indeed, there is a number of flaws in nostr's design that would take a restart of the network to fix, but the momentum is in favor of an incremental, backwards-compatibility path forward
just ignore the preachers and their chorus
Discussion
like, it's my opinion the easiest way to fix most of these problems is simply to create new network encodings, then you can put the fixes in them, and then it's just a matter of getting a few clients to adopt the new transport, and adding some small new parts that create requests that line up with the new changes
honestly, the biggest flaw in nostr is the REQ envelope, but maybe almost as big is the entire "envelope" nonsense, which should be changed to use an actual normal API
i think it would be easy to add a QUIC/HTTP/3 with JSONRPC2 and this would make it mesh better with apps that already work with bitcoin nodes as well
Nostr the culture and nostr the tech are different things and as I understood he's not as bullish because of the former
Sure lots of things could be improved
To be honest, I ultimately don't care what works so long as it allows an open sovereign identity and data portability