Meeting minutes
Community Updates
Phillip Long: I'm formerly with W3C on the team for about 8 years, now with GS1 for about 8 years, lots of W3C experience, looking forward to being chair.
Manu Sporny: Great news, CA DMV has made an announcement that they are putting W3C VCs every single physical driver's license and identification cards.
Manu Sporny: Every single one of these new cards going forward will have a W3C compliant VC embedded into the barcode on the back of the driver's license. There are verifiers that are out there that can check these already.
Manu Sporny: It uses the ecdsa-di-2023 cryptosuite and the VC barcode spec that has been incubated in the CCG.
Manu Sporny: CA DMV also issues digital W3C VCs and ISO mdoc mDLs.
Manu Sporny: Another update is from MOSIP. MOSIP creates an opensource digital identity government platform. They released some new numbers, they deployed a long time ago with VC 1.1. Over the last year they now issued over 100 million W3C VCs.
Manu Sporny: In the Philippines, India, Uganda, some other countries, will have more data at TPAC. By the end of this year there will be around 200 million W3C VCs just through their platform.
<Joe Andrieu> Is there a URL for the Mosip #s?
<Phillip Long> MOSIF is the acronym?
Manu Sporny: So great news for W3C VCs over the past couple of months.
render method and confidence method
Brent Zundel: We've received word that the CCG has published these specs as final community group drafts.
Brent Zundel: That means they have been through the process of IPR releases and that important step. That means that we as a WG can consider working on them.
Brent Zundel: Our charter allows us to adopt and bring those in without any need to recharter.
Brent Zundel: I've said in the past that I think how we handle those specs should be an indication of whether adding even more items through a new charter is feasible.
Brent Zundel: Given that, let's have a conversation about who will become editors for these specs. We're not changing our meeting cadence at this time and the plan is to have task forces for working on these specs and moving them forward and then bringing things back to this group.
Brent Zundel: We need editors other than our good friend, Manu Sporny, who has been carrying the burden.
<Phil Archer> Is there a more up to date link than https://
Brent Zundel: That is the conversation I'd like to have, who will do the editing work and who will do the talking work -- showing up to the task force meetings?
Brent Zundel: Please jump on the queue.
Manu Sporny: I agree, your not my Mom, you're much taller than that.
Brent Zundel: Other than that, we're the same.
Manu Sporny: I agree.
Manu Sporny: People filled out a poll and volunteered to be editors.
Manu Sporny: I passed that list to you, Brent Zundel, and Ivan.
Manu Sporny: We have people that have volunteered to do that work, some are not on this call or might not be part of the WG, so wanted to provide that as data. For render method and confidence method and almost every other spec, except the post quantum suites, we have 2-3 editors that aren't me that have volunteered.
Brent Zundel: Thank you, Manu.
Phil Archer: Wearing a co-chair hat, editors are a crucial thing and we need to make clear who those people will be. If we will take it on, I will want to know whether test suites exist, whether there is the capacity in the group to create them.
Phil Archer: I imagine that's a "yes" and I know there are good implementations of render method at least.
Phil Archer: I'd like to know how far away we are from candidate rec at this stage and I'd like to know what we need to do to get there.
Phil Archer: I know there are 9 open issues on render method, 4 on the confidence method -- a more recent issue on render method, no new issues from confidence method since May.
Phil Archer: I am less confidence about the confidence method, render method looks great, I'd like to know if I'm wrong about any of this.
Manu Sporny: We've had regular incubation meetings in the CCG. For render method, it has 4 different implementations and 6-7 if you count some in asia -- and confidence method has 1-2 and we need test suites to know. So for render methods we need test suites to figure out what's being tested, visual output or data model.
Manu Sporny: I know our company has committed to do an implementation.
Manu Sporny: As far as how far along render method is -- it's been incubated for 3 years now, there's the US/EU/decentralized approach and the asia approach, both participants on both sides looking to move things forward as well.
Manu Sporny: A number of education institutions want the decentralized PDF/SVG approach as well.
Manu Sporny: The spec is really over-incubated at this point, we've got maybe 3-4 months of spec to get to CR, that's my expectation. The core architecture and design is done but we have to work through details.
Manu Sporny: Confidence method -- the reason for not a lot of activity is that there isn't a lot to the spec. There are actually zero interoperable implementations -- the unknown things are not the data structure or the data model, getting a test suite is easy -- it's the protocol stuff that is hard to get to interop and test suites.
Manu Sporny: There's some stuff in the VC API work to define what that looks like -- do we put it into a protocol and test it or do we test the data model?
Manu Sporny: The other thing with confidence method, people have been pushing biometric confidence methods -- like a picture of a person -- there are privacy questions.
Manu Sporny: Faces, fingerprints, other biometric things -- the WG will need to make some decisions on what we think is appropriate to standardize.
Manu Sporny: Those are the remaining big rocks that need to be worked through in a WG. I have some thoughts on how we meet to move these things forward.
Phil Archer: Yes, that more or less confirms my view, thank you.
Ivan Herman: Two things, one is a very formal one. Which goes back to editorship.
Ivan Herman: After resolving that -- then we need to officially publish a FPWD for both docs or for render method to start.
Ivan Herman: In both cases it's required to have an editor who has signed on and given their name to the document.
Ivan Herman: If we need invited expert status for an editor we have to get that behind us before we can get to formal publication.
Ivan Herman: Less administration -- Manu you said that we may want to check visually -- and that raises one flag for me. Which we may not be able to avoid, as soon as we are having anything that is related to user-facing things then the accessibility and i18n review will be much more complicated.
Ivan Herman: So far we have been lucky with those being easier -- in this case -- who knows how it will work. If we stop at SVG and say nothing about what will be in the SVG we might get by, but let's be ready.
Phil Archer: I was also thinking about a11y and i18n, I was thinking of render method -- and we need to render to all users. I would expect that at the same time, assuming we've got one or both of them to FPWD either today or the next time we meet in Kobe, we would immediately seek horizontal review to get those communities onboard asap.
<Manu Sporny> agree w/ phila -- immediate HR for renderMethod
Joe Andrieu: I agree with Manu about confidence method. The spec is straightforward and the work is in the different types. Manu Sporny, which types do you think are easier and low-hanging fruit. But proof of control over private keys is an easy one. Could you paint a picture... could we pick one and get it in there without solving all of them?
Manu Sporny: It's exactly what you said, Joe, proof of possession of a key is super easy and straightfoward.
Manu Sporny: There are orgs that will roll out biometrics whether we get there or not.
Brent Zundel: I am happy to propose publishing FPWD for both of these today. But we need to name editors.
Ivan Herman: I think, formally, and administratively, it's perfectly fine to pass a resolution which says to have FPWD provided that we get one or more editors for each spec.
Ivan Herman: The only thing I'm worried about is ... because we have one meeting a month and we have everything lined up, then we have to wait 3 weeks for a formal resolution and that's a bit silly.
Brent Zundel: I'm happy with that as a path forward, next meeting is TPAC.
Manu Sporny: +1 to that.
Manu Sporny: +1 to saying we're going to publish FPWD based on editors signing up. There are 7 people who said they want to be editors of render method and we have 2 for confidence method. That's the data we have.
Brent Zundel: Good data. How many are members of the group?
Manu Sporny: Two are here today -- in addition to those two another that's a member of the group and me.
Joe Andrieu: I may have said I'd be an editor of the confidence method.
Manu Sporny: You did not.
Joe Andrieu: Ok, but I will now.
Manu Sporny: Dmitri has said he wants to be editor as well for render method.
Brent Zundel: Ok, we have one editor for at least each document.
Brent Zundel: So that's enough to make a proposal.
<Phillip Long> +1
<Dmitri Zagidulin> +1
<Phil Archer> +1
<Brent Zundel> PROPOSED: We will publish a First Public Working Draft of Render Method, with a shortname of vc-render-method, with initial content from https://
<Ivan Herman> +1
<Phil Archer> +1
<Manu Sporny> +1
<Phillip Long> +1
Dave Longley: +1
<Denken Chen> +1
<Benjamin Young> +1
<Joe Andrieu> +1
<Parth Bhatt> +1
<Dmitri Zagidulin> +1
<Brent Zundel> +1
<Jennie Meier> +1
<David Chadwick> +1
<Ted Thibodeau Jr.> +1
RESOLUTION: We will publish a First Public Working Draft of Render Method, with a shortname of vc-render-method, with initial content from https://
<Brent Zundel> PROPOSED: We will publish a First Public Working Draft of Confidence Method, with a shortname of vc-confidence-method, with initial content from https://
<Denken Chen> +1
<Phil Archer> +1
<Ivan Herman> +1
Dave Longley: +1
<Dmitri Zagidulin> +1
<Brent Zundel> +1
<Phillip Long> +1
<Manu Sporny> +1
<Joe Andrieu> +1
<Parth Bhatt> +1
<Benjamin Young> +1
<Jennie Meier> +1
<David Chadwick> +1
<Ted Thibodeau Jr.> +1
<Ted Thibodeau Jr.> +1
RESOLUTION: We will publish a First Public Working Draft of Confidence Method, with a shortname of vc-confidence-method, with initial content from https://
Manu Sporny: I'll propose something, but I don't think we should discuss today -- food for thought. I was wondering if we could get a new charter proposal in front of the group to talk about it at TPAC, that's our next meeting. I wanted to hear what the plan might be.
Manu Sporny: We have been meeting -- we have 3 weekly standing calls for different calls in CCG. We could shift those times over to work on these specs, is that the expectation or will we just get into a weekly cadence with this call -- Brent you had mentioned parallelizing the work but I'm worried about doing too much of that.
Manu Sporny: The question is -- do we want to have standing calls for these and alternate between render method and confidence method, that's one proposal. We have a new meeting infrastructure in the CCG, it auto-scribes, it records video, it does everything automatically and then emails the list with minutes, topics, summary, etc.
Manu Sporny: The downside here is that it IS recorded and it doesn't follow PA/Ivan's tooling. I'm suggesting that if we're going to massively parallelize that we use more automated tooling and that could get controversial.
Ivan Herman: First, before I forget about the tooling -- personally I don't care if we use my tooling or not, but a longer discussion would be need and could be discussed in Kobe or something. Not an easy thing to do, there has been experimentation with automated tools and the results were iffy at best.
Ivan Herman: Totally different thing: the group had a special slot for special calls -- the slots are there we could use those.
Ivan Herman: We can call the calls whatever we want.
Ivan Herman: The practicalities that I don't know -- I presume if I follow that your intention is to have each of these documents in its own github repository -- then I will need all the usual things to transfer to W3C to set them up as WG repos.
Ivan Herman: We can discuss that offline next.
Manu Sporny: Yes.
Phil Archer: Can I get a sense, please, what is the overlap in personnel -- is it the same people? If so you can't parallelize, if it's the same people we would alternate. If they are distinct groups you could parallelize.
Brent Zundel: Yup good point. We could meet every week and talk about those two specs and the maintenance stuff as well. Alternatively, if there are more dedicated groups and smaller groups of people whether they alternate or not -- I think that's the second option.
Brent Zundel: My inclination is to allow the render method and confidence method folks to each self-organize. You are welcome to use this meeting slot when this meeting isn't happening obviously -- and if people going to both meetings find we talk about the same things then that conversation can grow organically about of that. I don't want to say top down what must be done.
Brent Zundel: I want to allow you guys to work it out -- Dmitri for example has said he's only working on render method.
Brent Zundel: Feel empowered to start meeting if you care about these things and make sure that Phil and I are invited to those meetings.
Brent Zundel: Editors should run those calls. Any decisions in those decisions are proposed decisions until the WG gets to look at them.
Manu Sporny: Just some feedback: Ivan I'll work with you to move the repos over to the WG.
<Dmitri Zagidulin> +1 I'm fine with alternating / sharing a call
Manu Sporny: Phillip Long, the overlap is like 60% -- that is working on all these specs. I would suggest, even with what Dmitri said, that we just have one call to work on these two specs. Or we'll alternative if folks really don't like it. I'd rather start with just one call and just see how we do until we split further.
Manu Sporny: We already hold the incubation calls during this exact time and we could just say that we're going to move those calls over to render and confidence method and they will be WG calls.
Manu Sporny: We'll need help setting up a time for that -- a W3C calendar thing for that, but that's the proposal for that. The only thing for that is do we use this tooling or that -- I was anti the automation stuff for years and it reached a point where it now is great.
Manu Sporny: The machine does not get tired and it largely captures every word but messes up the acronyms -- you can't tell it not to scribe stuff, it doesn't auto-link like your tooling, Ivan Herman, but we can modify its output to generate an HTML page with links and stuff, but no auto-resolution stuff.
<Denken Chen> +1 for starting with one unified call for both methods
Manu Sporny: We don't have to start with it, but it has made meetings so much easier, and all the minutes and scribing is taken care of.
Ivan Herman: I don't want to get into the tooling issue right now -- that's a separate issue, but at to the rechartering. Do we go there now or do we discuss on the two documents still?
Brent Zundel: No plans to talk about chartering today.
Ivan Herman: Manu raised it.
Dmitri Zagidulin: Just to weigh in, I think I'd love to start with the CCG auto-scribe infrastructure for the render method calls until someone complains and would be happy to establish one call and alternate calls between render and confidence.
<Joe Andrieu> +1 to Dmitri's suggestion
Brent Zundel: Ok, either Ivan or I can set up the calendar stuff in the W3C WG page, so as you determine meeting times and links, we can assist, reach out to us.
<Manu Sporny> +1 to what Dmitri said.
Brent Zundel: I'd like to spend the last bit of time we have looking at issues.
Brent Zundel: I do want to say that, in the VCDM, we have 8 open issues and they are already triaged, nothing new there.
Brent Zundel: Which is fantastic.
VCDM Issues
<Ted Thibodeau Jr.> it will be worthwhile investigating a way to `pause/restart recorder`
<Brent Zundel> +1 Ted
Manu Sporny: Yes, they have been triaged. I did look through all of them, I don't know what to write for some of them, the WG might need to weigh in. It's just editorial text, but do we want to add that text or what are we trying to accomplish.
Manu Sporny: I'm not saying we need to go through them in this call, nothing time sensitive -- fine to just say ... we'll need to put some effort into processing them, maybe we do it at TPAC, maybe another call.
Brent Zundel: Thank you, Manu.
Brent Zundel: I am noting that Manu -- I'm seeing you added the class 2 determination on a number of these.
Manu Sporny: Yeah, maybe.
Brent Zundel: I'm looking at the VC-JOSE-COSE repo.
Manu Sporny: Oh, yes.
Brent Zundel: That's fine, glad you did that, I think that's a difference between editor/team of editors did an initial pass vs. group did it.
Brent Zundel: I am glad that you did that (categorized some issues on VC-JOSE-COSE).
Brent Zundel: I think it's important as a group that we say we agree, it speeds up processing, so thank you for that!
Manu Sporny: +1
CID Issues
Brent Zundel: Unfortunately, I don't remember where we were at with the VCDM, I don't think we got to controlled identifiers at all, so let's spend a couple minutes there.
<Brent Zundel> https://
w3c/cid#157
Brent Zundel: This is labeled "Possible erratum / enhancement". This was back in July.
Brent Zundel: I think it fits in with our ... I think this is clearly something that we can address.
Brent Zundel: But what's the extent to which this is a class 4 change -- we don't have a class on it.
Manu Sporny: I think you can argue it either way, the IANA consideration section is non-normative, could be class 2, but a new named information hash entry request ... sounds class 4 to me.
Manu Sporny: Someone would have to write the spec on the IETF side or we'd have to write something in our spec. I didn't necessarily see it as an errata, I think.
Brent Zundel: Any other input from folks with IANA experience?
Brent Zundel: I agree, technically with Manu -- it's a non-normative section, but it does feel somewhat substantive.
Brent Zundel: So any objections to labeling this class 2?
Brent Zundel: I'm hearing no objections, so this is a class 2 change.
Manu Sporny: Eh... ah ... ew ... I don't think it's appropriate to add something of this magnitude to our spec and someone needs to write the IETF doc for us to link to.
Ivan Herman: We are just classifying right now.
Brent Zundel: Ok, I'm writing class 4. Which means we'll do it in the future.
Brent Zundel: I don't think it's an errata, it's not an error in the spec.
Manu Sporny: Correct.
w3c/cid#161
Brent Zundel: Fix assertion narrative ... etc. Labeled class 2 ready for PR. I agree.
Brent Zundel: This was moved over from the DID spec originally, I believe. This is from 2022.
Brent Zundel: Any objections to keeping this class 2?
Brent Zundel: Ok, no objections.
Brent Zundel: One more.
w3c/cid#160
Brent Zundel: Service ID required or optional, number 160.
Brent Zundel: This is marked class 3 with a discuss label. I should remember deep in my bones what class 3 means ... but it's ... what?
Manu Sporny: You are clarifying or modifying conformance of existing implementations. Either making something previously non-conformant => conformant or vice versa. Here this is clearly contradictory and we need to clarify. I recommend we make this property optional.
Manu Sporny: We just need to strike the "id" bit that says each service needs to contain these properties.
Brent Zundel: Ok, I will label this as errata in addition to the class 3 label and we can and will address it.
Brent Zundel: Any objections?
Brent Zundel: Not hearing any objections. We're at time.
Brent Zundel: Thank you everyone for being here, thanks Dave for scribing and I look forward to seeing task force meeting invites, reach out if you need any help getting things on the official calendar, you are all fantastic, it's a pleasure working with you.
Dave Longley: You're welcome!