[MINUTES] CCG VCALM 2025-12-09

CCG VCALM Meeting Summary - 2025-12-09

*Topics Covered:*

   -

   *Pull Request Processing:*
   - Merged "updated get exchange response object" (PR #569) which
      clarifies the description of the get exchange response object to
      indicate it can be an empty object.
   -

   *Abstract API Model and Translation Layers:*
   - Discussion on the need for a higher abstraction of the current API to
      facilitate easier migration between different credential formats and
      protocols (e.g., Indy to WebV, AnonCreds, SDJ, MDL).
      - The goal is to reduce software changes required for ministries and
      issuers when new credential types or standards emerge, shifting towards
      configuration-driven updates.
      - Exploration of translation layers for formats like MDL to W3C
      Verifiable Credentials, and how abstracted APIs could handle different
      presentation request formats (e.g., DALE, VPR, DWLL, AnonCreds).
      - Concerns were raised about creating too many standards (XKCD
      effect) and the inherent difficulty in perfect translation
between evolving
      formats.
      - The potential for a "workflow" feature within specs like VCOM to
      handle these translations was discussed.
   -

   *Notification and Discovery of New Credential Schemas:*
   - A challenge was identified in how to notify ecosystem verifiers about
      new credential schemas being published, especially when issuers
migrate to
      new technologies (e.g., WebV alongside Indy).
      - Ideas included a public verifier onboarding process with
      subscription models or queryable resources, and potentially an "agentic
      chatbot" trained on credential specifications.
      - The possibility of using DID documents and associated services for
      credential discovery was also mentioned.
   -

   *Interoperability and Test Suites:*
   - Emphasis was placed on the importance of focusing on implementation
      and interoperability, with test suites playing a key role in
achieving this.
      - The goal is to move the VCOM specification towards Candidate
      Recommendation and beyond efficiently.
   -

   *VCOM Specification Development:*
   - A proposal to focus on reducing existing issues and PRs rather than
      adding speculative new features to the VCOM specification for Version 1.0
      was made.
      - The timeline for moving the specification into a working group and
      the subsequent focus on test suites and CR were discussed.
   -

   *Specific Issue Discussion:*
   - Issue #560 regarding table updates in the respect OAS repo was
      mentioned, with a PR expected by the end of the week.
      - An issue related to converting query languages (e.g., Query by
      Example to JSON pointers, DALE to JSON pointers) was discussed, with
      existing open-source libraries potentially superseding the
current example
      code. This was seen as a potential appendix item for the spec.

*Key Points:*

   - The *"updated get exchange response object" PR was merged*, indicating
   progress on the VCOM specification.
   - There's a strong interest in *abstracting APIs* to simplify credential
   management and migration, reducing vendor lock-in and the burden of
   software updates.
   - *Translation layers and support for multiple formats* are seen as
   crucial for interoperability, but the complexity and potential for creating
   new standards need careful consideration.
   - *Notifying and enabling verifiers* about new credential schemas is a
   significant ecosystem challenge that may require new mechanisms beyond the
   core VCOM API.
   - The *focus for VCOM moving forward should be on completing the
   specification, building test suites, and driving implementation* rather
   than introducing new, speculative features.
   - The meeting concluded with a discussion on *scheduling for the holiday
   period*.

Text: https://meet.w3c-ccg.org/archives/w3c-ccg-ccg-vcalm-2025-12-09.md

Video: https://meet.w3c-ccg.org/archives/w3c-ccg-ccg-vcalm-2025-12-09.mp4
*CCG VCALM - 2025/12/09 14:50 EST - Transcript* *Attendees*

Dave Longley, Dmitri Zagidulin, Eric Schuh, Joe Andrieu, John's Notetaker,
Kayode Ezike, Manu Sporny, Manu Sporny's Presentation, Nate Otto, Parth
Bhatt, Patrick St-Louis, Rodrigo
*Transcript*

Manu Sporny: Hey I thought Patrick was going to run the call today, but I
don't see him. so I can get us started and then I'll hand it over to him
when he joins. So, let's go ahead and get started. welcome to the VCOM
call, the BC API for life cycle management call. today is December 9th,
2025. we have an agenda which is basically our standard agenda which is to
process pull requests and then potentially process issues if we need to
although we have quite a number of issues that are ready for PR and are
just waiting on PR.

Manu Sporny: So it might be a short call today as we only have one poll
request to process. so that's our agenda. Are there any updates or changes
to the agenda? Anything else we want to discuss today? All right. If not,
let's go ahead and get started with poll requests. let me go ahead and
share my screen. let's see. All right. So, one pull request, opened by you,
ic, three weeks ago,

Manu Sporny: an updated get exchange response object. yes, please, take us
through it, Over to you.

Eric Schuh: Yeah, sure.

Eric Schuh: So yeah, 3 weeks ago I had put this in making some changes to
this get exchange response object. we had discussed and decided that my
changes were a little bit overboard. so I've updated this PR to simply
update the description of that object rather than the object itself. this
should then appear in section 3.66 of the respect document. and ended up
being, I think, a fairly simple language addition just to clarify that this
could be an empty object. and I think that's about it.

Manu Sporny: Looks good to me. it's been floating out there for a while to
reviews. any continued concerns, Dave? I know that you requested some
changes.

Dave Longley: No. I was just navigating there myself to approve the poll
request.

Manu Sporny: And because it's been floating out there for a while, I think
we can merge it on the call today. so this is pull request 569 569 updated
get exchange response object.
00:05:00

Manu Sporny: All And then mind if I squash this into a single commit?

Eric Schuh: Yep. Go ahead.

Eric Schuh: Shouldn't be a problem.

Manu Sporny: All squash and Didn't do that. this is Let's see.

Eric Schuh: Yeah, good.

Manu Sporny: get exchange response object to make it clear that this can be
an empty object. There we go. Squash merge and then 569 addresses this
issue loing issue Has been merged. Close comment. All right, that's that
item. okay, awesome.

Manu Sporny: Thank you very much, Eric, for processing that addresses the
only outstanding PR that we have. hey Patrick, I went ahead and started the
meeting, but happy to hand it over to you…

Manu Sporny: if you want to run the rest of the meeting. I noted that
almost everything we have is ready for PR, so really we're just waiting for
PRs, and I didn't know if you had, any thoughts on what we should talk
about today.

Patrick St-Louis: Yeah,…

Patrick St-Louis: I'm happy to take over. Sorry, I just really got deep in
the gutter in some other task and I didn't see the time go by. let me just
open the repo. I didn't have any specific items in mind. there is something
I wanted to share. So this morning there was the hakapog meeting and one of
the discussion was to have a higher abstraction of the current aipi API one
of the things that's happening now at least in BCG gov but I'm sure it's
thing that happened all over the place is there's this code base right so
occupy and then within bcgv obviously

Patrick St-Louis: different ministries use that interface to issue their
credentials and some of them the software provided to them or others might
need to have a more custom API integration and when it comes to a situation
right now so basic is looking to migrate away from indie towards webv but
this can also happen to any sort of migration whether could be moving away
from on creds or anything. U and one of the things that makes it difficult
right now is software changes and how to coordinate not only issuer
software but also verifier software to change their API calls mostly
they're going to make or their presentation request.

Patrick St-Louis: will be this idea of making a bit more abstracted API
configuration or API model that relies more on configuration and less on
the controller software knowing what it will issue and sending that
information when issuing something or making a presentation request came up
and obviously I've

Patrick St-Louis: there to strongly encourage to adopt something similar to
the VC API. so I just wanted to mention this likely will be some work that
will take on eventually to create a sort of abstract API plugin and I think
it would be interesting to model it maybe not like on implementation of the
VC API but at least try to meet in there somewhere something a bit closer.
Just wanted to share this. Yeah. Yes.
00:10:00

Manu Sporny: Yeah. Yeah. that's super interesting. I guess where do you see
the overlap? I guess meaning I think that's good and it's good to figure
out how to meet in the middle however we want to put it. but I'm wondering
kind of like what's the core thing that you're trying to address with the
abstract API?

Manu Sporny: Is it like just the calls that you make to issuance and
verification infrastructure or is it something else?

Patrick St-Louis: So the end goal is that in the future when …

Patrick St-Louis: because not knowing what technologies will come out right
things can change pretty fast that when BCG gov as an issuer of credentials
and having verifier service depending on it. when they want to make a
migration towards I don't know a new profile or a new stack of credential
types like SDJ or MDL or implement these different credential that it
doesn't require a custom update of every single software all the time and
try to tell people to update this

Patrick St-Louis: software to call this API and then when there's a
migration later on this migration can be more implemented through
configuration of the software rather than making that many code changes
does that make sense so for example when they send a payload of the
credential they want to issue using the VC data model right even if they're
issuing on creds

Patrick St-Louis: represent this with a credential subject and these sort
of terms instead of doing a anon creds payload for anon creds some other
payload for some other object and have the configured instance then this
payload from this issuer was meant to be issued as an enred and then
triggered a sort of nonredit based exchange which this is another thing in
itself. Yeah. Yeah.

Manu Sporny: Okay. that makes sense. yeah, I've been wondering for a while
about translation layers, meaning for example, in California at the
California DMV, their digital wallet supports both MDL and verifiable
credentials. and in fact, even their driver's license is issued both as an
MDL and…

Manu Sporny: as a W3C verifiable credential. and they are semantically
equivalent to each other. each one contains exactly the same data fields
than the other one.

Patrick St-Louis: Yeah.

Manu Sporny: It's just they're in different locations and things like that.

Manu Sporny: and as the MDL format is very different from everything else
right and so having those translation layers helps at least anyone
deploying this technology not to worry too much about the underlying data
format or the underlying protocol. ultimately are they just care about
attributes and issuing those attributes in whatever format works.

Patrick St-Louis: Yeah. I think translation layer is a another good example
is presentation requests, canon has a very specific model for that when you
want to request predicates or so on. maybe something like DAW, which seems
to be pretty popular right now, could be used, as the input if it's maybe a
little bit simpler and then could be translated to,…

Manu Sporny: Mhm.

Patrick St-Louis: different query language through a mapping or
translation. instead of having software to learn all these three different
query languages, which I think presentation request queries is probably one
of the more complex thing especially when you get into advanced features
predicate selective disclosure and so requesting multiple credential doing
a presentation request has maybe a higher level of complexity to learn
whether that's presentation request the VP request 2024 or dwell or anon
there's a lot of things that can be done a lot of logical gates and so on
so having this one abstracted endpoint that when
00:15:00

Patrick St-Louis: the payload comes in whether it's L a W3C credential a
anon creds or so on the payload looks more or less the same maybe some
additional options that are specific to for example with addon creds you
would need to provide some kind of connection identifier for which didcom
connection you want to send that payload to but that's just a connection ID
you can give as an option but at least follow this when want to represent
the attributes which is just a simple key value but you could just
represent that as a credential subject right use those terminologies and
then whether that needs to be translated to a MDL a SDJ or a add-on grids
can happen to like you mentioned a translation layer Dave

Dave Longley: Yeah, I'm on the queue to say I think what you're describing
is more or less what is envisioned for workflow features or it's already
the case that some implementations for workflows will take in for example a
VPR and can translate it to DALE and can translate it to presentation
request and presumably a nonredit could fit in to that model as well and
then whichever

Dave Longley: protocol or query language makes sense for the client that
happens to connect to the exchange can be served whatever they need to be
served and we've envisioned something very similar as well for translating
envelope based formats like MDL into a VC representation so VCs are
expected to cover the vast majority of credential cases and then there are
just a few that people are using for envelope cases is driver's licenses
and you could have a client provide a MDL to an exchange and the exchange
state could in return from the workflow service to the coordinator could
include a translation of that to VC format and then you could always just
consume in that format and not have to worry about whatever the thing was
provided and all of the verification of the envelope and all that stuff
would happen on those services.

Dave Longley: So any coordinator connecting up to a workflow could
implement to that single format. administrators could design their
workflows using whatever base presentation format they presentation request
format they want to use and it would be translated in the workflow.

Dave Longley: And I think it would cover most or maybe all of these use
cases and I think that it's designed to do that. So, if that wasn't clear,
I wanted to make sure it

Patrick St-Louis: Yeah, I think definitely workflows could be a good place.

Patrick St-Louis: My gut feeling right now I feel like workflow we're still
making a few changes here and there. So I don't know if we'll literally
implement the VCOM workflows but we could definitely have a look and get
inspired by this. But also when we talk about the vehicle one of the
advantage was this to prevent vendor locking and enabling software to be
abstracted for another one. But I think what I just explained here that if
a software needs to adopt a new credential type model something like the
VCOM if this already took that into consideration can make that process a
lot easier.

Patrick St-Louis: the other thing which is a bit more un I don't think it's
really related but for VCOM but not a issue we have but another challenge
everything is a challenge it's not a problem so issuers and verifiers that
are controlled by government entities we can coordinate these change. but
the other question is other verifiers in the ecosystem that so for example
BC has the person credential right that was one of their first credential
they issue and it's meant to be a citizen held credential on their wallet
and obviously it's issued by the government.

Patrick St-Louis: so if private verifiers in the ecosystem decide to start
to use the person credential to just verify different things for their
business need how can they be notify or updated when the issuers let's say
they just tangible example right now if DC is going to start to issue
credential using webv on top of the indie one so they will start issuing
with indie they're going to start issuing with WebVH, there's still going
to be like the IND credentials in circulation, but the verifiers will need
to know hey, by the way, you can also check for the webv identifier, So,
maybe that's a nonredit specific thing, but you can add a restriction in
your presentation request for a very specific credential schema that was
published.
00:20:00

Patrick St-Louis: So how do you notify or make a ecosystem aware of a new
schema that's published that they can also add in their verification
request. So that's like another challenge we're facing. But I don't think
the solution necessarily lies in the C API. This is more a sort of
notification or subscription model for verifiers. Nate

Nate Otto: Hey just briefly mentioning some relevant stuff from my recent
work. I've been doing some conversions between DEAL and other query formats
and in general some decel and presentation exchange verifiable presentation
requests with query by example have a lot of potential to get complex in
ways that do make it difficult to just translate between one to the other.

Patrick St-Louis: Okay.

Nate Otto: And there's some particular JSONLDLD type IRI expansion stuff in
DAL that is a little bit tricky to convert back from DALE.

Nate Otto: So in any case just to be aware that is not necessarily an easy
thing to do and it may be valuable to have some sort of restrictions on the
input just sort of start with simple queries if you want to do much
translating or…

Nate Otto: formats that can be easily converted into a variety of these
target formats. more simply in my experimentation I found going from DAL to
VPR was more difficult…

Patrick St-Louis: That's a good feedback.

Patrick St-Louis: Would you say it goes both ways? some VPR stuff is
difficult to represent in Dool, some Dwool stuff is difficult in VPR or
there's only one direction that's more difficult than the other?

Nate Otto: but I think it really has to do with just the level of
complexity of the query that is trying to be represented and it's really
easy to get

Nate Otto: into a complex situation with DAL because you have credential
sets and credentials and with all the field matching considerations.

Nate Otto: There's just a lot more machinery that you have very easy access
to than you do with a really simple query by example. yeah.

Patrick St-Louis: Okay, interesting.

Patrick St-Louis: Thank you for that feedback. It's definitely going to be
useful. Do you have source libraries for that kind of work that could have
a look or…

Nate Otto: No, not representing the current work I'm doing. although I will
note the digital bazaar does have a oid4 client that's open source that has
some VPR2 DAL conversion code in it.

Patrick St-Louis: okay translation? Okay, cool. I'll definitely maybe have
a look at that. Yeah, man. Yes.

Manu Sporny: One thing that I heard you say that that felt like a dangerous
thing was the old standards joke XKCD where you've got two standards and
then someone's like this doesn't address my issues and now you've got three
standards. I heard you say something I might have misheard you something
like if we had a more generalized format then we could convert to these
other ones which is exactly that XKCD joke right I don't think what we want
to do is create a higher level of abstraction from DAL and VPR certainly
not yet because they're both kind of evolving.

Manu Sporny: and then the other reason is these things are not entirely
interchangeable and as they mature I expect that to get harder and harder.
they may have some subset that overlaps but sometimes features exist for
very specific reasons. So, I'm wondering if translation is probably the
best we can do and there's only a subset that we can actually translate in
reality over the next 10 years. it may be that we can only translate maybe
60% of the things one format to the other. but that might be fine, I mean
that might address the vast majority of use cases.
00:25:00

Manu Sporny: I think the things that I'm most concerned about are that was
Statement two was about how do you let other people know that you've got a
new credential format or something that's fairly significantly different.
in a fully distributed system, you can't do that because if it's a truly
decentralized system, you have no idea who's using the system and you don't
know how to get in touch with them. and then the only option you have is
potentially a message that you transmit in point that makes it all the way
over to point Z somehow, right?

Manu Sporny: But I don't think that thing that you're saying is something
you're trying to solve is a solvable problem without adding a tremendous
amount of complexity and…

Patrick St-Louis: So for the first part I was definitely not suggesting to
invent a new data model to rule them all.

Manu Sporny: and things like that. So

Patrick St-Louis: I think my comment was more of choosing one that
currently exists that is the most universal right which I think the VCDM
2.0 if I understood correctly one of its goal was that anything could maybe
be a VCDM 2.0 Right. There was something and then for the presentation
request what I was saying is there was discussion in using dwell as the
format that can be mapped to other I'm also aware that in the future you
can't prepare for things that you don't know there might very well be a new
credential and you just cannot prepare for things that you don't know and
it's about finding the correct abst

Patrick St-Louis: ract abstraction level you don't want to get so abstract
that it's a bit of a balance definitely for the other thing with the
verifier I'm pretty so my kind of the way I'm approaching this is to have
verify your on boarding process like a public verifier on boarding process

Patrick St-Louis: is that when a verifier wants to start leveraging your
credentials in their software, there is a proper onboarding process for
them to follow. And in that process includes something that they can either
subscribe to or provide them a resource that they can query to get the
latest updates that are in there. And I had this kind of idea that maybe a
sort of a AI agentic chatbot that is trained on the type of credentials
that you will issue. So the chatbot knows the specification you're going to
use.

Patrick St-Louis: that knows the type the did method you're going to use
that verifiers could chat with and that it would let them know they say I'm
using the person credential is there a new method available or so on and
this chatbot could return to them the information that they would need to
update in their presentation request or it's a bit farfetched But I feel
like there could maybe be something to be done there. so yeah, at least the
thing I think could be a potential solution is the onboarding process for
verifiers, right?

Patrick St-Louis: there is a public on boarding process that they read
could be a guide they read and there is through that either a thing they
can subscribe to or a agent they can interface with that would let them
know of these changes. Dave

Dave Longley: Yeah. there's a lot to respond to there. First, I was
thinking you might be able to do something like a service and a did
document if you wanted to have a discovery service for any kind of
information around…

Patrick St-Louis: heat.

Dave Longley: what types of credentials are being issued by a particular
party. so that might be one way to discover such a service and then use
such a service. presumably if you're verifying the credentials from said
you would be able to get their DID document and that's where you could look
for that. the other thing I wanted to say, Nate mentioned there's an open
source library that Digital Bazaar has put out called OID4-Client. I put a
link to it in the chat. this library, the readme needs some help. So, you
probably won't find direct function calls in there today in the readme, but
they're in the code.
00:30:00

Dave Longley: You can convert between OD4 authorization request and VPR.
You can convert between different query by example and DALE and
presentation request. They're all convertible between each other. And the
way that we solved the expanded type IRI issue that Nate mentioned with
DALE is we just list the type as the expanded IRI of a verifiable
credential and then we express a context as a claim in the DALE language.
So by doing that there's only ever one type. It's the verifiable credential
type.

Dave Longley: no there's no other expansion that needs to be done and then
you can express any other type and various information through expressing
what the context fields are that are expected and then all of the claims
you're looking for the differences that I found between query by example
and DALE the main difference is DALE assumes that there is an order to
arrays and the query by example assumes the opposite and you can't do one
with query by example and you can't do the other with dacle. they both made
different choices about which way to go. the query by example one chose to
use to ignore order because it's expected that you would want to be able to
selectively disclose things in sets regardless of what the order is. so
that you don't reveal ordering information. so that's why it made that
choice.

Dave Longley: I can't speak to how the DACA one was made. otherwise they're
really really close. So I think if this expanded IRI thing that I've
suggested pans out in implementations into the future, I think it solves
the big problem Nate had I also noticed it and that's how we went about
solving it. and I think you would only for the most part edge cases where
that those two things don't convert cleanly. it is a much more significant
task to convert cleanly to presentation exchange because presentation
exchange has many more bells and whistles for a variety of different regax
patterns and…

Dave Longley: all kinds of stuff like that. so that's what I can report on

Patrick St-Louis: Interesting. …

Patrick St-Louis: thank you.

Manu Sporny: Yeah, I guess in addition to what everyone said, Patrick, I
don't know if you've looked at there's another approach to some of this
stuff, and that is to support multiple simultaneous queries in multiple
formats all at the same time. and effectively communicate for example,
here's a use case, I'm a website and I would like to see your driver's
license and I accept driver's licenses in VC format and in W3C verifiable
credential data model format right and I don't and as a server service
whatever

Manu Sporny: I don't care which one of those things you use as long as you
get me one of those types. the server logic ends up becoming kind of, way
more complicated because now I have to process three different formats. and
I could receive any of those three. but if I remember correctly, …

Patrick St-Louis: Yeah.

Manu Sporny: I think that's like a situation where we meaning the global
community working with digital credentials is going to find itself in. and
that feels like something how do you solve that problem? it's probably
you're going to have to process each one of those inputs as kind of a
separate thing.

Manu Sporny: the software might have some kind of unification layer to it
but the more formats you have the harder that becomes the more query
formats you have the harder it becomes.

Patrick St-Louis: Yeah.

Manu Sporny: So it may be that the software that ends up becoming the most
useful to people is the one the most useful editing applications read any
format and then they have a base level of kind of things you can do on the
raw image to process it and then when you can save to one of a hundred
different formats. Right.
00:35:00

Manu Sporny: it feels like hopefully the digital credentiing space doesn't
get as out of hand as the graphics editing space but that kind of feels
like where we might be headed which is we'll take in a whole bunch of
different inputs and we have kind of one working model that's like raw red
green blue to continue the graphics editor analogy right you've got a pixel
buffer you work in 24-bit bit and 32-bit color and then when you're done
processing it so you've got kind of this unified model that you do the
processing on and then when you're done processing it you serialize back
out to a very specific format MD SDJ the W3CVC as an example just some
thoughts

Patrick St-Louis: Yeah, that's interesting. So you have a base level of
feature you support and this translation layer. But then for very specific
stuff you would need to go at supporting multiple formats and using the
specific inputs. one thing and the other thing about supporting multiple
format I think that's a bit the same idea of the protocols right like the
interaction endpoint where it offers multiple protocols and the wallet can
choose the one they want.

Patrick St-Louis: but that would also need to be supported the wallet needs
to have this logic to be able to make a decision even if it only uses one
it needs to be able to make that decision to take the choice that's being
offered and pick one. So that's more work there. And another big reason is
just the reality that some of these ministries issuer they just don't have
the budget or the resource to go through all these changes. so I don't know
how other government events are organized but definitely a pattern that
I've seen in different provinces is that there's a ministry that's
responsible for most of the digital transformation efforts and then they're
kind of tasked with assisting other ministries and navigating this digital
transformation.

Patrick St-Louis: But these other ministries themselves, they often don't
have the expertise or the funding because they need to focus on what they
actually do, which is not to implement verifiable credentials. it's like
the credential is maybe something they will issue as part of the service,
but they need to focus on their service. a good example was when we did the
mining office they are not experts and verifiable credential. They are
experts in managing mining permits. The digital trust team was tasked with
providing a way to export these permits in a verifiable credential format.

Patrick St-Louis: But it wouldn't make sense for the mining office to
update all the time to the latest credential, format that exists. they only
care about providing the data that they manage, mining permits and so on
and having a almost unmanaged way to issue these as a credential.

Patrick St-Louis: So, I feel like that's probably a reality in many places.
yeah.

Manu Sporny: Yeah, certainly. I think a number of governments have bitten
off way more than they can chew. I think they're going to find that out in
the next couple of years as things keep changing, right? I mean of course
I'm going to say this as a vendor but there are reasons that there are
specialized companies that exist that deal in anything image editing
creating serving infrastructure data center servers hosting all that kind
of stuff.

Patrick St-Louis: Yeah.

Manu Sporny: all of that stuff is stuff that governments did in the very
beginning, when there are new technologies that come out and they find out
very quickly that that is not their area of expertise, And that's where you
kind of depend on a fair and competitive market to provide those things at
the best cost the market can muster. So yeah I have seen that happen
multiple times certainly in the United States and I think where that's
largely headed is the same place it's always headed which is government it
does an initial implementation they see the value in it they want to
continue with it but they administration politics whatever has different
priorities and
00:40:00

Manu Sporny: So they have to transition over to depending on some shared
services infrastructure in the government or the public market to provide
those services.

Patrick St-Louis: Yes. Yes.

Manu Sporny: And those companies do have the scale to keep up with all the
different ways things can be done but I don't think governments are there
yet, I mean, we still have governments in the European Union believing that
they're going to build and manage and run government wallet services. And I
think they will find out over the next decade that doing that is an
incredibly expensive proposition for a government to take on their own. and
of course, you'll see all the standard we're going to depend on open source
to do this, thinking it's free, but it's not really free. you still need
the personnel to take care of the open source and manage that kind of stuff.

Manu Sporny: So the good news here is that I think exactly the same thing
that has happened over the past 40 years of software boom cycles leveling
off followed by commoditization of the technology.

Manu Sporny: I don't think digital credentials are any different. I think
we'll see the same kind of things happen there as well.

Patrick St-Louis: interesting. that being said,…

Patrick St-Louis: that's all I had to say. So is there any more comments
otherwise we can go back into the VCOM issues? Awesome. Thank you everyone
for listening to me and the input. I'll definitely take that into
consideration and we'll keep this call updated. looking at this quickly was
there any issue related to Eric

Eric Schuh: Yeah, I just had one quick thing related to issue 560. I was
working on this and I just wanted to let Manu and Dave know that to update
these tables, I'm once again going to have to make modifications to the
respect OAS repo.

Eric Schuh: So, hopefully by the end of the week, there's a PR both here
and there, to handle this issue. That was it.

Patrick St-Louis: perfect. …

Patrick St-Louis: it's still on track and that's good. so this seems like
just a data model. I'm trying to see if there could be maybe an issue
that's related to maybe some of the thing we just discussed translation or
maybe this here. yes, man.

Manu Sporny: So I'm a bit concerned about I would like to see people
experiment with that out in the field before we kind of take that on. the
only reason I'm saying this is because we're getting ready to transition
the specification into a working group and it feels like we've got more
than enough in there functionalitywise for a version 10 global standard.
and I don't necessarily want to add new speculative features for us to,
work on. I mean, clearly if they have very immediate value and are very
immediate need, we should work on them. but I would like us to kind of
focus on, PRs and reducing the number of issues, not like adding more
functionality to VCOM.

Manu Sporny: …

Patrick St-Louis: I was just saying to see…

Patrick St-Louis: if there's an existing issue that was similar to that.

Manu Sporny: right. Yeah,…

Patrick St-Louis: I definitely agreed not to add more right now. I think
Yeah.

Manu Sporny: I don't think there is. I mean, I was, also contemplating. I
was like, this, translation feature might be super interesting and
important and we might want to talk about it. And then I was kind of like,
yeah, but we're not there yet. it doesn't feel like we're there yet in the
community industry deployments where that is a significant issue or barrier
at least I'm not our customers or whoever we're engaged with we're just now
starting to see maybe there's a potential to ask for a driver's license
that's both a MDOC like an MDL and…
00:45:00

Patrick St-Louis: Yeah. Diff.

Manu Sporny: and a VC.

Manu Sporny: it feels like we're two years out on that feature set. that's
it.

Dave Longley: I don't know that we're that far out on that particular
example. That might be one of the first things that gets incubated is
convert an MDOC MDL into a verifiable credential driver's license. so that
might happen, but there is something already in the spec today, the VCOM
spec on the presentation side where a authorization request can be
generated from a VPR. that's already in the spec today it's in an issue
with a ready for PR tag.

Dave Longley: I'm not sure.

Patrick St-Louis: Interesting. menu

Dave Longley: It's one or the other.

Manu Sporny: Yeah, I mean plus one on that.

Manu Sporny: I guess what I'm saying is I would rather that we experiment
with it and try to figure out how it might work rather than openly
pontificate on this call about what it could be. I mean, given the time, if
we don't have issues to talk about and we're just waiting on PRs, we could
just end the call and tell people to raise PRs, Or even we could be a
little more aggressive about and just start assigning people to issues and
see if they're willing to take, certain issues and spend our time that way.
again, just a proposal. I'm not saying we should totally do that.

Manu Sporny: I'm just trying to figure out what do the next couple of weeks
look like before this thing, is in a charter and then once it gets adopted,
where is our focus going to shift? my hope is that the focus shifts totally
finish off the spec text and start getting ready for candidate wreck and
try to start working on test suite shift our focus to just let's get the
test suite done let's get this thing into CR and…

Manu Sporny: and beyond as quickly as we can but I don't know if that's
shared among everybody else on the call.

Patrick St-Louis: Yeah,…

Patrick St-Louis: I think focusing on the test suite but really on
interoperability and implementation I think that's the goal of the test
suite but put the focus on the actual implementation of the vehicle makes a
lot of sense. yeah so it sound like at the point and it's the stage that we
are happy with is that there's no translation but it's more of a
multi-support model meaning don't translate from one to another but it
supports multiple formats or query language that a fair assumption of what
we want to propose for at least this first

Patrick St-Louis: version. Okay. I would like maybe to have a look at there
was one about this one here. There's the word conversion. So I'm just
curious. did someone raise their hand?

Dave Longley: I was just going to modulate…

Dave Longley: what you said because you said won't translate but will
support multiple formats. I think it will support multiple formats and we
don't have to talk about translation. I do think I do expect people to
experiment with

Patrick St-Louis: Okay. Yes.

Patrick St-Louis: I made more than it's not something that will be actively
worked on for an initial version. okay.

Dave Longley: We'll have to see, but I don't think this group is saying
we're going to talk about it on calls

Patrick St-Louis: So this one talks a bit about but this is more converting
a query by example to a list of JSON pointers. so this is more a code less
a sort of spec feature. about how Yeah, they've
00:50:00

Dave Longley: Yeah, I think the point of this issue would be to show how to
consume a VPR which is in a query by example which is a serialized format
into an abstract representation…

Dave Longley: which uses JSON pointers and by doing that you can go from
JSON pointers to a number of these other languages certainly to DALE and…

Patrick St-Louis: the curve.

Dave Longley: whatever example code is pasted here I think has been
superseded by some code that's linked in the open source library at the
bottom that actually implements this. So there is a library that has some
code in it that implements converting you can take query by example and…

Dave Longley: go to a map of JSON pointers. You can take a DALE and go to a
map of JSON pointers and you can do the same thing with presentation
exchange and then go into a different format.

Patrick St-Louis: and that's okay.

Dave Longley: So it does cover translation of query languages.

Patrick St-Louis: And that's the same project you shared in the call.

Dave Longley: Yeah, that's right.

Patrick St-Louis:

Patrick St-Louis: So that's good. So they seem like this issue is a little
bit related but more explains to provide a sort of algorithm and the code

Dave Longley: Yeah, I would expect this to go into an appendix. And we have
our VPR inquiry by example section where we talk about those languages and
we talk about how you can include the other DALE directly into a VPR but I
think an appendix would show and here's how you can convert from one to the
other.

Dave Longley: So if your use case supports pres just defining this once in
one of these languages and…

Dave Longley: then letting your workflow translation service serve it up to
however many formats are needed for the wallets or exchange consumers you
want to support then you can do it and I think it would be nice to have
example conversion code in an appendix

Patrick St-Louis: interesting. I might be interested in looking at that or…

Patrick St-Louis: not to put you on the spot, but I don't Nate if it's
something you'd like to have a look at. You were mentioning you did a
little bit of this.

Nate Otto: I'll think about it. I don't have time in before our next call,
but we can follow up on

Patrick St-Louis: Okay, this go I'll listen myself and we can definitely
follow up and I'll listen both of us. Is that okay with you?

Nate Otto: Yes.

Patrick St-Louis: I think it's permanent. So this kind of maybe a very
first appendex example of how it could be without diving too much into the
details of every possible translation. I think that could be a nice thing
to have. perfect. with this being said, it is five to the hour.

Patrick St-Louis: I think this is a great time to end today's meeting. I
was a bit late. I could see when I came in the PR was closed today or was
merged. So that's some good progress. And I know we talked last time about
the holiday period. I believe if I remember correctly, next week we will
have the meeting and then come back around the 13th of January. Is that
still what we decided? I did try to go in and cancel the 23rd, 30th, and 6.
however, when I log in, the cancelled button is sort of grayed out and I
cannot select it.

Patrick St-Louis: So I'm not sure if that's a user error or a permission.

Manu Sporny: No, you have to modify the event and…

Manu Sporny: only after you modify the event can you cancel it.

Manu Sporny: But I can take care of Patrick deleting the events.

Patrick St-Louis: …

Patrick St-Louis: I'll give it another shot.

Patrick St-Louis: So, you say I need to add a agenda description or
something and then cancel it.

Manu Sporny: Yes, I think so.

Manu Sporny: I think you can go in and for the agenda say this meeting is
cancelled and then when you save it then you get the option to cancel it
which is weird.

Patrick St-Louis: Okay, I'll give this a try. So, I'll just put in the
description canceled for holidays and…

Manu Sporny: Okay, sounds

Patrick St-Louis: try to cancel it and we'll give an update if it worked or
not. thank you very much everyone and see you next week.
Meeting ended after 00:55:05 👋

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