Several projects I’m thinking about. One in particular nostr:npub1jlrs53pkdfjnts29kveljul2sm0actt6n8dxrrzqcersttvcuv3qdjynqn and I were discussing in a thread a few days ago. In my mind, I’m calling it Grapevine Communities, and it will build on the Communities feature that already exists in Coracle.

The idea of a Grapevine Community is that you will delegate to your Grapevine the task of managing not only your list of Communities, but also the lists of pubkeys that are associated with each Community. To use this feature, Alice goes to grapevineCommunities.com (or whatever we decide to use) and attests that she endorses (or blocks) Bob to curate her Communities and associated pubkeys. In the future, she’ll be able to attest in a Community-specific fashion, e.g. she trusts Bob to curate the pubkeys associated with the Nostr Developers Community. But for initial rollout, her attestation will endorse Bob to curate all of her Communities. She will have a control panel that provides an overview of her Communities in their current state plus an indication of the updates that are suggested by her Grapevine, which would include: new Communities to add; existing Communities to delete; and changes to the pubkeys for individual Communities. Alice will have buttons to implement those changes, with options to implement all of them at once or just one Community at a time, maybe after she reviews what those changes are. In the future, we can provide a service so that this happens automatically without the need for Alice to visit the Grapevine Communities site each time she wants to update her lists.

In separate pages, Bob and other users will submit Communities to the Grapevine, endorse or block specific Communities, and endorse or block specific pubkeys to each Community.

If I understand correctly how Communities is currently implemented in Coracle, we may not need to make any changes at all to Coracle. Perhaps in the future, we could expand this service to allow the Grapevine to manage mute lists: one for NSFW, one for bots, etc, and that would require some changes to Coracle.

This will be a feature-rich site, with more to add in the future, and design is going to be essential so users understand what’s going on and don’t get overwhelmed with the complexities.

There are a lot of elements that need to be in place for web of trust to take off, among them:

- explicit attestations of trust, not proxy indicators

- trust needs to be contextual

- attestations are used to calculate influence scores, which are also contextual

- trust in a broad context should apply automatically to its sub-contexts, otherwise UI is too cumbersome

- there must be a way to track confidence, eg my Grapevine tells me Bob is an expert in X, but confidence is only 5% bc the estimate is based on scant data

- your WoT manages the contexts and their hierarchies

I’ve built a proof of concept (see link) that demonstrates most of the above elements, so I know it’s possible to achieve all of the above. But design and UX are bad, so it’s time now to focus on deciding what product to build first and how to make it awesome.

https://github.com/wds4/pretty-good/blob/main/appDescriptions/curatedLists/overview.md