Interesting; still an issue without final and agreed on solution!

What's wrong the most?

For real time updates, there essentially are two approaches: 1. Push: server is aware of the clients and maintains direct connections with them; once there is new data, it sends it to them 2. Pull: clients are aware of the server; they periodically ask - is there any new data? If there is - they get it, if there is not - they will ask again anyways As with everything, there are different tradeoffs to both approaches.

For most users, I think that custodian with proton-services-like trust-minimized standard auth flow will be better. Plan to work on the API that allows it 🫡

Exactly - if your product is great, profit is an ultimate sign of virtue

It depends - currently the global position of the dollar make it so that the standard of living and job opportunities for an average American are worse in many European countries. Plus, some countries in Europe are way more peaceful - Poland for example.

I hope so! I fear that incentives to run relays will make it so that there are two or three ones that have 99% of market share; if that's the case, we might end up with decentralization in theory, but centralization in practice. Although that's a completely different topic

Unfortunately, we might end up in the exactly same place as we are with centralized platforms.

NOSTR approach to the Identity makes it fundamentally unrecoverable, once you loose your keys. wasn't it a fundamental mistake? It might mean that it's impossible to create a recovery/rotation procedure that has acceptable UX.

I think that it is quite unfortunate that we gave up on NIP-26: https://github.com/nostr-protocol/nips/blob/master/26.md Ephemeral keys seems to be a natural and smart way to increase security and UX of NOSTR: keep your master key somewhere safe, in a cold storage/encrypted, and generate ephemeral keys for online clients. Maybe we should revisit the idea in a new NIP, improving on it? Because, what's the alternative?