Oddbean new post about | logout
 I don’t have an intuitive feeling for bandwidth and storage requirements. I suppose storage depends on archive policies, but can this system scale? 
 We are compressing older media, so far not a problem. If we do a 10-20x growth we might need to revisit some architecture, fortunately Fishcake has experience scaling enterprise and service provider systems, so sure we can handle it 🤙 
 Is there no mechanism for coordinating between nodes who keeps a copy of what? If all node store all things, the system by definition scales poorly. 

I haven’t a clue how nostr actually works.