Coming soon to Coracle: relay reviews based on NIP 32 👀
https://us-southeast-1.linodeobjects.com/dufflepud/uploads/aa237add-97a8-4f8e-ae6d-0512d7e993aa.mov
Discussion
Y'all better give my relays a good review or else 🔫🤠
🔥🔥🔥
How do users really know if a relay is good or not? What makes it good vs not good? Apps don’t show us info that would make us informed on the matter
(Just checking that I’m looking at the right nip-32 since it’s not yet been merged)
Yep, that's the one
Cool. Trying to understand how it works now. It’s making me think of JSON-LD, linked data. I’m thinking a context in JSON-LD plays a similar role to a namespace in nip 32, although I haven’t fully grokked nip 32 yet (and it’s been a while since I’ve thought about linked data) so I could be misunderstanding.
If I’m understanding correctly, a namespace L is basically a list of tags that we can use to look up what any particular tag means? …
nostr:npub1cpstx8lzhwctunfe80rugz5qsj9ztw8surec9j6mf8phha68dj6qhm8j5e
Meaning is entirwly based on convention, but yeah, I plan to create a page with my nomenclature on it that people can reference
Your nomenclature will be custom tailored to the purpose of a rating system for relays, I assume.
I wonder: Could the list of types at:
https://schema.org/docs/schemas.html
fulfill the role of a namespace? (Where a “type” at schema.org = an l-tag label in nip-32.)
And specifically, could
be useful for coracle’s relays rating system?
I’m wondering to what extent a variety of customized namespaces could be made interoperable / overlapping. Maybe schema.org augmented by whatever additional types a dev might need for some specific application.
I’ve been reading through the pull requests in the nips repo related to nip-32 and I can def sympathize with concerns by #[3], #[4] and others, particularly here: https://github.com/nostr-protocol/nips/pull/457, regarding defined vocabularies. We need them, but they are awful in so many ways. I may add some of my thoughts over there, but while the thought occurs to me I wanted to wonder out loud whether crowdsourcing a NIP-32 namespace — which is basically a defined vocabulary, if I am understanding correctly (and I may not be) — might be a potential application of (and potential way to test out the feasibility of) loose consensus using DCoSL.
This is great!!