Request for Explicit Interop Test Suites from Mozilla

[mprorock]

per email on the AC list

---

Preview | Diff

[iherman]

cc @tantek

[clehner]

Public copy of the request: https://lists.w3.org/Archives/Public/public-new-work/2022Jun/0000.html

[brentzundel]

I've made some suggested changes for better clarity.

[brentzundel]

@mprorock if you accept my suggestions I will approve this PR, thank you for raising it.

[mprorock]

@mprorock if you accept my suggestions I will approve this PR, thank you for raising it.

Accepted - thanks for the extra eyes!

[Sakurann]

please fix the typo in implementations before merging. Thank you for the PR

[brentzundel]

@msporny as you probably have the most direct experience with the current VC test suite, I am very interested in your opinion of these proposed changes.

[mprorock]

please fix the typo in implementations before merging. Thank you for the PR

corrected

[tantek]

LGTM and this would resolve the specific issue that Mozilla's Formal Objection was based on. Thanks for putting in this work!

[msporny]

@msporny as you probably have the most direct experience with the current VC test suite, I am very interested in your opinion of these proposed changes.

The more modern test suites that we're planning on using for the VC 2.0 work will be able to test interoperability in the way this PR is suggesting.

The only question I have is whether or not the test suites can utilize protocols to demonstrate interoperability? Given that test suites are non-normative, I suggest the answer to this question is "Yes, the test suite can utilize VC protocols to demonstrate interoperability." For example, the credential refresh and credential status require protocols to be used. I believe this would address Google and Mozilla's concerns, while not crossing the hard line that Microsoft has about the WG not working on protocols.

IOW, one way through this is to say the WG will not normatively define protocols, but can utilize protocols to demonstrate interoperability on VC 2.0 data model features. This all keeps going back to whether or not the VC2WG is generating data model specifications or data model + protocol specifications. I believe we have consensus to work on data models normatively and protocols non-normatively... and we can prove interoperability by using a combination of normative specification text and non-normative test suites (that utilize particular protocols).

[iherman]

@msporny I do not think there is a problem. The text in the charter does not say how the interchange happens, that can be done via protocol agreed upon by parties, emails, or by snail mail...

[iherman]

I will start the formal process by W3M for the approval of the charter.

[iherman]

LGTM and this would resolve the specific issue that Mozilla's Formal Objection was based on. Thanks for putting in this work!

Thanks Tantek.