Open peacekeeper opened 4 years ago
Quick question - what's the argument against using a query parameter (instead of appending the DID to the resolver endpoint)?
No argument against that, I think this would work just as well. I guess the question is how tight or how flexible we want to define the HTTP(S) binding.
For a DID Resolver exposed at an HTTP(S) endpoint, the current thinking is that one would append the DID or DID URL to the endpoint.
E.g. if a DID Resolver is exposed at
https://uniresolver.io/1.0/identifiers/
and the DID to resolve isdid:example:1234
, then an HTTP GET is executed against the URLhttps://uniresolver.io/1.0/identifiers/did:example:1234
in order to resolve the DID.In cases where a DID Resolver supports only a single method, it may be sufficient to only append the method-specific identifier.
E.g. if a DID Resolver is exposed at
https://example.resolver.com/
and the DID to resolve isdid:example:1234
, then an HTTP GET may be executed against the URLhttps://example.resolver.com/1234
in order to resolve the DID.Discuss if this idea makes sense and if/how it should be incorporated into the spec.