Meeting minutes
Brent Zundel: bulk of meeting on VCDM issues.
Brent Zundel: Agenda updates?
Manu Sporny: plan for an transcription tool for meetings (perhaps)?
Brent Zundel: intros?
Transcription tool
Manu Sporny: transcription tool ---
Manu Sporny: current shortcomings of transcription system - hard to say off record comments; URLs not shared in minutes; proposals and resolutions don't get saved hence not shared
Manu Sporny: subtopic: any URL or is detected as a 'special thing', will be pulled into the minutes automatically.
Manu Sporny: saying things off the record - only easy way to do that is going into a breakout room, but high friction. Or say things without colon in the chat channel.
Manu Sporny: put "off the record" in the chat window" but AI likely get it wrong at times. Shouldn't depend on the AI accuracy
Manu Sporny: concerns?
Brent Zundel: discussed topic last week during advisory bd f2f meeting. Part of session on open source software dev in W3C. Came up with simple requirements toward a transcription system. Outlined in a way the community might be able to do this work
Ivan Herman: may want to check with Otto from DID working group. May be approaching this differently. Considering speech to text engine it inject them into IRC, making current tools usable.
Manu Sporny: talking with Otto. Goal is to move to a transcription system ASAP. In discovery phase so many options being explored. Previous attempts (3 prior iterations) all of which generated issues, complaints.
Manu Sporny: ideal outcome is moving to a new system sooner rather than later
Brent Zundel: if the earlier problems can be solved there is great interest in adopting it,. Anything based on Google Meet is not usable in China, hence is a non-starter.
Ivan Herman: China problem is on issue re: Google Meet. But there are other concerns around Meet, including the AI use of Meet by Google.
Possible F2F
Brent Zundel: agrees with many flowers blooming approach. There is encouragement by W3C
Brent Zundel: F2F in Brussels in June, June 2 to 4, 2026. But don't make travel arrangements yet . But have room arranged by GS1
Charter update
Brent Zundel: As before this meeting, Charter voting results 18 responses supporting charter as written . Voting open through 27th of Feb. Objections at last minute are common. So stay tuned
Ivan Herman: not public yet but Ivan and the Chinese team haven't given up on a cryptosuite SM2 and SM3 into the charter. Might get comments from members in China to get it into the charter and members found later to do it. Heads up this might happen when voting closes
Brent Zundel: urges encouraging Chinese members to make this change.
Brent Zundel: encourage voting members to do so.
VC Data Model Issues
<Brent Zundel> https://
Brent Zundel: VCDM issues looked at yesterday and everything has been triaged.
w3c/vc-data-model#1480
Brent Zundel: First is Example of use of renderMethod
Brent Zundel: assigned person isn't on the call. May not match current plans for the renderMethod
Manu Sporny: Issue may be overtaken by events.
Manu Sporny: core of issue is the EU wants to have something to say about the renderMethod. There displayParameter is being used. Significant number of needs being met in education sector are met by current HTML renderMethod approach.
Manu Sporny: Patrick creates PRs but this may have fallen off his list. Manu offers to do it so.
Ivan Herman: If we do that we should do with confidenceMethod.
<Dmitri Zagidulin Zagidulin> +1 to what Manu Sporny said, we should just point to the Render Method spec
Ivan Herman: but a reference to both
w3c/vc-data-model#1593
Brent Zundel: Add SD-JWT-VC to Ecosystem Compatibility section - minor change. Just adding another line to the list in the spec. Anyone willing to be assigned to it? Find a link to JD-JWT-VC and adding a line to it.
Manu Sporny: Concerned that the word VC is use in the SD-JWT-VC spec incompatible about the way W3C uses Verifiable Credential. This has no relation to the W3C VC work. No mapping to VCs in general so would have to do this individually per data model
Brent Zundel: understanding this request w
<Dave Longley> perhaps we should add a line that says: "Despite the use of the term 'verifiable credentials' in the SD-JWT-VC specifications, these specifications explicitly state they are incompatible with W3C Verifiable Credentials"
Understanding of the request would go into the ecosystem 5.11 ecosystem compatibility section.
Manu Sporny: supports adding that text and clarifies it is not directly related to the W3C VCDM workgroup activity
Brent Zundel: may be worthwhile to point out that we don't know all the things the transformation section items will result in successful transformations.
Brent Zundel: anyone willing to be assigned to add this to the incompatibility section.
Manu Sporny: willing to give this a shot
w3c/vc-data-model#1603
Manu Sporny willing to address this
w3c/vc-data-model#1617
Brent Zundel: Manu Sporny raised this in Oct. haven't yet discussed it.
Manu Sporny: way back in another universe thought the best thing to do would be issuing one credential claim. A driver's license would be split into 35 different credentials, for example. Now we have selective disclosure so atomic credentials may not be the best approach.
Manu Sporny: if you want to selectively disclose use it not individual credentials
<Manu Sporny> I'd be fine w/ removing the section, too.
Ivan Herman: is that a normative section?
Brent Zundel: no it's not
Dmitri Zagidulin: hasn't come in a while - is a VC implementation Guide a responsibility of this group.
Dmitri Zagidulin: minimize the amount of claims, kitchen sink credentials done with selective disclosure - belongs in implementation Guide
Dmitri Zagidulin: Have had the conversation about composite verifiable credentials. I.e., credentials which contain credentials.
Dmitri Zagidulin: CLRv2 from 1Edtech - expresses school transcript as a VC. It is compound as a credential for each class is then wrapped by an outer credential
Dmitri Zagidulin: technically this is correct vis-a-vis VCDM but awkward.
Dmitri Zagidulin: should instead of representing compound credentials in compound form use id digest hash to link to them
Dmitri Zagidulin: do we want to repurpose dependent claims to address this concept?>
<Dmitri Zagidulin Zagidulin> agreed, we don't want to make removi8ng the section dependent
<Dmitri Zagidulin Zagidulin> let's remove
<Manu Sporny> +1 to that ^
Brent Zundel: if anyone is volunteering to become an editor of the implementation guide (not revised since 2019)? Bundling dependent claims fits well into the revised implementation guide. But doesn't want to close this issue dependent on revising the implementation guide.
<Dmitri Zagidulin Zagidulin> so, +1 from me to removing this current section & closing this issue
<Dave Longley> +1 to remove the section and let the future take whatever direction wrt. linked VCs
<Dmitri Zagidulin Zagidulin> I can do that.
Brent Zundel: will open an issue in the implementation guide repo addressing this issue, link to it and remove the section PR.
Brent Zundel: Dmitri Zagidulin has volunteered
Brent Zundel: would raise the issue in the implementation guide or Dmitri Zagidulin would you like to?
<Brent Zundel> w3c/
Dmitri Zagidulin: willing to take this on. Close this one after PR re: implementation Guide
w3c/vc-data-model#1592
Brent Zundel: Multiple comments on intro section. Manu Sporny assigned to himself a year ago.
Manu Sporny: happy for someone else to take this one. Maybe history of why it is what is would be something Manu has knowledge about.
Brent Zundel: neglected to ask Manu Sporny in last conversation - apologies
Brent Zundel: not opposed to having a conversation about it. More productive to discuss with concrete changes in a PR
Manu Sporny: to make it easier for someone else to raise a PR....
Manu Sporny: Paragraph 1 doesn't define what VC and link to said definition should be made.
Manu Sporny: physical equivalence understood by average user by not for other part isn't.
<Dave Longley> note: various physical "watermarks"/other security mechanisms, etc. ... aren't necessarily very intuitive either
Manu Sporny: unclear if mentioning business rules would help. The process a verifier runs is helpful to understand this.
Manu Sporny: examples of holders is provided. Concern is might cause confusion but was added because leaving it out did cause confusion
Manu Sporny: Should we link to something about subject being the holder and point to that implementation guide section
Manu Sporny: Examples of verifying VC are confusing. Have a section on what is processed in the document and should repeat that.
<Dave Longley> "Does the role 'verifier' not include the process of 'verifying' the correctness ... of a claim" <-- definitely not, this is specifically avoided through the use of trust in the issuer instead
Manu Sporny: bullet 4 - confusing and definitions missing. Have to take a look and see what can be done.
Manu Sporny: figure 1 how do you identify identifiers? See proof of possession orypto key, might be useful
<Dave Longley> ecosystem is not used as a proper noun
Many: why are we using ecosystem? Is it not also a protected environment or proprietary system? Ecosystems come in many forms. Can do it multiple ways. Shouldn't remove the language as request isn't clear.
Manu Sporny: federated ID model and what we're presenting is not clear. We aren't clear about federated ID and what we are doing Federated ID is 2-party model, W3C is a 3-party model. May need to explain the differences. Ask for clarification
Brent Zundel: currently Manu Sporny is assigned to this issue and wishes not to be.
Brent Zundel: who else can take this on and make one or more PRs to address the concerns
Brent Zundel: long pause.....
lost my audio can some take this while i fix it?
<Manu Sporny> Dmitri Zagidulin Zagidulin: Maybe we can come back to this
Dmitri Zagidulin: should we address another time?
Brent Zundel: yes come back to it later
w3c/vc-data-model#1583
Brent Zundel: security review recommended the security considerations should be modified. Didn't tackle it then and said it would be future work
Manu Sporny: threat modeling guide has been written and its comprehensive an well written. Working groups should include threat modeling in each of the groups.
Brent Zundel: expects a request for threat model for the W3C ecosystem.
Manu Sporny: every spec should have a threat modeling guide. But that's wishful thinking as they are significant piece of work
Manu Sporny: Threat modeling might become a web of guides pointing to each other anchored to the VCDM threat model - may become the future state. Likely need to do one for VCDM v1.1 - SHould close this and agree to do one for VCDM
Ivan Herman: is uneasy because many of these are for specs in document mode. This could be an issue for existing specs which are more complicated. Should do this going forward, not retrospectively.
Ivan Herman: these re in document maintenance mode.
Manu Sporny: lets talk to Joe and Simone for suggestions
Brent Zundel: At time.