[MINUTES] CCG VCALM 2026-02-17

W3C CCG VCALM Meeting Summary - 2026/02/17

*Attendees:* Benjamin Young, Dave Longley, Dmitri Zagidulin, Elaine Wooton,
Eric Schuh, Joe Andrieu, John's Notetaker, Parth Bhatt, Patrick St-Louis,
Ted Thibodeau Jr

*Topics Covered:*

   - *Community Updates:*
      - The DID "Web" method has reached the Recommendation state, marking
      a significant milestone as the first DID method to do so.
   - *PR Reviews and Merges:*
      - *PR for adding links to component tables:* Merged after a minor
      conflict resolution. The update adds a third column linking directly to
      endpoint sections in the specification.
      - *PR 599 (Addressing Issue 399 - Client initiated VPR):* Merged
      after discussion. This PR adds examples and sequence diagrams to
      demonstrate how an exchange client can send a Verifiable Presentation
      Request (VPR) when initiating an exchange. Key discussion points included
      the structure of credential queries, the use of JSON comments, and
      clarifying "empty request body" to "empty JSON object" in diagrams.
      - *PR for Callbacks Endpoint:* Merged. This PR adds the callbacks
      endpoint to the component tables, which was previously missing.
      - *PR for JSON Schema for Query By Example:* Discussion held, with
      plans to review next week. Key points involved ensuring the
schema supports
      other query types and doesn't prematurely restrict valid spec extensions.
      - *PR for Delete Presentation by ID Endpoint:* Merged. This adds a
      DELETE endpoint for presentations, mirroring the existing DELETE
credential
      endpoint.
   - *Issue Discussion:*
      - *Issue 585 (Minimal viable credential templates and step templates
      for workflows):* This issue is being kept open to track annotations
      for examples in an appendix. The group discussed whether to keep
or delete
      the examples if no annotations are provided. A comment will be left for
      "John's Notetaker" to get feedback before closing.
      - *Issue 603 (Index Allocator):* Assigned to Patrick St-Louis and
      Dave Longley for further description and potential PR.
      - *Issue 606 (Redirect URL):* Discussion confirmed the need for both
      a spec description and OAS definition for redirect URLs.
      - *Issue 595 (Add delete presentation by ID endpoint):* Closed,
      addressed by a PR.
      - *Issue 610 (Unifying example requests):* This issue aims to
      consolidate requests for various examples related to workflows. A comment
      will be left for "Coyote" (not present) and reviewed next week.
   - *Workflow Discussion:*
      - Dave Longley proposed the need for normative guidance or an
      algorithm for processing workflow steps, similar to a state machine, to
      handle complex navigation, branching, and repeated steps. This was
      discussed in relation to the current spec's approach to workflows.

*Key Points:*

   - The "Web" DID method reaching Recommendation status is a significant
   achievement.
   - The meeting focused heavily on reviewing and merging Pull Requests to
   refine the VCALM specification.
   - Discussions around credential queries, JSON schema validity, and
   clarifying wording (e.g., "empty JSON object" vs. "empty response") are
   crucial for spec accuracy.
   - The group is working towards defining clearer mechanisms for workflow
   step processing and example clarity.
   - Four PRs were merged during the meeting, indicating good progress on
   the VCALM work item.

Text: https://meet.w3c-ccg.org/archives/w3c-ccg-ccg-vcalm-2026-02-17.md

Video: https://meet.w3c-ccg.org/archives/w3c-ccg-ccg-vcalm-2026-02-17.mp4
*CCG VCALM - 2026/02/17 14:58 EST - Transcript* *Attendees*

Benjamin Young, Dave Longley, Dmitri Zagidulin, Elaine Wooton, Eric Schuh,
Joe Andrieu, John's Notetaker, Parth Bhatt, Patrick St-Louis, Ted Thibodeau
Jr
*Transcript*

Patrick St-Louis: Welcome we'll get started in a couple minutes usual. Some
time for people to come in.

Patrick St-Louis: Okay, we're going to get started with the call. If people
can keep joining then, they will catch up with where we are. Welcome
everyone to the W3CCCG VCOM meeting. Today is the 17th of February 2026.
This is a W3C meeting. So all W3C policies are into effect for this call.
so we will be discussing today the VCOM work item which is a API for
managing verifiable credential throughout their life cycle from issuing
verifying and updating status on the agenda today.

Patrick St-Louis: So we'll first have some time for introduction and
community updates and then we'll move into review and issue assignment. If
someone has topics they want to bring up we can also make place to discuss
these specific topics. so I will leave a moment if someone wants to
reintroduce themselves to other people in this call. please take the
microphone now. Now is the time.

Patrick St-Louis: And if anyone has a community updates they would like to
share anything that is relevant to not only VCOM but other adjacent
specification when it comes to verifiable credentials and exchange please
let us
00:05:00

Patrick St-Louis: No, no, I can give a little update. so we had a
presentation at the CCG meeting today. Stephen Karan was presenting a
status update for the did webv method. this topic was actually registered
back in September, but there's quite a backlog of presentation at the CCG.
So today was a time it was a great call gave a lot of status update or
current deployments and a bit of the direction of the did webvh method and
I learned on the call this morning that it's the first did method to reach
the recommendation state. So that is pretty significant in my opinion.

Patrick St-Louis: so that's for community updates. So let's get into the
agenda for today. So we will go to the PR reviews. there were a couple of
PRs last time that I think we'll be able to close. So we will start with
these. And I also saw there were a couple of new PRs that were opened. So
let's go in reverse time order. So the older PRs first I have still not had
time to go back to the pseudo code R for selective disclosure and request
mapping from one format to the other. I will get to that when I have some
time. so the first thing we will want to discuss I think this was just to
add links to the components table.

Patrick St-Louis: I think this was dependent on a respspec PR.

Eric Schuh: Yeah,…

Patrick St-Louis: I think one so this has been merged. So this is great. so
why don't you take us through this, Eric?

Eric Schuh: so nothing's really changed from last week. so this adds a
third column to the component tables that links directly to the endpoint
section of the specification. there is one final step that I've pinged Manu
about which is updating the CDN server that is actually hosting this
main.js which I don't have access to. Dave, I wasn't sure if maybe you
could just ping Manu, but I sent one ping a couple hours ago, so he might
have just not gotten to it yet.

Eric Schuh: But I think we could probably Yeah,…

Dave Longley: Yeah, he's traveling.

Dave Longley: I'll just say he's traveling today.

Eric Schuh: no worries then.

Eric Schuh: So I think that we can probably merge this and once that server
is updated we might need to do a PR to update the link if Monty rules the
version number as well…

Eric Schuh: but effectively I don't think there's any harm in merging this
at the moment. the tables will just render the old tables until that
resource server gets updated.

Patrick St-Louis: That sounds great to me.

Patrick St-Louis: So I see here there seems to be a conflict. So if you
could be simple.

Eric Schuh: fix those.

Eric Schuh: Yep. I'll do that. Okay.

Patrick St-Louis: It should be just a one button. and then we'll come back
to this at the end and if there's an objection we'll merge this. I don't
see a problem. so there is a PR that was open last week from Par. Would you
like to take us through this PR?

Parth Bhatt: So,…

Patrick St-Louis: All

Parth Bhatt: 599 address issues 399, which requested, examples showing that
an exchange client can send a verifiable presentation request when
initiating an exchange and not just an empty object. So currently the spec
states that either party can request information from the other party I am
referring to client and the server but we had no concrete examples
demonstrating this for the client side. So I have added two examples to
address that in the exchange example section which is 3.6.8.

Parth Bhatt: The first shows a client specifying which cryptographic proof
format and envelope types it will accept. if you scroll down below the
sequence diagram. Yes, this one. and then the second example shows a client
requesting credentials from the server specifically a business registration
credential before it's willing to proceed further. so this were the new
examples and apart from that I have also added the two sequence diagram
illustrating the same client initiated VPR flows.
00:10:00

Parth Bhatt: Which are basically if you scroll up a bit. Yep. This two
sequence diagrams. That's it.

Patrick St-Louis: Thank you.

Patrick St-Louis: So I just have a question. So this is a query by example.

Patrick St-Louis: But there's no example in it. I thought the example was
required field in the query by example. Yes, Dave.

Parth Bhatt: Go ahead.

Dave Longley: I don't know that it is.

Dave Longley: We'd have to see specifically what the spec says, but I don't
know that we necessarily require an example. I take the meaning there to be
a little bit confusing.

Patrick St-Louis: What? Yeah. this basically says send me anything,…

Patrick St-Louis: right? Okay.

Dave Longley: It says,…

Dave Longley: "Send me anything with these accepted crypto suites in these
envelopes."

Patrick St-Louis: Is able to the credential query? could you add just a
little example in here an example key with a type?

Patrick St-Louis: Even if it's just a type or just a context, just so we
have a little example thing, I think it would make a bit more sense. Is
that possible?

Dave Longley: I want to make two comments.

Dave Longley: Sorry no one's on the queue so I'll just jump on. I think it
might be a good idea to just put an empty example to indicate that you will
accept anything so that you always have to put example there. which might
come into conflict with whether or not there are any required properties in
an example itself. but I wanted to highlight that I think this spec example
is showing that a wallet or client for the exchange can announce the crypto
suites that it would accept and the envelopes that it would accept without
having to specifically say what credential they want to accept.

Dave Longley: they can say that as well, but it is important that we
support the use case that they don't care about that specifically.

Dave Longley: They're just engaging in an exchange. They're happy to
receive any type of credential provided that it's secured using these
crypto suites or those envelopes, right?

Patrick St-Louis: Okay, so actually everything is optional in the
credential query.

Patrick St-Louis: So you could just have an empty credential query and that
would be valid technically. Yeah. I think this all seems fine.

Patrick St-Louis: Not seeing anything here. Okay.

Dave Longley: So I did have one comment on the PR …

Dave Longley: which is in the mermaid diagram we say we talk about empty
request body and empty response and empty you we say empty a few times it's
not we need to be care I think we should be a little more careful with that…

Dave Longley: because it's not an empty request body. It's an empty JSON
object. So it's just close curly brace in all of those cases.

Patrick St-Louis: Right. Right.

Patrick St-Louis: You want Yeah.

Dave Longley: So we should clarify that.

Patrick St-Louis: Yeah. I think that is a valid point. so this is mostly a
wording context. I think we both understood. I think the meaning here was
to return an empty object, but it's mostly so an empty JSON object in the
request body would make sense. Okay, I think this is a valid comment. so
I'll let you fix this part and if this can be done during the call, we'll
loop back in and that's all good.

Parth Bhatt: Yeah, sounds good. I have one more question.

Parth Bhatt: So, if you scroll down on the issue maybe click on the issue.
question in very first comment.

Patrick St-Louis: Which grate are we looking at?

Parth Bhatt: So in this example the credential query are commented and I
added those back in the example. So do we want to just keep it at a high
level and…
00:15:00

Patrick St-Louis: Even in these yeah I would remove…

Parth Bhatt: remove the credential query command the nesting of that? Yes.
Yes.

Patrick St-Louis: then if you have an example JSON like that I don't think
that's valid JSON I think you can these are just comment but I don't think
JSON supports comment

Patrick St-Louis: I might be wrong.

Dave Longley: if you're asking about removing the comments. so two points I
think I put this in here and I think I was wrong about the format of a
query by example. I think the accepted crypto suites and accept accepted
envelopes go are nested underneath credential query which is what I think
you put in the PR and I think that's right. we do happen to support putting
comments not that these comments should be used. So just we can get rid we
don't have to consider it at all for what's on the page here. But I do
think there's something in respect to let you put comments into the JSON
such that even when you copy and paste the example it will drop those
comments. I think there's something that does that which is helpful in the
spec sometimes, but I don't think we have to do that here because I think I
just messed up the input.

Patrick St-Louis: interesting. I believe that if the accepted crypto sweet
and envelope are at this level they are talking more about the proof on the
presentation or…

Parth Bhatt: That is correct. Sorry.

Patrick St-Louis: do they just not belong here?

Dave Longley: I think you have to put those in did all query if you were to
do that.

Patrick St-Louis: Okay. Yeah.

Patrick St-Louis: Yeah, I think that might be correct, right? Yeah.

Patrick St-Louis: Yeah. Yeah. Yeah. Okay. Pretty good. yeah, Joe. They've

Joe Andrieu: Yeah, it's a slightly different note.

Joe Andrieu: Is it should we Benjamin brought up this JSON C approach that
should we switch to that? I think that's just a hack on the front end to
you. We are putting comments in the GitHub page which is invalid JSON but
they can use the copy and paste functionality to get just the JSON out of
it or we could just stick with pure JSON and say don't put comments in
there but the comments are handy.

Dave Longley: So, if you're talking about when we put stuff into GitHub,
when we put it into issues and stuff, yeah, that sounds great. Use JSON C
if that's going to accomplish that. There's some other mechanism that I was
talking about in W3C specs. I don't remember the mechanics of it. I don't
think it uses JSON C. I think it's something else, but I could be wrong.

Patrick St-Louis: Okay. Benjamin

Dave Longley: So, it's not when we talk about what goes into the spec. I
definitely think the comments are helpful, but I don't think it
specifically uses Jason C to accomplish

Benjamin Young: Yeah, I can do some poking around, but I think what
respspec shells out to which I think is highlight J whenever it's
highlighting code can take JSON C as I don't know if it's a media type or
what it uses. but I think it's JS that actually does the proper
highlighting. I could be wrong, but it's worth testing. But generally
speaking, yeah, you can use JSON C where you would say JS or JSON in many
modern tools and it'll highlight it correctly. but obviously…

Benjamin Young: if you copy and paste out of that and stick it in a JSON
file, it'll be wrong.

Patrick St-Louis: Okay. …

Patrick St-Louis: so let me go back a bit to something I've heard. are we
talking about here where the holder is proposing a presentation to a
verifier?

Dave Longley: the short answer to that is I think this example is just
showing that when the holder connects they may optionally request something
from the server indicating that they have some constraints that the server
must meet.

Patrick St-Louis: Okay. Okay.

Dave Longley: So I think the two use cases that are covered here the client
is saying I only accept these crypto suites or envelopes. And then the
second use case is I need you to send me a business credential.
00:20:00

Dave Longley: I think that's what was used in here before I'm willing to
continue the exchange.

Patrick St-Louis: But I thought this was…

Patrick St-Louis: what workflows were done, And these would be different
steps of a workflow. Yeah. It's got a bit card.

Dave Longley: they are a workflow when you implement a workflow you can
choose to accept a VPR from a client and…

Dave Longley: so they're able to make certain requests and you can set up
your workflow such that if a client makes a request for a type of
credential that you're willing to present if there's a common credential
that I'm on this trust list or I have this business I am of this business
or so Yep.

Patrick St-Louis: Yeah. Yeah.

Dave Longley: Then you can elect in the workflow to send that if the client
requests

Patrick St-Louis: Because these are the type of interactions that workflows
are really useful for.

Patrick St-Louis: Sure you can use a workflow for just send me I request a
presentation or I want to issue you something but the real thing it can
enable is these multi-step I want you to send me this first and then I'll
either ask you something else or issue you this and yeah okay so this is…

Patrick St-Louis: what this is trying to enable Dave

Dave Longley: Yeah. I might frame that a little differently.

Dave Longley: I would say workflows themselves enable coordinators to
offload I mean they're vital no matter how many steps you have. They enable
coordinators to offload the communication with the wallet based on
workflows that they support. And however, those workflows can also have
multiple steps and do a number of interesting things all using some basic
primitives. So it allows you to do a lot of powerful things with just a few
primitives

Dave Longley: where you don't have to worry about what protocol the client
is choosing or deal with any of those things. The coordinator just once it
creates an exchange for a workflow that it supports, it then just pulls
that exchange to get the state information from whatever choices the user
has made with their wallet.

Patrick St-Louis: Very good.

Patrick St-Louis: So, we'll wait for this small edit and then we'll come
back at Any other comments on this call backs. We like those. All right.
Let's get into this.

Patrick St-Louis: Eric, why don't you take us to this part?

Eric Schuh: Yeah, sure.

Eric Schuh: Pretty simple. as I was going through updating the links for
the component tables, I noticed that the callbacks endpoint was not listed
in them. I had misunderstood stood originally I think where this lived Dave
so I understand your confusion on the initial issue but this callbacks
endpoint was added to and as well as the expected caller should match up
with what Dave posted in the issue. and I think Dave it looks like you
already had a chance to take a look at this.

Eric Schuh: That is the new X dash component that I added to support adding
the links to the callback tables or…

Dave Longley: Yeah, that's right. It looked like the PR matched what I had
written on my thoughts there. And anyone else is welcome to take a look or
say whatever they need

Patrick St-Louis: What is this table link?

Eric Schuh: sorry to the component tables. that it's literally the link
shortcut so that the respect OAS can generate the component tables. Yes.

Patrick St-Louis: I'm confused. Are we talk is this something that will
need to be made in the request that's sending the call back and the
headers? This seems like a header to me.

Dave Longley: Yeah, this is a spec feature.

Patrick St-Louis: Okay.

Patrick St-Louis: So, this is not a protocol thing.

Dave Longley: It has nothing to do with the …

Patrick St-Louis: Sorry, I thought we're talking about header parameters.

Dave Longley: there's nothing to do with the protocol.

Patrick St-Louis: So, this is just like some Okay.

Dave Longley: Yep. Yeah.

Patrick St-Louis: So, then yeah, I don't see I think I mean I'll trust that
this does what it's meant to be doing. yeah,…

Dave Longley: Yeah, if you look OAS at the bottom, Patrick, and just in
case you want to know that that's where those things are used.

Patrick St-Louis: Yeah, that makes a lot more sense. Thank you. any
objection to merge this?
00:25:00

Dave Longley: The wait I was going to say you got to hit the squash button.
It didn't have a clean commit history because it has for a future note,
that merge branch in we now have two branches in the history.

Dave Longley: And so whenever that shows up, that probably should have been
a rebase to fix the conflict, but it was a merge. And if you had squashed
it, we would have avoided that.

Patrick St-Louis: Okay, noted.

Patrick St-Louis: It'll probably be fine.

Patrick St-Louis: We could revert it, which we could squash after.

Dave Longley: I think reverting it.

Dave Longley: We'll just add more commits. So, the history will have to be
rewritten…

Dave Longley: if we want to fix that up. That can happen at some point.

Patrick St-Louis: Yeah. …

Patrick St-Louis: so we're happy we're leaving it like this.

Dave Longley: I think surgery would have to be performed to fix it right
now. So, we can just leave a note that we might want to fix that up.

Patrick St-Louis: Might want next commentary. There we go.

Patrick St-Louis: com this and someone will pick this up. All Let's go see
this one. Manu is not on a call today. We can still have a look. so what we
want to address here. So, query by example. This is a really old issue.
required so it seems to be okay. I'm not too sure what's happening just
semantic information about query by example.

Patrick St-Louis: keeping in mind this is from 2023. So things could have
been a bit different at the time. It was discussed some months ago under
specified JSON scheme. So we want a JSON schema for query by example. Okay.

Patrick St-Louis: So, it sounds like there were some comments from Dave
that will need to be addressed probably related to the latest things we've
been talking about accepted issuers and crypto suites which is an array of
objects. So JS schema to take account fields structure right object

Patrick St-Louis: array of objects or strings. So, it sounds like just the
accepted fields need to be just changed a little bit. I think overall it's
probably going to look pretty good. Presentation request query.

Patrick St-Louis: Can you define? Yeah. a problem I see here is that this
is a credential query for a query by example. but we might not have all
these fields on other query type. So,
00:30:00

Dave Longley: I think if you scroll up there's actually an query and if you
see type on line 36 right now we only support const query by example in
this VR. I think we have some others in the spec. So, we might want to make
an one of for this I think it's this object starting on line 32 that we
need to Yeah,…

Patrick St-Louis: But this going to be problematic because surely if we
support many different types, They're all not going to the credential query
would also need to be a one of,…

Dave Longley: we Yeah,…

Patrick St-Louis: right? Yeah.

Dave Longley:

Dave Longley: but possibly we might have to figure. Yes. we should look at
what else we have in the spec today and we can namespace this or make it or
maybe it becomes a credential query object. We got to sort that out. I'm
sure there will be more edits to this thing both before the PR and…

Patrick St-Louis: Okay. yeah.

Dave Longley: probably after or

Patrick St-Louis: And it's true that the issue was specifically about query
by example.

Patrick St-Louis: I wondering if instead we just want to define this query
by example, and…

Dave Longley: I would say we can make more loose at least.

Patrick St-Louis: leave this outside the J.

Dave Longley: I think it's probably good to define at least have
placeholders for those other things, but we can make the schemas less
restrictive. So you could still have query and it could just be an object
and maybe comment out we'll still have query by example in there.

Patrick St-Louis: Yeah, but then it doesn't really address the issue
anymore, which is okay.

Dave Longley: I'm just saying the higher level stuff where we don't quite
know exactly what we want to do. We could make sure that it's flexible
enough so that people can put other things in there which it needs to be
anyway because the whole design is that you can extend it and put whatever
you'd want in there.

Dave Longley: So we just need to adjust the schema I think to make sure
that it supports

Patrick St-Louis: and future to support other query types.

Patrick St-Louis: because if we provide a JSON schema most likely people
will want to use this JSON schema to validate stuff and we want to make
sure that we don't shoot ourselves in the foot that the schema actually
prevents accepting something that's totally valid according to the spec
which is not right. So

Dave Longley: Yeah, that's exactly right. We want this schema to say if you
happen to put query by example in there, we're going to check all of that
because we know how to check that.

Patrick St-Louis: Interesting. Yeah,…

Dave Longley: If you put some other type, it's still going to pass, but
there won't be any checking rules on it unless it's another type that we
have in the spec. That's what we want.

Patrick St-Louis: I think that's a good idea. Yeah, and as we have well
definfined types, we can add them and leave people free to do whatever they
want. let's put other query types. So, I think this is great. So, we'll
probably review this next week. Overall, very good and very relevant for
what we're discussing these days. So, the last one here and then we'll do
another quick run around see if we can merge some of these other ones. so
add a delete presentation by ID endpoint section.

Patrick St-Louis: So before we get into that okay so we do get presentation
endpoint and it would make sense to add a delete. let's open up the issue
so very recent issue. Yeah this seems to make sense. presentation and
credentials should both have a way to delete any more details part to this.
So it seems pretty straightforward…

Patrick St-Louis: but if you could still just take us through.

Parth Bhatt: No, I think you covered everything.

Parth Bhatt: That was a missing end point in the section spec. So, I just
simply added that's it.

Patrick St-Louis: Yeah, very good.

Patrick St-Louis: And these are the same things that are on this matches
the delete credential endpoint pretty much.
00:35:00

Parth Bhatt: That is correct.

Patrick St-Louis: As is this Yeah, it's a pretty clean commit history. any
objection to merging Any objection to rebasing and merging this? Maybe I'll
start to say this so we can be more precise. There This has been deleted.
Awesome. Okay, let's go have a look.

Patrick St-Louis: Did you have the time to rebase your branch, Eric, or has
it been a bit more complicated than anticipated?

Eric Schuh: No, it shows that it's ready for squash and…

Eric Schuh: merge on my screen.

Patrick St-Louis: We just merged something else, so you'll probably need to
just hit that button one more time.

Eric Schuh: Okay. Yeah,…

Patrick St-Louis: Hopefully, it's not going to be too much. Let me know
even…

Eric Schuh: I guess it shows on my screen that all checks have passed and I
only have a squash and merge option.

Patrick St-Louis: if you refresh the page.

Eric Schuh: Yeah. Yeah,…

Patrick St-Louis: Any objection to have Eric squash and merge this on
behalf of the group? I'm not sure I'm not able to. It clearly says here
that there's some conflicts. maybe it's Yeah.

Eric Schuh: I'd be happy to share if someone has maybe they spot something.

Patrick St-Louis: Any objection to just let Eric hit the button?

Dave Longley: No objection. Just make sure to squash it. I imagine if I'm
sure there's some eventual consistency bug here and either it will fail
Eric and then we'll know that Patrick's is more consistent or it will pass
and we'll know that yours is more up to date.

Eric Schuh: Okay.

Patrick St-Louis: Joe. …

Joe Andrieu: Yeah, I just wanted to add I'm seeing the same thing Eric is.
So,

Patrick St-Louis: So, I'm the problem. That's good. All do it, There we go.
Maybe my browser just didn't like the PR. perfect. So, this has been
merged. Let's have a look at this last one. I think this one, if we're
lucky, we might be able to merge this as well. Yes, we are. Let's review
the comet history. this is pretty good.

Patrick St-Louis: self assigned little force push here. Think this is fine.
yes part.

Parth Bhatt: Yeah, I rebased because at that point last week I think we
merged couple of PR.

Patrick St-Louis: What do we do here Dave?

Patrick St-Louis: Do we squash or rebates?

Dave Longley: This one is still hold on.

Dave Longley: This one is still showing if I go to the files changed and
maybe mine is out of date. It's still saying empty response or if complete.
It doesn't say empty JSON object. I just see one place where that got
updated. So, there's four other empty places that might be confusing to a
reader.

Patrick St-Louis: I think part probably just updated the line you commented
on …

Dave Longley: Yeah.

Patrick St-Louis: if there's other ones where are we seeing them? So
everywhere it says empty we want to say empty there for example.

Dave Longley: Empty JSON object. Yep.

Patrick St-Louis: Can you go through these part?

Parth Bhatt: s***.

Patrick St-Louis: And then if we can close this today it' be good. so while
you do that we're just going to go through issues. we'll leave a little
five minute at the end to come back and merge this.

Patrick St-Louis: and that will be very good. the index allocator. I don't
really know what it's used for use case multiple issuers or clients can I
assign this one to Dave and myself and maybe Dave I could ask you to leave
just a description here of what this is used for and I'll take the task of
putting that in a PR
00:40:00

Patrick St-Louis: or I mean if you want to open a PR, you're more than
welcome Xator.

Dave Longley: I'm unlikely to get to a PR, but I might be able to get to
putting a description in an issue. So, thank you for your suggestion. I
will try to do

Patrick St-Louis: So we move this for PR state. it's actually in ready for
PR, but we can move it

Patrick St-Louis: into move it into the SP. So, I'm going to assign myself
and I'm going to put you on there as well. We can kind of work together on
this. There you go. at the redirect URL. So, this was open last week. you
want to take us through this, Dave? Yes.

Dave Longley: Yeah, I think we were just missing this in the spec. I see
pars your hand is up. I don't know if that's an old hand or not. Okay, I
think you just put it down. All right. yeah, we'd had this in the concept
of a redirect URL and at least some text around it in the spec for quite a
while. but we don't have it in the OAS and so it's just missing there. so
one of the things you can put in your step description is a redirect URL
that will be sent at the end of the step that can redirect the client to
any URL, but that URL can be regular web page. it can also be an
interaction URL to allow you to continue doing exchanges somewhere else
without having to go back out to the web.

Dave Longley: and we just need some text that says that's where you put it
so that when your workflow runs, you'll send that to the client. Yes.

Patrick St-Louis: So we need two things this and…

Patrick St-Louis: the OAS and we need descriptive text and the spec. so PR
should include descript let's configure redirect URL and should also OS

Patrick St-Louis: There we go. Let's do one more. I believe we can close
rge. I'm actually going to do the right thing and just go really quick. is
it Yeah.

Dave Longley: It was actually linked in that issue. Should have the number
in there. Add delete number 595.

Patrick St-Louis: Okay. …

Dave Longley: If You can just say clo,…

Patrick St-Louis: it's because I referred to it, I thought it would close
it automatically.

Patrick St-Louis: I'll just say close.

Dave Longley: closed addressed by PR. Yep. There you go.

Patrick St-Louis: That's it. good. Let's keep going. okay. wait. Let's put
this one. so, this is already assigned. yeah, you can talk about it. Is
there more that needs to be discussed here?

Patrick St-Louis: And are you in a good place to says help wanted? I don't
know if it's you that is asking for help or okay.

Eric Schuh: Yeah, that is me.

Eric Schuh: Mostly so there's a couple PRs linked. I'll try not to repeat
myself from the text there. but basically Dave when you added the examples
for the minimal viable credential templates and step templates for the
workflows. you also included a couple of fairly long examples of workflows
and had mentioned that it would be nice to get them annotated with some of
the interesting choices made for those particular examples and right just
to explain the optionality. Pyote, you also had an outstanding request for
a note or comment or an example of how to use a particular VPR object the
domain object I believe it was.
00:45:00

Eric Schuh: so my proposal here was basically to have this issue open to
track annotations to the appendix that was added with that previous PR and
basically give people two to four weeks to annotate the examples that are
already in there or add their own examples to that appendix. and then maybe
in a month we come back to this issue and if that the appendix has grown or
is in a form that we think is acceptable we keep it around and if no one
has annotated any examples I'll probably come in and plan on raising a PR
to delete the examples that are currently included in that appendix. so
yeah I think that's it.

Eric Schuh: And Dave, I don't know if I know you're busy.

Eric Schuh: I don't know if it would be helpful if maybe we could find 30
minutes and talk through those examples and then I could do the leg work of
actually putting a PR in. but I'm not exactly sure what you were
considering interesting in those examples. So, I think be hard for me.
Never.

Patrick St-Louis: No, Dave.

Patrick St-Louis: Dave is not busy at all. He

Dave Longley: I would also suggest I don't think we need to necessar these
are already in an appendix.

Dave Longley: So, even if we don't annotate them anymore right now, I think
they're useful examples for people to see more complex workflows. and I
wouldn't say that they need to be deleted. I would say it's probably okay
for us to just keep them there in the appendix and if anyone gets to doing
any kind of annotation in the near term that's fine and if not it's also
fine. So, we can keep the issue open just as a reminder for anybody…

Dave Longley: who might have that kind of time, but I think it's okay as
This

Patrick St-Louis: Okay. …

Patrick St-Louis: so there was a lot of information TLDDR, what do you need
from us, here? What are we doing?

Eric Schuh: So I think the main things here is in point one there of this
issue I list two other issues that I think we can close out as c this one
captures the points that were remaining from those previous issues.

Eric Schuh: So part of this issue was trying to let us close out those
other two that had been left hanging after prior PRs. yes.

Patrick St-Louis: …

Patrick St-Louis: you're suggesting that we can close these right now
without additional Let's just go over them really quick. see where it's at.
And you're saying that these have somewhat already been addressed by some
PR, but they were like some outstanding kind of details that were not
directly addressed. Does that capture discussion?

Eric Schuh: Correct. Yes.

Patrick St-Louis: It's a lot about examples.

Eric Schuh: Yeah. …

Patrick St-Louis: Okay.

Eric Schuh: these are the examples that went into the appendix originally.

Patrick St-Louis: And basically the question is do these example help to
clarify the workflow steps? does anyone here feel like maybe we can ask
John I know that he's kind of moved on to other things. I will leave a
comment, ask him if what he thinks about it, and if next week he hasn't
replied, we will go ahead and close this. Dave

Dave Longley: Yeah, I think that's fine. I think something that the spec
should eventually have is ideally either a non-normative algorithm for
workflow step processing.

Dave Longley: And I think that would help a lot as well. I'm hoping to
produce something kind of like that that I could put into an issue that
could get turned into a PR in the future.

Patrick St-Louis: Yeah. …

Patrick St-Louis: not presented I brought this didcom workflow thing and
they have a pretty strong sort of state machine. and so one thing that they
do a bit different because I've been looking a bit closer is you will not
necessarily have a start and an end to a workflow. this workflow can move
back and forth between different states and depending what state it's in
different information will be displayed with the user.
00:50:00

Patrick St-Louis: the user will be able to interact with different
direction they want to go in the workflow. So it's not so much like a
sequential thing like A B C D it could go to C and then to B and then to E
if the workflow allows it and the way that it's structured. So they have
these transitions that are enabled and then they have a set of actions that
can be taken and a set of states. Depending what state you're in, you can
take certain actions and depending what action you do, it's going to
trigger a specific not transaction, what was the word I say transition to
another state. And then there's display data to be added.

Patrick St-Louis: I feel like the workflow here I would need to review it
but they're mostly a sort of sequential steps of operation in the workflow
as defined in the VCOM.

Dave Longley: When you say the workflow here, do you mean the example
that's given? VCOM, the steps are also semantic, so you can move around
however the workflow designer would like.

Patrick St-Louis: Wasn't there a next step thing.

Dave Longley: There is any given step can go to a next step but what next
step that can be computed by variables and data so it could change
depending on what inputs are given and…

Patrick St-Louis: Okay.

Dave Longley: you might come back to another step at a certain point with
different data. So there's a number of ways that it can Yeah.

Patrick St-Louis: and what you're saying is so because what I heard you say
before is the spec is missing normative information on how to navigate this
state machine. that I'm going to use as term that represent how you
progress to a workflow. It just says a bit here's the workflow, here's how
it work, but there's no normative way that a software would navigate this
workflow.

Dave Longley: And so what I'd be suggesting would be would we can say that
it's normative we can give some normative guidance but you can do something
equivalent that has the same behavior from the client's perspective but if
you would like an algorithm for how to process the current step and move to
the next one is really all I'm talking about. So this would be evaluate the
current step if it's convert it into a more static step by running your
template code check to see if any input is required from the client if you
need a VPR to be satisfied check to see if there's any verifiable
presentation in the step that you're supposed to put any issued credentials
on top of check for issue requests issue

Dave Longley: those credentials either storing them in the state or putting
them into the VP. and then check to see if there's a redirect add all of
that stuff to your response based on what you found. set your next step and
then save your state and send it to your client. It's more or less that so
there's just a few steps there a few instructions there to follow when
you're processing a step and then you just repeat that for every step in
the workflow. So it's not an overall here's how you start and… here's how
you end. It's for any given step that you read, which is an exchange is
always going to start on some step.

Dave Longley:

Patrick St-Louis: It sounds like you touched maybe on something that's not
quite related to this issue directly.

Patrick St-Louis: I think what you're proposing is a bit more specific than
this. take the describe So, support branching, repeated step, that's kind
of what I was referring to, when you want to be able to go in different
direction. did the PR add some descriptive text or only added examples,
Eric?
00:55:00

Eric Schuh: So the PR that I submitted was from…

Patrick St-Louis: Okay.

Eric Schuh: if you scroll up I think just a tad right above that example I
believe Manu had raised an initial PR that added the descriptive text in
this initial issue. and then he added that note there that asks what does a
minimal viable step template look like and what does a minimal viable
credential template look like which is what the PRI raised for the appendix
addressed from Dave's examples that he provided in this issue. and so that
those two minimum viable templates were in my mind what were keeping this
issue open.

Eric Schuh: And so those were addressed, but in that discussion, it was
mentioned that it might be nice to also have the other examples Dave
provided here,…

Eric Schuh: provided as part of the appendix and possibly have them
annotated in the future, which is what drove that new issue there. Yes.

Patrick St-Louis: and this captured here.

Patrick St-Louis: So, I'll ask John if he's happy with some of the
resolution. I think it does address his initial issue. there were
subsequent things discussed. we'd like to close the session and follow up
with 585. So I'll commend this and next week if we haven't heard back we'll
just go ahead because I know John has moved on to other things and he may
just not be following this anymore. yes remebrance John.

Patrick St-Louis: So those who don't know John, he was here for quite a
while and he helped build an implementation and stuff, but he's …

Patrick St-Louis: he's just found another job that's not as close with
verifiable credentials. I met him at IW. He is very And then what was the
This one. Okay.

Eric Schuh: So this one Coyote had a request for a particular example as
part of a workflow.

Eric Schuh: So basically I was just trying to unify these multiple issues
that we had asking for different examples in workflows into a single issue.
also because a couple of these issues were very old. So, just trying to do
some cleanup. yeah,…

Patrick St-Louis: is not on the call today. the old issue we also want to
be careful that we're not just moving issues related to examples and…

Eric Schuh: the only reason I did this was just because we had multiple
ones that were basically in my mind asking for very similar things.

Patrick St-Louis: for workflows we are going to group these issues. So they

Patrick St-Louis: be So, same thing I'll put this comment usually it might
be around next week and we can follow up on this. Perfect. and now two
minutes left and have you had time part to make the change?

Parth Bhatt: Yes, I did.

Patrick St-Louis: Let's have a final review. I would love if we could There
we go. Let's look for The JSON objectivation object. very good. any
objection to closing Merging this PR. I'm really bad at saying if we need
to rebase or merge.

Patrick St-Louis: Dave, I will need your help on this.

Dave Longley: This one's kind of in between. One of the ways I analyze
this. So, the first thing is make sure there are no merge commits. There
aren't any of those. That's one of the big things.

Patrick St-Louis: Yeah, they all look like fairly,…

Dave Longley:

Dave Longley: Yeah. Yeah.

Patrick St-Louis: one line.

Dave Longley: when I read one of the commits it's the only one that's
giving me pause is the last commit that just says add curly brace like
someone reading that commit yeah they can go look at the context for it…

Dave Longley: but they would have no idea really what that's for and so
that's really the only commit that doesn't really explain what's going on I
think it's okay …

Patrick St-Louis: Okay. Yeah. Okay.

Dave Longley: but in the future I would suggest that it says something more
like a little more descriptive.
01:00:00

Patrick St-Louis: So let's all collectively to just try to be somewhat
precise about what has been done in the commit…

Dave Longley: Yeah, if you look at the other commit,…

Patrick St-Louis: if we need to go too crazy…

Dave Longley: they all look really good.

Patrick St-Louis: but yeah and it's fine if it's just a general thing this
makes sense. it's just when we are working sometime we don't realize that
the commit message can be seen without the rest of the context…

Patrick St-Louis: because in the context it makes sense but yeah perfect we
merged four PRs today…

Dave Longley: Yeah, that's exactly right.

Dave Longley: Always remember you're gonna lose the context when people are
looking at

Patrick St-Louis: which is pretty happy with that closed a few issues
assigned a few so with that said the time has come so I will adjoin this
meeting. Thank you very much for attending. Good progress today and I will
see you all again next week. Bye-bye.
Meeting ended after 01:01:14 👋

*This editable transcript was computer generated and might contain errors.
People can also change the text after it was created.*