Oddbean

I dont think anyone did an analysis on this. But a local database is an absolulte requirement for many clients right now. We are all already duplicating downloads and duplicating disk space. This is besides the fact that we all developing our own databases separatedly, which is a huge waste of time compared to centralizing the development in relay apps. Then there is the offline use case. I believe most of our event writting will have to operate off-line. And the app must find each other events if that is the case. Otherwise users will just be super confused.

FWIW I think an “in device relay” is a great idea. I also see the problem with having “multiple” apps to dl, when multiple clients are already a head scratcher. For “most” people the value prop of a “signer” app will be as an “identity manager” for accessing the entire nostr ecosystem (native apps and PWAs). Given this use case, having “personal storage” built into the actual signer app makes total sense. Il even raise the ante, and go so far as to suggest this app (the “identity manager” class of apps) could even act as a nostr on-ramp for profile building, client discovery, and relay management. Maybe on paper these are separate NIPs, but in practice a single “this is me on nostr” app wd be bomb!