Imagine using the Decentralized Lists custom NIP to create a list of “pubkeys whose npubs I’ve verified with the owner in detail in the flesh in real life”. Maybe a shorter title but the description could be as detailed as desired.
Then use Trusted Assertions to fetch the “rank” metric and using that to curate the list.
There you go, decentralized proof of personhood.
This is so meta. This nip ALMOST specifies a method by which individuals can define their own content types (for others to publish content on Nostr) simply by publishing an event. No nip needed for simple new kinds to interoperate across clients.
I’m writing up an article that describes how a client like tunestr can use this nip to do exactly that - let users define content types.
Example: tunestr create a category called Musician Categories. Then users add items to this list to create content types:
- Country Music Artists
- Pop Artists
- Jazz Artists
etc
Each of the above is a kind 9999 note which specifies that items expect a p tag (the musician). Your grapevine decides which Artist Categories to accept.
Users also contribute individual musicians to each category, also using kind 9999 notes. Other users endorse or reject items using kind 7 reaction events. And your grapevine users all this data to decide which list items to accept / reject.
This kind of content curation requires lots of thought on the part of the app devs. You want to make it complex enough to fulfill the needs of the community but not too complex to make for a bad UX.
Example: you could create categories of “musical experts” with exceptional taste in music, then give those experts more control over content curation. Then maybe create a metric that ranges 0-5 for how much control to give to each expert. Is this too complicated? Not complex enough? Are these the wrong lists to curate? What are the right lists? Who makes these decisions, anyway?
A case can be made that a client like tunestr should run its own WoT service (which could be a Brainstorm instance, or could be a competitor service which I hope will exist someday), which will calculate personalized metrics that the tunestr community decides are worthy of being calculated. The leader(s) of the community will carry a lot of responsibility deciding which details to delegate to individual users and which decisions should be handled by the community leaders.
Assuming the users and leaders can be found, GrapeRank provides a clear and detailed path to implement all of the above. I don’t know of any other protocol for which that can be said.
