Re: Current status of DID Core implementations (June 2021)

Hi.

I checked the following points.

1.  I ran and checked raw output in JSON 
   (which is at did-test-suite/packages/did-core-test-server/report/tmp/did-spec-test-run.latest.json  this file is gitignore'ed)

Only three outputs for the following statement:

> 7.1.3 DID Document Metadata - If a DID has been deactivated, DID document metadata MUST include this property with the boolean value true.

These are following three DIDs:
         did:btcr:x705-jznz-q3nl-srs
         did:btcr:xz35-jznz-q9yu-ply
         did:orb:interim:EiDQF58KdqxovQinUrhplB8P83_F8aMP7tTwLD6UrgVzNA

One test output for the following statement:

> 7.1.3 DID Document Metadata - If a DID has not been deactivated, this property is OPTIONAL, but if included, MUST have the boolean value false.

This is for this DID:  did:btcr:xkrn-xz7q-qsye-28p

Thus, I believe this is not a tallying code problem.


2. Markus listed following files:
> https://github.com/w3c/did-test-suite/blob/main/packages/did-core-test-server/suites/implementations/universal-resolver-did-btcr.json
> https://github.com/w3c/did-test-suite/blob/main/packages/did-core-test-server/suites/implementations/resolver-did-ion.json
> https://github.com/w3c/did-test-suite/blob/main/packages/did-core-test-server/suites/implementations/resolver-did-orb.json

Since `resolver-did-on.json` is in `brokenFixtures`, this one will not appear in test results.

So I think test output and report are in sync.


3. Before checking the above conditions, I set several breakpoints to observe the situation.
    At least, I found the following possibly incorrect values
   (I'm not reading this part of the code well. so I might be wrong)

   - `testDidDocumentMetadata`  may ran empty `didDocumentMetadata`
   - expectedOutcome can be `defaultOutcome` but code checks against with
     equal to `activatedOutcome` or `deactivatedOutcome`
    -> https://github.com/w3c/did-test-suite/blob/992369bda6fd78314a4bae7f969e0a5fb11dbb2f/packages/did-core-test-server/suites/did-resolution/did-resolution.js#L203

Hope this helps.

shigeya

On Mon, Jun 28, 2021, at 16:29, Markus Sabadello wrote:
> I don't think it's true that we are lacking enough implementations for
> "deactivated":
> 
> https://github.com/w3c/did-test-suite/blob/main/packages/did-core-test-server/suites/implementations/universal-resolver-did-btcr.json
> https://github.com/w3c/did-test-suite/blob/main/packages/did-core-test-server/suites/implementations/resolver-did-ion.json
> https://github.com/w3c/did-test-suite/blob/main/packages/did-core-test-server/suites/implementations/resolver-did-orb.json
> 
> Same for "nextUpdate" and "nextVersionId" I think.
> 
> Maybe the test suite isn't counting some implemented features correctly?
> 
> Markus
> 
> On 27.06.21 21:57, Manu Sporny wrote:
> > Hi all,
> >
> > Our latest implementation report for DID Core is available here:
> >
> > https://w3c.github.io/did-test-suite/#spec-statement-summary
> >
> > Here are the remaining items that the WG needs to discuss on the upcoming call:
> >
> > #1: Are the hl, relativeRef, and service implementations
> >     independent enough?
> >
> > Our numbers are looking good. We now have two method implementations for hl,
> > relativeRef, and service -- although, by the same implementer (Mattr). We'll
> > have to debate if that counts or not on the upcoming WG call. If there are two
> > independent implementers at Mattr that created their did:web and did:sov
> > implementations, then it should count. If not, there may still be a good
> > argument to count these as two independent METHOD implementations.
> >
> > #2: Are we letting the JSON serialization keep
> >     unimplemented features?
> >
> > The JSON serialization continues to lack implementations for datetime, double,
> > integer, and null. The WG has noted that they're going to keep those features
> > in the spec anyway in order to support the entire JSON data model.
> >
> > #3: What are we going to do with deactivated, nextUpdate,
> >     and nextVersionId?
> >
> > We are currently lacking implementation support for the following DID
> > Resolution features:
> >
> > * deactivated
> > * nextUpdate
> > * nextVersionId
> >
> > DID URL Dereferencing now has two independent implementations for most of the
> > features, but just barely.
> >
> > Those are the three remaining items to discuss wrt. implementations, AFAIK.
> >
> > -- manu
> >
> 
>

Received on Monday, 28 June 2021 10:00:14 UTC