Re: New CR Snapshot

On 5/29/21 2:42 PM, Manu Sporny wrote:
> On 5/28/21 3:01 PM, Brent Zundel wrote:
>> Feedback from our first CR has shown that it is possible some normative 
>> features may not have implementation support in time for them to be
>> included in our final Recommendation.
> 
> This email details the features that might not have implementation support by
> the time we enter the Proposed Recommendation phase. I have attached the
> latest implementation report for reference.
> 
> There are 47 normative features that do not currently have at least two
> independent and conformant implementations. Yes, that's a lot of features to
> not have independent implementations of by this time in the process... but
> there's hope!
> 
> Many of these unimplemented features are in the DID Resolution section and we
> are told that the Universal Resolver will be implementing a number of these
> features bringing the total count up to two implementations. If that fails to
> happen, we have wording that will allow us to remove features that are not
> implemented during CR2.
> 
> However, there are features with zero implementations and even if the
> Universal Resolver implements the following features, it won't be enough and
> the features will have to be removed from the specification:
> 
> * 7.1.2 DID Resolution Metadata - invalidDid
> * 7.1.2 DID Resolution Metadata - notFound
> * 7.1.2 DID Resolution Metadata - error
> * 7.1.2 DID Resolution Metadata - invalidDidUrl

Should these be marked at-risk for CR2 in case they don't get
implemented? I'd like us to be careful to avoid CR3.

> 
> Some of the zero implementation concerns are in the JSON Production section,
> and we expect the DID WG to resolve to support these features even without
> implementation support (as a general principle to have this support in the
> Abstract Data Model):
> 
> * 6.2.1 JSON Production - datetime
> * 6.2.1 JSON Production - integer
> * 6.2.1 JSON Production - double
> * 6.2.1 JSON Production - boolean
> * 6.2.1 JSON Production - null

Should these or the JSON section be marked at-risk to prevent a
potential CR3 (and charter extension) should the basis for accepting
these without independent implementations be rejected for any reason?

It seems we should be careful here. We should probably mark anything
that didn't get implementation support in CR1 as at-risk in CR2 -- even
if the WG is able to develop acceptance criteria on some other basis
post CR2.


-- 
Dave Longley
CTO
Digital Bazaar, Inc.

Received on Saturday, 29 May 2021 20:06:24 UTC