search

decentralized-identity / peer-did-method-spec

A rich DID method that has no blockchain dependencies. The verifiable data registry is a synchronization protocol between peers.
https://decentralized-identity.github.io/peer-did-method-spec/index.html
Apache License 2.0
27 stars 17 forks source link

Remove the editors note 2022 from the specification per issue #47 #48

Closed swcurran closed 1 year ago

swcurran commented 1 year ago

See #47 .

Signed-off-by: Stephen Curran swcurran@gmail.com

dhh1128 commented 1 year ago

Hey, Stephen. I don't have a strong opinion about this. I wrote the text that you're proposing to remove, but I did it at the request of some DIF folks who felt like the spec was stale -- not because I felt a compelling need. So I guess I could go either way. Maybe we need input from @bumblefudge ?

KERI went into production last week, but you're still correct about its status in the sense that did:keri is not currently separated from the larger KERI ecosystem as a standalone DID library.

TelegramSam commented 1 year ago

I'd be in favor of a more nuanced deprecation notice that focuses on the potential of keri to replace method 1, while noting the utility of method 2.

bumblefudge commented 1 year ago

A KERI-Light stand-alone library that has feature parity with DID:peer in Aries usecases would make that nuanced deprecation notice much easier to write without offending any one either side!

bumblefudge commented 1 year ago

And before anyone volunteers me, I'm not familiar enough with that use-case or the test vectors to provide the nuance here, hehe

swcurran commented 1 year ago

Can we get this discussion going again. The did:peer spec is (to me) extremely relevant, and there is no (AFAIK) "KERI-lite" equivalent that might be available. A KERI did method that is intended for public (globally published) DIDs, but I'm not aware of any work towards the use of KERI-based DIDs being used in a peer to peer scenario. AFAIK, there is nothing to link to.

bumblefudge commented 1 year ago

I'm guessing the IETF internet draft from May of this year isn't the light/peer version you're looking for? It's written as though it covers p2p use cases, but I'm not sure where to look for a description of how to generate and handle logs and prove root-of-trust in ephemeral/p2p contexts. i'm very out of the loop here but would love to read up on where this landed if anyone has links.

swcurran commented 1 year ago

No, as it is incomplete. Lots of terms undefined, and does not have enough detail to create a practical implementation.

dhh1128 commented 1 year ago

I think that new work on various KERI trajectories (e.g., did:webs) mean that the deprecation language should be retracted. Eventually we'll have something better to say, but in the meantime, the notice is creating confusion. So I vote to take it out. I'll leave the PR open for a day and a half to see if contrary comments come in; otherwise, I'll merge this at EOD Friday.