hodlbod on Nostr: I think (as usual) that more relay de-commodification will help with this. This is ...

I think (as usual) that more relay de-commodification will help with this. This is counter-intuitive, because it means that for each use case there are one or two choke points that handle most traffic.

- Looking for a profile? Use purplepag.es or relay.nostr.band
- Searching? Nostr.land or relay.nostr.band
- Looking for safe notes from your network? Use a wot relay

A new category of relay might be "newcomers welcome, spam managed by heuristics". A relay that's really good at assessing whether notes by a fresh pubkey are worth showing. Clients then would hard-code this relay when publishing notes from newcomers or fetching notes from new accounts, just like they already do for search/indexer relays.

Obviously, it would be ideal for many instances of each relay type to exist, and for discovery of these relays to be automatic rather than hard-coded, and I think we can get there, but it will probably require defining these use cases and advertising support for them via nip 11 or something.

quoting
nevent1q…pp3z

A new user who is getting shadow-banned from nostr:

nevent1q…kvzh

How can we improve this? I think the main fault is with primal here (just because pbeta (nprofile…c4d0) is using it), for blocking notes from its own users, and not showing relay publish failures (maybe?). But the criticisms apply to all clients.

Probably the root issue is that relay policy is hard to infer via nip 11 and notice/ok messages. It's also one of those unglamorous chores like error handling that no devs (except go devs) enjoy.