You said "data on nostr becomes even more difficult to recover over time, and eventually irrecoverable if not hosted elsehwere, or regularly reposted"
As well as:
'Storing data on nostr long-term is too high maintenence'
Given that what makes a relay a relay is just a handfull of query and indexing conventions as described in NIP-01, but other than that just a server. What about that all the sudden impacts the ability of a computer to store data? What makes other computers possibly so special that they can do what computers running the nostr conventions cant?
Given that Nostrs defines nothing about the server, other than these interaction conventions, could you not just apply to that server whatever it is in mind you have for this special alternative-non-relay-server?
(Nothing is stopping someone to run a relay with a tripple redundant geographically distributed tape storage back-end storing all the things to the heatdeath of the universe, a thing big cloud providers have been providing since forever)
