Gregory's ServerThe W3C Decentralized Identifiers standard is now a 1.0 draft.
https://www.w3.org/TR/did-core/
(I only just heard of it today, I have no opinions on it right now.)
 6 comments | Expand all CWs
@darius The Tor DID spec actually looks pretty interesting tho: https://blockchaincommons.github.io/did-method-onion/ since it's self-validating (the DID uri is literally an Onion site url aka public key) and builds upon an existing decentralized registration system that does not require use of a blockchain. And there's a web-based spec as well, which appears to use DNS for validation. @darius Anyways, in the end, excluding the blockchain junk (I'm not gonna pay a cryptocurrency speculator in order to prove my own identity!), if the DID format does take off it's almost certainly going to be because a significant number of people get identities through a small number of big identity providers. :/ @darius yep, i suppose so the big problem i see with DID as it stands right now is that there's no generic protocol/resolver for these theoretical large identity providers to use - the current spec would probably have each provider end up creating their own incompatible DID method, and different resolvers would support different subsets of methods. It's a big mess. @darius worked at a startup doing identity verification stuff a few years back. founders were very into blockchain-y things. very into DIDs. had meetings about them and how we should use them to revolutionize identity. none of us knew what anyone was talking about. to summarize: i still have no opinions on it. |
@darius Hmm. In terms of a generic data format, sure, that spec seems fine but kind of boring. It doesn't say anything other than "here's a URI format", "here's a document format", and "you have to use a resolver (not specified here) to get from one to the other".
The actual meat of the spec is over in https://w3c.github.io/did-spec-registries/#did-methods and the heavy association with proof-of-work cryptojunk really turns me off :(