Closed AlexAndrei98 closed 1 year ago
@dhh1128 -- the motivation of did:peer:3 is to have the capability/ease of use of did:peer:2 (in DIDComm v1 and v2), without always having to have the super long, info rich DID. A party could send a single did:peer:2 (thus transmitting the DIDDoc) and then send a peer:did:3 with every subsequent message.
Practically, we also want to transition from the current unqualified peer DIDs (with no formal DID method) in Aries to did:peer:3s, but canonicalizing the existing DIDDoc to a did:peer:2, and then converting it to a did:peer:3.
From an implementation perspective, support would be needed for "synonym" DIDs -- finding a connection based on either DID format.
in separate issue/PR, I would further suggest that we deprecate did:peer:0 (always use did:key) and did:peer:1 (no canoncialization defined, and the DID rotation support is "unwieldy"). Instead, we should use DID rotation, which I think is already supported in DIDComm v2 (right).
Hi, @AlexAndrei98 . Thank you for the PR!
Can you provide some background on the goals that motivated this PR, and how it relates to goals @TelegramSam might have had when he wrote method2? This will help me to know whether I should just accept the PR directly, or try to look for ways to align methods 2 and 3 more closely.