I do think we can implement nostr. It's just that it's not relevant for the initial phase which is about achieving user adoption. I think nostr is great for handling identity long-term, but that would require an already popular platform educating users about nostr rather than the other way around.