Open b1conrad opened 3 years ago
A) we're currently using just the random string that might appear after did:sov:
when using Indy (which we aren't currently) as if it were a DID
B) we're currently building an ad hoc DID document rather than a proper did:peer:
document
C) that the io.picolabs.did
ruleset probably should be part of packages/pico-engine/krl
because we could want a peer DID in other applications besides ACA-Pico.
So, there is a refactoring opportunity, along with the proper use of a real DID, a peer DID, instead of the pseudo DID we are using now. It wouldn't break our conformance with DIDComm version one, because there the DID isn't actually used (rather the private key is used) in connections.
Phil pointed out that the identifier is connected to a pair of keys
Specifically, the DID is generated from the initial public key. The keys can later be rotated, but the DID is not changed. The binding between the DID and the initial public key protects the initiation event.
We are currently just using a DID-like identifier, and minimal DIDDoc, but are not following the peer DID spec.
Our "DID" looks like "P5H3STwNjrhJ7jtEXchzEt" and the DIDDoc is incomplete