W3C home > Mailing lists > Public > public-credentials@w3.org > June 2021

June 21, 2021 W3C CCG VC-EDU Task Force Minutes

From: Kim Hamilton Duffy <kimhd@mit.edu>
Date: Fri, 25 Jun 2021 10:29:45 -0700
Message-ID: <CAFmmOzdf+jG0wR3deP_d2hnrxSX0x7B4g-4kxF5mQrFhq7kh_A@mail.gmail.com>
To: "W3C Credentials CG (Public List)" <public-credentials@w3.org>
Thanks to Kim for scribing this week! The minutes
for this week's CCG Verifiable Credentials for Education Task Force
telecon are now available:
https://w3c-ccg.github.io/meetings/2021-06-21-vc-education

Full text of the discussion follows for W3C archival purposes.
Audio from the meeting is available as well (link provided below).

----------------------------------------------------------------
CCG Verifiable Credentials for Education Task Force Telecon Minutes
for 2021-06-21

Agenda:
  https://lists.w3.org/Archives/Public/public-credentials/2021Jun/0193.html
Topics:
  1. Introductions and Re-introductions
  2. Announcements and Reminders
  3. OBs and VCs, continued
Organizer:
  Kim Hamilton Duffy and Kerri Lemoie and Anthony Camilleri
Scribe:

Present:
  Kim Hamilton Duffy, Kerri Lemoie, Eric Kuhn, Juan Caballero,
  Marty Reed, Keith Kowal, Phil_Barker, Phil Long
Audio:
  https://w3c-ccg.github.io/meetings/2021-06-21/audio.ogg

<kimhd> Okay IP note anyone can participate in these calls,
  however, all substantive contributors to any CCD work items must
  be members of the cct with full IPR agreement signed, and you can
  go to this w three CC G site to join.
<kimhd> That it's located here at this link and that will require
  you to have a w3 account first.
<kimhd> it's free.
<kimhd> And the text of the Community contributor license
  agreement is available at this link.
<kimhd> Call notes these Minutes and an audio recording of
  everything, Sam is call are archived in the meetings repo of cce
  github account and the link to that is here.
<kimhd> And just to yourself to speak on the call just type q+
  into chat.

Topic: Introductions and Re-introductions

<kimhd> Okay introductions and reminders anyone new to the call
  today that would like to introduce themselves.
<kimhd> So just unmute yourself and talk.
<kimhd> Anyone like to reintroduce themselves.
<kimhd> Okay looks like we have a bunch of regulars today.

Topic: Announcements and Reminders

<kimhd> Okay announcements and reminders we were supposed to be
  changing it today, but I had a couple of problems, so we will be
  honest, which digit see next week, so we are here on zoom so.
<kimhd> that's probably obvious but yes next week, we will switch
  to just see we'll start with the in orientation at the beginning
  of it to just quickly show you how to use it scribe and
  everything.
<kimhd> will send out the link, which is just a web link you just
  join you don't have to install anything you can connect.
<kimhd> On a mobile APP they have Jesse has an APP you can use if
  you want, and then Lastly, we do have some conference bridges, if
  you prefer to call in over voice so those will all be in next
  week's meeting invite.
<kimhd> That is all that we have for the intro stuff so carrie if
  you're ready to take over with the ongoing discussion and open
  badge verifiable credentials think we're ready for you and I can
  paste the slides into here.

Topic: OBs and VCs, continued

<kerri_lemoie> awesome Thank you hear me okay cuz I haven't air
  conditioner on behind me, make sure i'm clear enough to be part
  yeah yeah okay.
<kerri_lemoie> i'm going to go ahead and share my screen.
<kerri_lemoie> And Hello everybody today I am working from from
  Rhode island i'm visiting family in England, so I am on one
  laptop about my my dual screen see my presentations pulling from
  a day.
<kerri_lemoie> And we this these slides are the same size we've
  been using over the past few weeks and i'm just adding content to
  them so For those of you haven't been on the call are all costs
  of farms me show you that what we're doing is discussing.
<kerri_lemoie> Open badges and verifiable credential matters as
  verifiable contentious we're just we're working on this at IMS
  billable badges work.
<kerri_lemoie> And we wanted to have these discussions here
  because we've been thinking this group has been talking about
  verifiable credentials and education.
<kerri_lemoie> And we thought well why don't we discuss what is
  one way of doing this one way of doing this one simple one
  credential is open badges.
<kerri_lemoie> So i'm not going to go through our decides because
  it's what the discussions have been every time but feel free to
  to walk through these.
<kerri_lemoie> and see my pretty diagrams here, especially
  insides two and four, and what I want to point us to, though, is
  side five because we're going to be sort of starting from here,
  and then going to slide seven so side five is like an overall
  example of an open badges vc assertion.
<kerri_lemoie> And so, this is a full on this is what it would
  look like or a close to what it should look like for an open
  badge and as a verifiable credential, this is still being worked
  on being proposed, at my Ms Ms global right now.
<kerri_lemoie> And then i'm going to take you here to slide
  seven, where we have two other examples, this is what we're going
  to talk about today, which is more focused on the actual.
<kerri_lemoie> content of the of the credential and in verifiable
  credentials is they were talking about the credentials subject.
<kerri_lemoie> And today we're talking about what a badge classic
  example of a verifiable credential crystal subject would look
  like So the first thing, given this side goes directly to the vc
  data model um spec i'm going to open it up here.
<kerri_lemoie> And so, this is a credential subject property.
<kerri_lemoie> And in this example, it has when ID, which is the
  in open badges this would be the recipient, this is a good
  example of a recipient.
<kerri_lemoie> And what they did in this data model example is
  they created from the College degree and i'm the type is a
  bachelor degree, that is a very, very simple use example of a
  credential subject.
<kerri_lemoie> The next one, is a little bit more complex, and
  this would illustrate something i'd be different in verifiable
  credentials and would be different in open badges which a
  perennial subject can be an array of claims, so it could actually
  be two different claims going to two different recipients.
<kerri_lemoie> And, but you'll see that the issuer you'll you
  don't see actually the issuer isn't in here, but it would be the
  same issuer for both of these.
<kerri_lemoie> So then, if you go back to the slides.
<kerri_lemoie> And you click on the second thing is i'm an
  example of an open badges batch class inside of credentials.
<kerri_lemoie> And this is what we're working on now in the
  proposal at IMS little.
<kerri_lemoie> So here looks a little bit different, and let me
  just good to have you here, so this is the same setup right, this
  is the credential subject.
<kerri_lemoie> And we have this idea, which is the recipient ID
  and i'm in a week or so when we have a chance we can talk about
  what could go in here for another bad right now in this example
  it's a decentralized identify.
<kerri_lemoie> And then in here, where we were talking what we've
  been talking about in this group around looking at this is Level
  two, which is a using schema.org property has credentials.
<kerri_lemoie> Which references and education or occupational
  credential and what we have in here you'll see it looks very
  simple it's very similar to a batch class.
<kerri_lemoie> And so we have this ID, which is the canonical URL
  that references a batch class batch class Jason O D we're saying
  this is a type batch class.
<kerri_lemoie> My name is summer summer you'll see in the
  documentation for of open badges in the open awesome robotics add
  the description.
<kerri_lemoie> And image of the badge and this year we can come
  back to this at some other point because there's issuer could be
  the same or different issuer.
<kerri_lemoie> The nature of the verifiable credentials going to
  make about today, but we can pass on if we don't get to it.
<kerri_lemoie> And then, this criteria and the tags some property
  alignments and then down here, where we have a little bit
  different is we have.
<kerri_lemoie> Evidence which in open badges was part of the
  overall i'm assertion what we're doing here, and I believe I got
  it right and that's part of what we're discussing is evidence has
  been reference to the hospital.
<kerri_lemoie> So i'm sorry the credential subject now has
  credential because it has credential pipe describes would
  essentially be the batch class for the content of the open badge
  being issued as her credentials.
<kerri_lemoie> And slide eight i'm going to go through all of
  these, and then we can just talk about it all on slide eight.
<kerri_lemoie> I just put some primary comparison points so in
  laughter verify look into subject can have an array of claims
  with multiple recipients, and right now there are no actual.
<kerri_lemoie> attributes for education and verifiable
  credentials just an example attributes that are in the data model
  and verifiable claims, with the batch class, we are, we are
  shooting the one recipient and one claim.
<kerri_lemoie> And they're already a bunch of badges attributes
  where we can describe this plan, so these already exists as
  context for this.
<kerri_lemoie> And we can actually add to this so, for instance,
  one of the properties, we would like to add as part of the magic
  or no proposal were suggesting is something called achievement
  type without solid lines is global's comprehensive learning.
<kerri_lemoie> Learning record the car we don't like what is this
  is achievement type and then you can add a type of.
<kerri_lemoie> This list of achievements these strings associated
  with a badge which we never had in batches before which I think
  is pretty great actually so he's been in badges for a while,
  because.
<kerri_lemoie> You know there's always been this whole like oh
  bad just something like small, it can be very significant and
  it's just as one thing that actually.
<kerri_lemoie> Open batch could always have been anything right
  given any of these things and others so um I think getting this
  type will actually add a lot of power to badness.
<kerri_lemoie> Okay, so i'm going to pause there that that those
  the only slides i'm showing today just so we can talk about this
  aspect, this part of open badges verifiable bit of just.
<kerri_lemoie> Stop sharing now.
<kimhd> And Eric is on the queue.
<kerri_lemoie> awesome.
<> hey.
<eric_kuhn> Sorry, can you hear me.
<kerri_lemoie> I can.
<eric_kuhn> Okay, so can you explain the part of that you
  mentioned about the biomedical car and how the achievement, I
  worked with that because, are they isn't open badges that spec
  and the highest level to our spec totally different things So how
  do they work together and I guess, could you just.
<eric_kuhn> dive a little deeper into that.
<kerri_lemoie> I can, and then there are folks here who wants to
  steal our but i'm going to take a stab at it and then let's jump
  in here.
<eric_kuhn> So an open badge.
<kerri_lemoie> With first raised about 10 years old.
<kerri_lemoie> That specification and described to single
  achievement.
<kerri_lemoie> And it had a better two point O has basically
  three aspects main aspect, it has an assertion inside of
  assertion as reference for batch class that batch class opens is
  an issue.
<kerri_lemoie> And I see a lot is intended to be a collection of
  achievements that are published any of those achievements could
  be open badges.
<kerri_lemoie> So it is actually an array of achievements, so
  what would be great as it actually got more harmonized so if a
  car, has an achievement type.
<kerri_lemoie> It would be an array of achievements that each
  event would be saved in a car, has an open badge that specific
  achievement in the array would be badge.
<kerri_lemoie> But it could also be something out to be a degree
  or something else, so an open badges what we're saying is you can
  deliver a single if you've been like a single degree and a
  single.
<kerri_lemoie> certificate, as with the open badge specification.
<eric_kuhn> For would it also be correct to say that many open
  badges as vcs that are a combination of degrees certificate, you
  know, etc.
<eric_kuhn> All of those fees in your wallet people I feel are,
  or is it.
<kerri_lemoie> Now it's something different, so what we're not
  talking about yet in July, which will be taught after we figure
  out this magic part of it is, is what does a car version of a vc
  look like.
<kerri_lemoie> But what you could do is you can take open badges
  as verifiable credentials and present them, so the learner could
  present them together as their achievements.
<kerri_lemoie> But the car right now, mostly is from someone else
  may jump into the car can be presented by a learner as it is as
  one set of achievements to so that is something else, but right
  now.
<kerri_lemoie> I think it might be awkward right now to put
  verifiable credentials inside of a seal I think people are trying
  to do it, but I think we should pause on that and do it right, as
  we get to that stage that's that's my opinion on it anyway.
<kerri_lemoie> Did I confuse you even more with that answer.
<eric_kuhn> um.
<eric_kuhn> I still have some questions, but our next person.
<kerri_lemoie> Okay, well, if we can get back to what Anthony.
<anthony_camilleri> Sorry, I was talking into the ether ah so two
  questions um first of all.
<anthony_camilleri> In terms of the properties, you are showing
  on the credentials subject I Is this the idea that this is what a
  batch class would look like in a vc.
<anthony_camilleri> Or do you feel that some of these properties
  could be universal, for any verifiable credential and education,
  and then the second comment is.
<anthony_camilleri> In terms of let's say referring to the image
  that represents the badge i'm just wondering allowed if that
  belongs inside.
<anthony_camilleri> Let the inside the credential subject, or
  possibly in some kind of display properties tab one level higher.
<kerri_lemoie> So we are only talking about open badges map DC
  edgy Bradley we are specifically saying this is what it would
  look like with open badges in a vc.
<kerri_lemoie> And right now the image property for an open
  badges in the batch class.
<kerri_lemoie> To that that's all we're discussing right now,
  very simply, this is the work has been proposed today and master
  but.
<kerri_lemoie> Not not vc Thank you.
<kerri_lemoie> um but it's applicable, obviously, we would like
  folks at BCG to have a look at this and have some boxing yeah,
  let us know what you think about this does that make sense.
<anthony_camilleri> It does.
<kerri_lemoie> Okay.
<kerri_lemoie> Eric i'm looking at the queue here I go one did
  you want to jump in.
<juan_caballero> No i'm good.
<kerri_lemoie> you're good okay.
<kerri_lemoie> i'm Andy.
<andy_miller> Carry.
<andy_miller> The example with multiple credential subjects is it
  important that the.
<andy_miller> The essentially the recipient be different, in
  other words, could you have multiple credential subjects all all
  with the same recipient in a single PC.
<kerri_lemoie> And a vc I believe you could but not an open badge
  so.
<andy_miller> Okay right, so the I got it okay.
<kerri_lemoie> I wanted to make a comparison.
<kerri_lemoie> Between these two, which is why should the vc
  example.
<andy_miller> got it and they did, but you wouldn't recommend
  that when it's when it's open badges you not do that, you don't
  don't.
<> that's.
<andy_miller> Great.
<kerri_lemoie> Because will open badges is simply is is one claim
  to one recipient, I want to sure, so that is why it is one single
  focus there.
<kerri_lemoie> I know that folks are very interested in having
  multiple claims with NBC and I know that's going to come up in
  the future hall that's just not what today is about.
<andy_miller> For sure yeah I think it would have been actually a
  little confused me at least if if you could do that so i'm glad
  you thanks.
<kerri_lemoie> Great Thank you.
<ottomomy> hey I just respond to any they're only I think if you
  read the the vcs back, I think it is technically possible to do
  that, to put two different claim sets in a vc where the ID is the
  same between them and you're basically making two different sets
  of claims about the same entity.
<ottomomy> Why, you know suppose I suppose that a standards body
  or something could recommend that that approach would be valid
  that they essentially want to express multiple credentials all
  signed under a single bundle.
<ottomomy> But if you think about what you lose by doing that you
  lose the ability for the recipient or anyone to express only one
  of those credentials, at a time in a verifiable sense if you
  wanted to include them in a verifiable presentation.
<ottomomy> To prove the credential that you have you would have
  to present them together, you would not have the capability of
  presenting them apart.
<ottomomy> And I think if we wanted to achieve a capability sort
  of in like a transcript like use case.
<ottomomy> Where there are a set of credentials that do go
  together, and it is important at certain times to be able to see
  that that entire set has been delivered in whole.
<ottomomy> I don't think this is the approach that I would
  recommend using in order to accomplish that use case I would
  probably recommend just sort of wrapping each credential that is
  independently verifiable inside some broader structure.
<ottomomy> which potentially could be a verifiable presentation,
  or it could be just another receipt that includes you know,
  several others with the different schema.
<kerri_lemoie> That.
<kerri_lemoie> Maybe.
<marty_reed> Yes, I just wanted to find great work the clues in
  collision of achievement tab I think is hugely valuable and kind
  of being able to bring these things.
<marty_reed> Together, and I will say that.
<marty_reed> You know, based on the guidance of this group
  there's a complex credential work group seven group is that
  we've.
<marty_reed> been working through discussing.
<marty_reed> You know what those what those multi expression
  complex credentials look like, so if anyone is interested in
  joining that please reach out to me and i'll happily invite you,
  and then and then i'm curious just from the.
<marty_reed> From the group's.
<marty_reed> Understanding.
<marty_reed> We hear a lot about.
<marty_reed> verifiable presentation I just haven't seen it, and
  so you know utility requires.
<marty_reed> That you know, then you embed a vc inside of vc
  which is not that hard.
<marty_reed> Then you can re express it individually again not
  elegant but but timely and and so just curious how the group
  thinks about verifiable presentations or any any.
<marty_reed> example.
<marty_reed> Examples of that out there in the field.
<kerri_lemoie> Or the ice cubes in the QC she has a brief answer
  for this mighty can let her go.
<kimhd> In we're gonna say the more complete treatment to another
  call, I think the only other person I might ask to chime in would
  be one who's been working on these through the.
<kimhd> The vc http API groups, but essentially if the verifiable
  presentation is a way to repackage and prove control over
  identifiers in the embedded verifiable credentials.
<kimhd> And there are different things you know, depending on the
  signature suites that you're using you know you could that's
  where you can also start doing more interesting selective
  disclosure kinds of.
<juan_caballero> Work bad.
<kimhd> You can think of it as a way to just bundle a set of
  credentials together and say also proved control over it, and
  i'll call on one I see him on.
<juan_caballero> yeah so.
<juan_caballero> So.
<juan_caballero> I think VP is i've always been a little less
  standardized than vcs so some people's weepies look like a vc.
<juan_caballero> And some get a little more creative, but the,
  but I think the, the main thing is that.
<juan_caballero> Like nate was pointing out there's there's what
  is allowed by the spec and then there's a separate question of
  like what's realistic for interoperability or portability so
  sometimes.
<juan_caballero> There are default assumptions made by lots of
  people that you wouldn't be breaking spec to vary from but it
  might make it harder to interoperate if you did, and one of those
  is like.
<juan_caballero> The you know, a VP is signed by the holder and
  it signs over the contents, a challenge and with the holders key
  material and.
<juan_caballero> How the vcs relate to each other is a little
  under specified I mean, some people have have been trying to
  specify it, but it may be, in a way that isn't really universal
  enough to be helpful.
<juan_caballero> So I think the the thing there, I mean the
  general assumption is that you know the the use case that most
  people that the assumptions, most people bring to every VP use
  case is that the subject of at least one of the credentials is
  the holder signing them, and I think some of them more.
<juan_caballero> hotly contested corner cases our VP is where the
  presenter who holds all the credentials is the subject of none of
  the sorry yeah and so that might be.
<juan_caballero> Maybe an unusually academic answer.
<juan_caballero> But it's what I got.
<marty_reed> make sense thanks one.
<kimhd> and carry out just real fast.
<kimhd> there's, as you can see in the comments there's a lot of
  extra stuff going on, and I think.
<kimhd> we'll come back to that later because, certainly when it
  comes to packaging up a set of verifiable credentials it gets
  interesting I think one of the reasons we wanted to.
<kimhd> defer this discussion is in line with why we're focusing
  on open badges first just to sort of pick something very concrete
  and specific that we can make more progress on before getting
  into the more complicated cases.
<kerri_lemoie> yeah.
<kerri_lemoie> Exactly we're really just hoping to work on this
  like one simple aspect, because we know there are a lot more
  complex cases and that.
<kerri_lemoie> Presenting one of them badge is very different
  than presenting more than one, amongst many other credentials and
  when we started looking at all of this, and especially.
<kerri_lemoie> You know, with the learning and employment or
  record rapper we were trying to figure out okay let's just really
  sort of focused, because we know that that much market happened
  in this space.
<kerri_lemoie> But we know that this is a good, solid spec that a
  lot of people understand that already using so it makes for good
  discussion.
<kerri_lemoie> They could actually lead into these other points
  it's really great to see the discussion in the chat so that we
  know how to help them direction of the spiritual to distress.
<kerri_lemoie> And do any folks have any sort of questions other
  questions about.
<kerri_lemoie> You know the properties of open badges batch class
  and how and like has credential and how that works.
<kerri_lemoie> I know many of you do already understand open
  badges, but I just want to make sure that I will share that
  again.
<kerri_lemoie> This isn't a complete open badge, but it has quite
  a bit of of what could go into in a batch in the batch class.
<kerri_lemoie> Well then, i'm there was a discussion that I was
  going to pause on, but we have a bunch of time here, so I might
  as well get into it a little bit Oh, the mighty mighty fees
  before.
<marty_reed> I had when I didn't jump on the plane.
<marty_reed> In the evidence.
<marty_reed> Is there going to be a consideration for attachments
  or how are you thinking about that.
<kerri_lemoie> um I don't think so, I mean we haven't gotten to
  that discussion, yet, so I don't have a definitive answer on
  that, but I don't know I don't think so.
<ottomomy> You know, we can't answer questions here about sort of
  like what scope will IMS select for what open badges becomes, but
  we can talk about the generic sentence in verifiable credentials
  community about.
<ottomomy> If we have a known relationship between a credential
  subject and it sort of has evidence claim.
<ottomomy> There are other properties in schema.org that can be
  used to define evidence, and so, if there's a use case for
  attachments you know you might want to go through a standards
  body.
<ottomomy> To bring that up there, or we might come up with a
  very beneficial approach here, because schema.org does have
  properties, such as like full text or something like that of a
  creative work and evidence is based on creative work, so we can
  draw.
<ottomomy> You know extra stuff if we want if there's a profile
  link I definitely encourage small file sizes inside of
  credentials, if possible, so you might have some sort of like
  hash link approach that would be useful, sorry to jump the queue.
<marty_reed> yeah there's definitely I mean there's definitely
  use case but I agree 100% nate there's there's.
<marty_reed> A lot of size constraints out there.
<marty_reed> As you traverse different networks, especially
  whenever you're repackaging and resigning with.
<marty_reed> Did within that arena so just just curious about how
  everyone's thinking about that or how we, you know influence.
<marty_reed> Or at least understand best practice as everyone's
  thinking about thing.
<kerri_lemoie> yeah I think you know not that long ago there was
  a really great discussion on the CCD pile about embedding data in
  pdfs.
<kerri_lemoie> And I don't know when that was this is carmine
  weiner here I mean do you have any thoughts on embedding evidence
  in verifiable credentials in a broadly speaking.
<kerri_lemoie> Or, or about baking data into baking but embedding
  data into into graphics or images or files.
<kimhd> Oh, I mean on the general evidence side, I mean nothing
  really specific I we made some approaches in the past around sort
  of different visual representations, but I would say nothing very
  mature enough to really more discussion right now.
<kimhd> We can get into that later one do you have anything.
<juan_caballero> i'm not sure I understood the question is that
  was the question about linking to external files in the evidence
  document or was it more generally about ways of getting a visual
  representation to travel with the vc.
<kerri_lemoie> Yes, I went when it was both it was.
<marty_reed> It was about embedding.
<kerri_lemoie> And data like we have an evidence property badges
  and it was bad, including.
<kerri_lemoie> i'm not just a link where URL to the data but also
  like embedding the evidence like an image, or something like
  that, but also, I sort of parlayed that to.
<kerri_lemoie> embedding me see data into a file and with open
  badges we also have this history of calling badge baking.
<kerri_lemoie> where you can like deliver a badge with data
  inside of it which badge itself, right now, their images, but
  sometimes they could be evidence right, so I I put them together,
  but they are two separate things.
<> Okay.
<kerri_lemoie> um but yeah I would imagine like keeping vc size
  small would would say it's probably not a great idea right now to
  embed like photos and files into a vc.
<kerri_lemoie> Is that true.
<juan_caballero> yeah I think I think that's generally the
  consensus.
<juan_caballero> I think that the issue with what what Marty
  thing in the chat here like linking to an APP dfs location is.
<juan_caballero> pretty good if you can pin the I pfs and sort of
  ensure the long term stability of that URL my my concern, always
  with embedding links to static content is that you know the
  signature breaks if that link breaks.
<juan_caballero> and
<juan_caballero> In terms of him bedding so there, there has been
  a lot of talk about.
<juan_caballero> Other yeah other ways to keep visual visual
  information it aligned sort of external to a vc there is some
  weird stuff happening in depth with.
<juan_caballero> visual stuff being sort of a verify your side
  maintained resource, but yeah sorry I maybe maybe go down rabbit
  holes.
<kerri_lemoie> No okay it's great to hear your your understanding
  of what's going on, this is about this.
<kerri_lemoie> i'll there's a lot of chat going on and then here
  and I haven't been following along Kim if you're following along
  you mind sort of like books in on this discussion.
<kimhd> yeah we basically we don't want to talk about this
  anymore.
<kimhd> supposed to see no, no, no, no, no need to apologize to
  just you know, trying to keep things back to the original let's
  see so.
<kimhd> Anthony can I call on you, because um let's see we had an
  agenda focused on open bad structure, plus.
<anthony_camilleri> years or so.
<anthony_camilleri> Because i'd like to make sure that we deal
  with carrie slides and that we closed them today ah so.
<anthony_camilleri> First of all, okay Would you mind explaining
  the slides back up it's a lot easier when I know it's challenging
  critical monitor so let's go back to the slide with the
  credential subject structure festival.
<anthony_camilleri> So.
<anthony_camilleri> let's, so I think the question to the group,
  this is yes, so if we're looking at under the credentials subject
  and the way that has been proposed to move batch class on the
  credential from tech.
<anthony_camilleri> um does anybody have any comments on this
  other does anyone have a issues they think that this could create
  or problems, or do we think that this would actually be quite a
  good way to move forward in terms of a vc is an open batch let's
  start there.
<anthony_camilleri> Go ahead king.
<kimhd> Yes, I lost a where was the QA keep that up because there
  is a week where did the achievement type go i'm just not
  remembering where that was.
<kerri_lemoie> A prayer this example, let me put it in the other.
<anthony_camilleri> example if that was going to be my second
  question, then, just focus on the achievement type, if you
  prefer, we can focus on each event type first.
<kerri_lemoie> I perseids in this example here so, so it is type
  and then achievement type, this one is just a batch in this
  example.
<kimhd> And maybe stay on the second so achievement types, so
  this one is just usual eXtensible through Google json ld you get
  to define what you mean by are you to like badges to find
  somewhere in the context listed of i'm guessing.
<kerri_lemoie> Yes, so this is an open badge type of verifiable
  credential.
<kerri_lemoie> Okay, and achievement type would be in the context
  we're open badges be three.
<kimhd> Okay okay.
<kerri_lemoie> As part of a batch class.
<kimhd> Okay, great yeah this looks good but can you keep it here
  okay great.
<anthony_camilleri> Okay, I see key from the killer.
<keith_kowal> yeah I just I mean I think there's a lot of
  conversation before around this schema has credential I thought I
  understood it, but then I guess when we talked about some of the
  new.
<keith_kowal> What the spec was looking towards and I kind of
  lost context, again, I mean, first of all.
<keith_kowal> It doesn't seem like this is normally how schemas
  are represented in the specific the vc specification, I mean
  normally there's a different block where you represent schema and
  you provide a.
<keith_kowal> URL to it, but more importantly, I mean mostly it
  within normally within vcs if you declare, you have a vc and your
  subject the vc.
<keith_kowal> That you mean you basically are implicitly saying
  that this person has this credential, so why would you even need
  to reference this reference back to the schema.org has credential
  thing.
<kerri_lemoie> that's an excellent question i'm going to refer a
  different, it is still here because they can work on this a bit
  before I got too involved as much as I am now me Would you mind.
<ottomomy> Taking a quick stab so first of all, the the usage of.
<ottomomy> The word schema here on the screen is not the same as
  the usage of the schema to describe the stigma of this.
<ottomomy> Of this whole verifiable credential the schema here is
  just a reference to the schema.org namespace.
<ottomomy> which contains the property has credential, and so the
  credential example that you're looking at here makes one simple
  couple couple claims makes it evidence claim and a credential has
  credential claim.
<ottomomy> The host credential claim is sort of the meat and
  potatoes of what we've been talking about in this group for the
  last couple months, which is the idea that there is a defined
  achievement and that the subject of this credential has that
  achieved they have met the criteria of a.
<ottomomy> That has been defined by the credentials Creator, you
  know it's a bundle of criteria educational opportunity, maybe an
  image metadata it's all this this this define the thing like here
  is a degree in computer science.
<ottomomy> A bachelor's of science, you know in computer science
  separately that could be a defined the credential and we're
  making the achieve the claim that.
<ottomomy> This credential subject has that thing, and so, if you
  wanted to express, what is the schema of this credential.
<ottomomy> That would be what you're saying is that this is a
  defined achievement credential where there is a has credential
  relationships or at its core, and then the the properties of the
  specific credential that this learner has.
<ottomomy> This is an example drawn from IMS open badges were
  which has the vocabulary for all those different properties,
  mostly drawn from the schema.org vocabulary, as well as the
  humps.
<ottomomy> May I.
<keith_kowal> just say, like every vc that we issue today has
  contents in a bit and we don't have this I mean we issue a
  degrees we issued diplomas today as vcs using vc attributes per
  the vc spec and we don't need to have this has credential thing
  so.
<keith_kowal> If we're trying to align open badges to vcs, why do
  we need this thing when it's to me it's no different it's a
  different type of payload it's but you're trying to do the same
  thing Oh, thank you Lee.
<ottomomy> And i'll make only one brief response when you go back
  to the queue.
<ottomomy> That the generic concept of a defined achievement is
  powerful and a defined achievement that has a different type and
  has come in representation and ability to be serialized across.
<ottomomy> Many different types of that achievement is a powerful
  concept, and so, if you had eight different schemas to represent
  a different credential types, one being degree, one being.
<ottomomy> certificate, one being volunteer opportunity and all
  these different things that have common attributes, such as
  criteria.
<ottomomy> For completion and the notion that the the creator of
  this thing is the entity that can confer this on you.
<ottomomy> If you did those all in different ways, then you've
  got to build consumer software to handle all of those different
  schemas whereas if you build all of these different use cases
  with.
<ottomomy> One common defined credential schema that's pretty
  powerful in terms of building an ecosystem, where it's affordable
  to build consumer software.
<anthony_camilleri> Okay, thanks for being sort of I believe TIM
  isn't.
<kimhd> Actually nate address what I was going to cover did we do
  I see Eric on the queue up did we cover what eric's question was.
<eric_kuhn> uh no I was gonna ask, or is there, like a use case
  document tying back to when these this open badge type of model
  would be used like sister kp 12 buyer and.
<eric_kuhn> You know coursera life courses, you know what's the
  It is this whole discussion tied back to a set of real world
  examples that we discussed earlier that I missed.
<kerri_lemoie> Is K, I think that open badges or is a decade the
  old spec that started at mozilla foundation and over the IMS
  global in 2017 and.
<kerri_lemoie> It can be used for all of those books and to there
  are like 10 years worth of use cases and samples, so we could
  probably pull something up, but all of those you mentioned are
  true, they are used for everything from you know after school
  programs to.
<kerri_lemoie> Your recognizing courses to recognizing
  certificate programs really runs the gamut and they read they
  really are used to recognize one single type of achievement or
  still that could happen like anywhere that you type.
<kerri_lemoie> Two is very broad, in education, occupational
  applications.
<eric_kuhn> I see So the idea here is that we get this sort of
  nomenclature right and then, if i'm a student in the fifth grade
  graduating, then I would get one of these for that achievement.
<eric_kuhn> If i'm completing a course in college I would get one
  of these, or if I completed a course on Microsoft learn i've get
  you know, a vc with this type of data object that.
<eric_kuhn> That outlines what this achievement is and like that
  that whole system is based on what we're talking about now, or is
  there a big looming feel our type of other spec that we're not
  talking about now that totally changes what I just spent in.
<kerri_lemoie> that's right, so what we're talking about see I
  don't even see, I have highlighted inside here is the IMS global
  open badges specification that describes when issue.
<kerri_lemoie> And let's do our specification is a separate
  specification but open badges can be inside of a car as a single
  achievement that are listed in a car.
<eric_kuhn> Okay, and a car would be its own BC that's like a
  single vc with lots of different achievements line item.
<kerri_lemoie> I tested excess that's it that's a potential
  possibility.
<eric_kuhn> Okay.
<anthony_camilleri> And maybe this is a good time to take.
<anthony_camilleri> My own spot on the tool.
<anthony_camilleri> First of all.
<anthony_camilleri> I would like to just reinforce what an eight
  sets about the achievement type in the sense that understanding
  what the has credential is actually representing has.
<anthony_camilleri> Gone on use cases all afternoon it's it's a
  very, very, very useful feature because it allows us to explain
  it, this is a badge if it's an assessment it's an entitlement.
<anthony_camilleri> And as we move further along it would also
  allow us to relate these concepts to each other and maybe we
  wouldn't necessarily wants to let's say specify the relations
  ourselves but it opens up those possibilities.
<anthony_camilleri> To answer a little bit more, the question on
  whereas this asked for more specifically in the use cases and
  the, why are we focusing on this, and the reason we're doing this
  is.
<anthony_camilleri> What firstly as Kerry said, I mean open
  badges is a it's an established standard, it has a very strong.
<anthony_camilleri> Open Source Community behind it and you know
  you can't do anything in this space without treating open badges
  but secondly open badges have the lovely feature that they are
  relatively simple so starting with an open badge and figuring out
  what an open badge could look like.
<anthony_camilleri> is a good way to move the discussions forward
  about what a basic verifiable credential for education moves
  forward.
<anthony_camilleri> In the meantime mighty group has started to
  work on the complex credentials and is trying to figure out, then
  how let's say these things would work with more complex
  descriptions.
<anthony_camilleri> Of the basis of this work and will slowly
  begin to bring this back to the main group for once we finalize
  the open bachelor.
<anthony_camilleri> So I just wanted to let say remind of this
  approach because let's say there's a logic there all the building
  it in in this way and isn't really about.
<anthony_camilleri> The preferring open badges, but it really is
  a matter that would say open badges being so foundational also
  give us a good foundation to move the world forward.
<kerri_lemoie> As well said, thank you very much, I can see the
  queue so someone could let me know who the next.
<anthony_camilleri> Keith why don't you just beat your comment
  it's a lot easier, I think.
<keith_kowal> Well, I think I totally get the the need for
  achievement types and I, and I echo that.
<keith_kowal> ability to group badges but I guess the question
  i've been why don't we use the the type fields provided within
  vcs because I think that would you know so.
<keith_kowal> You could you can, in the context you could have a
  listing of types and then you could actually list out the type of
  credential in a type field.
<keith_kowal> And then you know that would make it much easier
  for standard like presentation exchange to consume that type
  right, because I think.
<keith_kowal> You know presentation exchange and the input
  descriptor allows you to specify well it's pretty open, but it
  makes a lot of sense that you would put type in there, I think
  some companies are already doing it that way, instead of schema.
<keith_kowal> So why not, why not move more towards a tight
  construct instead of this has credential achievement construct
  which still seems like something new to me like no i'm not really
  seen before, where I have seen pipes before.
<anthony_camilleri> May I call on needs to answer that because I
  believe you have already answered this in the chat.
<anthony_camilleri> Or if nate isn't available, maybe carrie
  could go.
<kimhd> nate seems to be unmuted I think.
<kimhd> That we should carry as well.
<anthony_camilleri> Okay, the comments that was written was that.
<anthony_camilleri> i'm not sure which way around it is but
  that's one of the i'm trying to find it in the chat history but
  uh pair one of the concepts only supports IRI is, while the other
  one supports references to formalize to 40 tables with the onset
  of the nate has written earlier.
<kerri_lemoie> I don't know if I can speak to let me quote about
  that i'm sorry.
<ottomomy> Sorry guys i'm on vacation and I was.
<ottomomy> Making coffee and breakfast on what I was mentioning
  is that achievement type is a specific term that originated in
  IMS global car.
<ottomomy> And that term expects a string and there is a
  particular defined vocabulary ncr it's likely that open badges
  will sort of pick that up kind of as is it's an optional
  property.
<ottomomy> And the overall concept of credential type is also a
  useful concept but, again, we have this idea of a.
<ottomomy> sort of reusable defined credential structure where
  there are several different types of credential that can use the
  same schema and how that gives us benefits for cost of
  implementation.
<ottomomy> And sure yeah we make great sense if you wanted to put
  it in the type property and may, and so you know I am so i'll
  have a conversation about this in the open badges work urban
  maybe when the achievement type.
<ottomomy> feature is discussed, maybe, people will say, well,
  now that we're here in a world of a verifiable credentials, it
  makes more sense to just have three types and.
<ottomomy> credential type and will do verifiable credential then
  open badge then some other IRI for the ultimate achievement Type
  two layers down the schema over.
<ottomomy> we're defining the relationship here between the the
  credential definition and what its type is and the ability to be
  able to express credentials of different types in the common
  schema.
<ottomomy> I think that has a lot of benefits and i'm really
  interested in exploring and building an ecosystem out around that
  idea, so I see that there is a useful difference between the
  credential type and achievement type.
<anthony_camilleri> nighttime not for that you've fully answered
  the question because I think the I if i'm not mistaken, the
  question is about having type and achievements type, both within
  the has credential rather than.
<anthony_camilleri> Because, if you look at what's on screen
  under ID we have tied.
<ottomomy> yeah ones and irin ones not.
<ottomomy> The difference that we start with.
<ottomomy> I mean it's just putting a term from car in the way
  that car did it It may be that if this group advises the bad be
  done a different way that.
<ottomomy> The open badges work group I mean we've got
  representatives here on the call will hear it, and I think it It
  certainly makes sense to maybe say oh OK, we can have an array of
  types in the credential type property and the second one will
  just be an IRI for the concept of batch.
<ottomomy> Totally reasonable.
<ottomomy> says in the chat it makes sense that there's a
  technical type in a conceptual type and that those are not
  completed there's an argument to be made that direction as well.
<kerri_lemoie> yeah These are all good points and it's the first
  time we've actually really discussed talking about achievement
  type with open badges specifically in this group to so it's great
  to hear all of those perspectives.
<kerri_lemoie> Are there other folks in the queue right now.
<anthony_camilleri> Who is currently empty.
<kerri_lemoie> Okay Anthony did you have thoughts of where you
  want to take it from here, we have about 10 minutes left.
<anthony_camilleri> I didn't have any particular thoughts.
<anthony_camilleri> Where where we wanted to take it.
<anthony_camilleri> Further from here, honestly, so I think he
  either over to you, or we could open it for just general
  comments.
<kerri_lemoie> yeah why don't we do that does anybody have any
  general comments about this, as it stands.
<kerri_lemoie> i'll leave this up here for now.
<kerri_lemoie> Oh yeah.
<phil_barker> Can I jump in film.
<phil_barker> i've just got a question we.
<phil_barker> a while back, we spent quite a few.
<phil_barker> coins discussing sort of high level conceptual
  things that might be covered by verifiable credentials.
<phil_barker> honing down on three in the end, which were.
<phil_barker> can't entirely remember but activities credentials
  stroke, qualifications and a third one, which I think something
  like affiliations, where do they fit into this type of not on
  your where's the information that this is a.
<phil_barker> credential stroke qualification.
<kerri_lemoie> That Anthony percent from the cues and answer to
  those questions.
<anthony_camilleri> Ah, well, I actually would.
<anthony_camilleri> be similar and they can answer it while at
  it.
<anthony_camilleri> Oh.
<anthony_camilleri> Oh, under this model, these would be
  different, what I pulled.
<anthony_camilleri> Different types here um.
<anthony_camilleri> I think when we finish all the work we might
  need to go through the naming of things I really don't want to
  facilitate naming of things right now.
<anthony_camilleri> But, to answer your question directly and
  achievements type could describe an activity to describe an
  entitlement you could describe the qualification I believe
  there's some examples here um but my.
<anthony_camilleri> My other question which was really.
<anthony_camilleri> hairy is.
<anthony_camilleri> If you were going to go in this direction
  with open badges what I have been wondering to myself allow that
  is that yeah, what do you need the type called bad because.
<anthony_camilleri> in the sense that we would have saved this
  new batch class structure that, especially if it included
  achievement type of could be used to describe pretty much any
  educational concepts.
<anthony_camilleri> And a badge isn't an educational concept per
  se and badges own way of describing educational concepts unless
  i'm getting very confused.
<anthony_camilleri> I would say, the whole thing is that I would
  say the badge would contain one of all these other things now, I
  say this from the position of not being an open badge expert, so
  I might have some conceptual confusion over there, but would
  appreciate it if you could clarify.
<kerri_lemoie> Sure um well this achievement.
<kerri_lemoie> Type comes from the car, so it is a property of
  the car and this is so that if a badge is one of the achievements
  in a car, it can be indicated as such.
<kerri_lemoie> And so we are proposing right now just just to
  start this discussion at the working group to you know, use the
  same properties that are already being used in a car.
<kerri_lemoie> i'm an overall the Anthony and open badge is both
  a technology, but also a and a description of an achievement
  right, so it is both at the same time.
<kerri_lemoie> But yes, this specific property, we would still
  need that because it is needed in car and if we're using the same
  properties and that's what decided on them we would keep that.
<kerri_lemoie> In the list and into in a way, I think, actually
  would be great if we kept it in the same we do do it this way is
  others said because.
<kerri_lemoie> If I have in my verified presentation, or just
  like generally in the world, I have a vc that is a badge, but I
  also have a vc that is.
<kerri_lemoie> A potential that could be a difference between the
  two, and what I don't see on this list for sealer, though, is.
<kerri_lemoie> Is micro credential and and i've noticed over the
  year so there's been what I think, in my opinion, is a conflation
  between open badges and micro credentials and this assumption
  that every open badge isn't maker credential.
<kerri_lemoie> which I don't think that's true my potential is
  its own type of credential that can be represented by an open
  badge so that opens up a whole philosophical discussion that
  probably goes beyond assessing some of the specifics here.
<anthony_camilleri> um okay so um.
<anthony_camilleri> i'm next on the cube I will only take one
  sentence, because I see others with it um I from my personal
  perspective, I very much support the idea of the achievement
  type, but they would have significant reservations about taking
  the cla list of achievement types, as even recommended.
<anthony_camilleri> We tried doing something with this list in
  Europe, and it cost us nothing but pain and we gave up trying to
  work with it, I could speak more at length and other time as to
  the reasons of that.
<anthony_camilleri> I see.
<anthony_camilleri> On the queue that are there are comments from
  Keith which I would urge you to say or a lead and I also see
  Phillip on the cube, so I will put Philip on the queue first and
  then Keith if you don't mind justify your chats verbally, I think
  it will be a lot easier.
<phil-t3> yeah.
<phil-t3> I just wanted to point out that, in the context of
  badging I think one of the reasons that has largely been.
<phil-t3> redefined, if you will, and colloquial use as a micro
  credential is twofold one is that in doing the in doing so in the
  academic world it moves it away from assertions that have credit
  involves, which means a whole other round of.
<phil-t3> complicated.
<phil-t3> political discussions with institutions, because only
  certain parts of institutions are are allowed to assign and award
  credit, and so one of the reasons why that.
<phil-t3> Open badge to call in that caught that connotation was
  to separate it from the things that typically are under the
  control of the provost and a registrar so there's there's that
  history that has to be.
<phil-t3> Considered in this, and the second is that the open
  badge in one sense, was defined initially to be as.
<phil-t3> As applicable to the granularity of the issuer as they
  wish, so you could define an open badge to be a very, very tiny
  thing or an open badge to be a culmination of a several years
  worth of work, because the badge itself had no other way of.
<phil-t3> of providing anything about the magnitude of the
  achievement and the work that was involved in it.
<phil-t3> And, as a consequence that was both a plus and a minus
  because it meant that people receiving the badges had no real
  sense of Is this a big piece of work or to the person just show
  up for class that day.
<phil-t3> versus, this is a PhD know leave it at that.
<anthony_camilleri> Okay, thank you very much, Philip i'm going
  to pass the floor to Keith.
<keith_kowal> um yeah basis echo what I said I think i've been
  talking with Kerry and others it's like you know.
<keith_kowal> So let's say we come up with this standard, you
  know I think you know and we're interested in this, but we have
  to build you know let's say P O co things like that, so we would
  most likely use.
<keith_kowal> Common you know standards, as we see them
  developing and you know, in the US, at least presentation
  exchanges is emerging as the dominant standard for making proof
  requests.
<keith_kowal> For these type of credentials, so I think my
  question always come back to like.
<keith_kowal> How we're going to approach that like today proof
  requests often ask for things like schemas or four types, which
  is you know still out a lot of complexity about it, but now
  we're.
<keith_kowal> adding another layer of complexity this concept of
  achievement type and I can definitely see.
<keith_kowal> The us having a treatment types in the Europe
  having achievement types and Australia having achievement types,
  because no one can agree on what achievement types are what the
  list should be.
<keith_kowal> I think it's just you're adding another layer into
  complexity, for how to verify or would actually get this data and
  I understand we're not like there's a lot of runway, we need to
  figure out this stuff, but I think these are just some that
  really questions to start maybe thinking about.
<anthony_camilleri> I see Kim on the cube can, I believe.
<anthony_camilleri> You will be next on the queue and also have
  the honors of closing the meeting.
<kimhd> Okay, great.
<kimhd> yeah I would just have to Keith.
<kimhd> provide some specific structural examples of what he's
  talking about is unfamiliar with the presentation change spec and
  so you know, I think.
<kimhd> i'm not i'm not really seeing.
<kimhd> The comment about like how to say i'm not seeing how
  anything that we're doing here precludes or makes presentation, a
  change more difficult, I mean maybe if it's.
<kimhd> A certain interpretation that he's making that Keith
  you're making about how it would be used in presentation I shane
  show, I was wondering if you could just write up something to
  show what you mean and we could, in fact.
<kimhd> Talk about that in the next meeting, or, if you want to
  send it to the mailing list that would be good to.
<kimhd> Very also very interested in using these with
  presentation, a change but i'm not really seeing the specific
  problems so.
<anthony_camilleri> Okay, I see a lot of consensus.
<anthony_camilleri> From the group on this is 6pm CET European
  time i'm sorry I don't know Pacific.
<anthony_camilleri> Thank you everyone for joining the call see
  you next week, hopefully, we will kick the tires on the gypsy
  Thank you very much for a super productive call.
<kerri_lemoie> Thanks everybody.
<stuart_freeman_(georgia_tech)> Thanks everyone.



-- 
*Kim Hamilton Duffy*

Senior Technology Architect

*MIT Open Learning | Digital Credentials Consortium*

kimhd@mit.edu
Received on Friday, 25 June 2021 17:30:20 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 June 2021 17:30:21 UTC