumers will need to pay some portion or else we end up with the same perverse incentive structure as the advertisement model.

With ads the consumer isn’t the paying customer so the service provider’s incentives aren’t aligned with those of the consumer. That led to monetizing the consumer’s attention and a meteoric rise in essentially surveillance tech.

We need to align all interested parties.