Re: did:did

Thanks all for the helpful feedback and suggestions. Clearly the community has been thinking about these and related issues for a long time - I have lots to learn.

On Fri, 2 Apr 2021 20:51:46 +0000
Pamela Dingle <Pamela.Dingle@microsoft.com> wrote:

> Improvement request:
> 
> While there may be no delete operation in the spec, it might be
> helpful to know that a did:did is not in use - I propose an
> extensible custom operation called "Done", that denotes intentional
> retirement of identifier.   Ideally that operation should either
> return the strings "good" or "bad" so that verifiers could know
> determine whether a given did:did DID done(good) or whether that
> did:did DID done(bad).

Good idea - I hope the underlying DID methods adopt this as well.

On Fri, 2 Apr 2021 09:45:50 -0700
Kim Hamilton <kimdhamilton@gmail.com> wrote:
> 
> Bug reports:
> 1. As a user, I would expect the “identity” method to be idempotent

It's true that as a transformation it doesn't return the same DID - and the DID document is not the same as the original DID's DID document either. Maybe the title should use the term "Identifier" instead of "Identity"?

> 2. As another user, I would expect undid and did to commute, which is
> not necessarily the case at the moment

That would be interesting to make work. There are definitely possibilities for interop and collaboration with the undid community.

> We’re getting there guys, keep pushing!

I hope so!

On Fri, 2 Apr 2021 17:15:41 +0200
Markus Sabadello <markus@danubetech.com> wrote:

> Okay.. So this is just for fun, or could it actually serve a concrete
> purpose (even if it's some edge case)?

It demonstrates use of the DID controller property, including transitively. This could be useful for testing implementation support of the DID controller property.

> I'd maybe add a sentence to the spec to clarify this... It looks cool
> :) But the motivation wasn't quite clear to me...

Thanks. We will try to address this in the next version. :)

-- 
Charles