Fwd: [MINUTES] W3C CCG CCG Verifiable Credentials for Education Task Force Call - 2022-08-15

> Begin forwarded message:
> 
> From: CCG Minutes Bot <minutes@w3c-ccg.org>
> Subject: [MINUTES] W3C CCG CCG Verifiable Credentials for Education Task Force Call - 2022-08-15
> Date: August 17, 2022 at 2:01:59 PM EDT
> To: public-credentials@w3.org
> Resent-From: public-credentials@w3.org
> 
> Thanks to Our Robot Overlords for scribing this week!
> 
> The transcript for the call is now available here:
> 
> https://w3c-ccg.github.io/meetings/2022-08-15-vc-education/
> 
> Full text of the discussion follows for W3C archival purposes.
> Audio of the meeting is available at the following location:
> 
> https://w3c-ccg.github.io/meetings/2022-08-15-vc-education/audio.ogg
> 
> ----------------------------------------------------------------
> VC for Education Task Force Transcript for 2022-08-15
> 
> Agenda:
>  https://lists.w3.org/Archives/Public/public-vc-edu/2022Aug/0005.html
> Topics:
>  1. IP Note
>  2. Call Notes
>  3. Introductions & Reintroductions
>  4. Announcements
>  5. Main Topic: Kayode Ezike with Updates on MIT/DCC Credential 
>    Status WOrk
> Organizer:
>  Kerri Lemoie
> Scribe:
>  Our Robot Overlords
> Present:
>  Xander, Andy Miller, John Kuo, Kerri Lemoie, Jonathan Bethune, 
>  David Chadwick, Stuart Freeman, Chandi Cumaranatunge, Mike Peck, 
>  Shawn Butterfield, Kayode Ezike, Dave McCool (Muzzy Lane), Joe 
>  Kaplan, Andy Griebel, Kimberly Linson, Mahesh Balan - 
>  pocketcred.com, David Ward, James Chartrand, Deb Everhart, Dmitri 
>  Zagidulin, JennGreene, Janko, Jon St. John, Nate Otto, Akshar 
>  Patel, Jim Kelly, Jeff O - HumanOS, Simone Ravaoli, Kaliya Young, 
>  Marty Reed, TallTed // Ted Thibodeau (he/him) (OpenLinkSw.com), 
>  Allyson Parco, Eric Shepherd, Phil L (P1)
> 
> Our Robot Overlords are scribing.
> Kerri Lemoie:  Hello everybody Welcome to the August 15th BC edu 
>  call today our main topic will be on credential status kod a DK 
>  from MIT digital credential Consortium is going to present us on 
>  some of the latest work that he's been doing on that over there 
>  and we can pour we get started with that let's go through a few 
>  boilerplate items first is IP note everybody for anyone.
> 
> Topic: IP Note
> 
> Kerri Lemoie:   Came to speed in these calls how any.
> <kerri_lemoie> create an account: 
>  https://www.w3.org/accounts/request
> Kerri Lemoie:  You substitute of contributions to any of the ccg 
>  work items must be done by members of the ccg with full IP are 
>  agreements signed and to do that you can create an account at w3c 
>  and put this in the chat for you and then also use this link to 
>  join the ccg.
> <kerri_lemoie> join the CCG: 
>  https://www.w3.org/community/credentials/join
> 
> Topic: Call Notes
> 
> Kerri Lemoie:  Hey second All call notes for this call are 
>  recorded and there are minutes being taken by are called our 
>  robot Overlord is do the transcription I wish you'll see in the 
>  chat we are also doing a Wheels do a video recording of This call 
>  which can be found later we can send out if you need it so coyote 
>  will have some slides today so it'll be nice to have that 
>  recorded we use a cue system.
> Kerri Lemoie:   To to ask questions and participate.
> Kerri Lemoie:  Conversation so if you would have a question or 
>  would like to say something please kill yourself first you do 
>  that by putting a q+ and the chat just like I did there you could 
>  also do to U plus and then say a note about something you want to 
>  say and that lets us know if you're responding to something 
>  that's being said right away so we know when to pull you into the 
>  conversation so that is very helpful and then to remove yourself 
>  from the queue just uq-.
> Kerri Lemoie:   And we get something.
> Kerri Lemoie:  They did that wrong okay remove something from the 
>  queue okay.
> Kerri Lemoie:  Because the floor all right when I say we hit q- 
>  we're actually looking to see who is in the queue.
> 
> Topic: Introductions & Reintroductions
> 
> Kerri Lemoie:  Okay next let's do some introductions and 
>  reintroductions is there anybody here joining us today for the 
>  first time that I would like to introduce themselves or anybody 
>  who's been here for a while and I would just like to say hello 
>  again and let us know a little bit about something they're 
>  working on if so go ahead and put yourself in the queue.
> Kerri Lemoie:  Yeah I'm going to queue myself up real quick here 
>  one thing I haven't mentioned is that I started a new role at the 
>  beginning of August I have the digital credential contortion so 
>  I'm going to be there director of Technology it doesn't really 
>  change much for BC I do I will still be here doing all of that 
>  work because the DCC has made her is really been backing all of 
>  this work and really significant way making sure that we have 
>  this open Community to work and so I'm really happy to be able to 
>  continue the.
> Kerri Lemoie:   Work there and help to do what I want to do which 
>  is in.
> <deb_everhart_(credential_engine)> congrats Kerri!
> Kerri Lemoie:  Adoption understanding of verifiable credentials 
>  that had that has been my my shift recently.
> 
> Topic: Announcements
> 
> <kerri_lemoie> VC email list: 
>  https://lists.w3.org/Archives/Public/public-vc-edu/
> Kerri Lemoie:  There's no other introduction to reintroductions 
>  next we have announcements and reminders one is that if you 
>  haven't joined it yet we do have an email list and I met email 
>  this can be joined by going here put this in the chat for you so 
>  we have it in our notes later.
> Kerri Lemoie:  The great list to join you don't have to be 
>  necessarily technical drawing this or you work on technical 
>  projects if you just want to stay in the loop on what's going on 
>  in Education and Training and achievement credentials that's the 
>  list to join this is where we try to have all those 
>  conversations.
> Kerri Lemoie:  And the other announcement that I have is that 
>  tomorrow at 11:00 and noon sorry tomorrow at noon Eastern is the 
>  weekly ccg meeting.
> Kerri Lemoie:  They recommend doing that as much as you possibly 
>  can to learn what's going on in the community as a whole.
> Kerri Lemoie:  Does anybody else have any announcements or 
>  anything they'd like to share.
> Kerri Lemoie:  It's the money I see that you just joined us would 
>  you be interested in telling folks about our next week's call 
>  will be about it sounds really great.
> Simone Ravaoli:  Yes hi hello everyone I hope you can hear me so 
>  we are going to have as a guest get anything on this sea of 
>  Gattaca Gattaca is a nest site a company based out of Spain and 
>  they've been involved in the end building and implementing all 
>  the ab c-- specification so that's that would be the European 
>  blockchain service infrastructure particularly they have been 
>  looking into version 2 of.
> Simone Ravaoli:   Of the did method did.
> Simone Ravaoli:  And they've risked to share some Reflections 
>  around what is the difference between version 1 and 2 and some of 
>  the implications that version 2 has with regards to privacy 
>  regulations in Europe also known as GDP are but they are a 
>  Community member that has really just been doing a lot of work 
>  from Europe and so that would be I think an interesting update 
>  from what's Happening outside of North America.
> Kerri Lemoie:  Thank you so much that sounds great I'm really 
>  looking forward to it okay anybody else have anything else before 
>  we move on to our main topic.
> 
> Topic: Main Topic: Kayode Ezike with Updates on MIT/DCC Credential Status WOrk
> 
> Kerri Lemoie:  Okay why don't I then introduce Kayode. Kayode is 
>  working on this great project related to credential status he's 
>  working on making it possible for issuers to manage the status of 
>  their credentials and for verifiers to check status info, 
>  especially a revocation or suspension so he's going to show you 
>  some of his work today and then feel free to ask questions and 
>  then we can see what we can learn about it well.
> Kerri Lemoie:   Tim Kennedy.
> Kayode Ezike: Slides: 
>  https://docs.google.com/presentation/d/1UYFcVLYaz8jhmmYM8l43cBg-mtINlxlFMESCbYupjwo
> Kayode Ezike:  Thank you thank you characters interesting here's 
>  the link to the presentation that be using for today I know we 
>  generally do not show video for van with purposes I would do that 
>  really quickly just to introduce myself and turn off every but 
>  this guy is he gay again and brief introduction is that I've been 
>  involved with digital credentials since around 2018 but that time 
>  that I began my master's program around proper credentials and 
>  personal data storage management.
> Kayode Ezike:   Since then I've been contributing to number of 
>  Open Source projects.
> Kayode Ezike:  Optical pulse of my own and these days I primarily 
>  support the digital credentials Consortium through MIT as well as 
>  Gobekli which is the startup it's also doing great work in this 
>  space so thank you for having me and I'll put up my screen so 
>  that an alternative you'll see how can see we have.
> Kerri Lemoie:  Kayode I'll keep an eye on the chat for you.
> Kayode Ezike:  Thank you okay so.
> Kayode Ezike:  If explosion one second.
> Kayode Ezike:  Spell with me.
> Kayode Ezike:  Okay I think if I try none.
> Kerri Lemoie:  Not yet not yet nope.
> Kayode Ezike:  Are you able to see my screen now.
> Dmitri Zagidulin:  If you're transmitting from a Mac there you 
>  might need to update the permissions in the control center which 
>  means you have to restart the browser and redrawing.
> Kayode Ezike:  But you're trying to do something like that but it 
>  was it.
> Dmitri Zagidulin:  One of us can also.
> Dmitri Zagidulin:  Screen share the slides and just thanks Kerry.
> Kerri Lemoie:  Yeah I can do that that's true.
> Dmitri Zagidulin:  I'll probably be faster chaotic.
> Dmitri Zagidulin:  Say next slide.
> Kerri Lemoie:  Yeah just let me know in and we pull it up here 
>  hold on one second.
> Kayode Ezike:  Animations hopefully syncs
> Kerri Lemoie:  I am just clearing everything out of the way here 
>  on the way everybody's happy about the desert play.
> Kerri Lemoie:  Okay can you see.
> Dmitri Zagidulin:  Yep we can see your screen.
> Kerri Lemoie:  To be this clarity.
> Kayode Ezike:  Great yes I can see.
> Kerri Lemoie:  Oh there you are okay you can see the head you can 
>  take it from here now just let me know when you want to move to 
>  the next slide.
> Kayode Ezike:  Awesome thank you Kerry so today we'll be 
>  discussing a topic I was cussing to what that would get a 
>  credential status hosting and for scare anybody away this is not 
>  a new specification for credential status is more so concerning 
>  ourselves with how can we help issuers to expedite their 
>  deployment of criticized infrastructure by leveraging familiar 
>  services to them and and hence sort of taking care of it.
> Kayode Ezike:  Origin authorization concerns that.
> Kayode Ezike:  To them and so this in terms of what we're 
>  discussing today I'll be giving a brief background of the topic 
>  and we'll be going through a bit of the state-of-the-art who tell 
>  you what our goals were let me Design This solution as well as 
>  what the actual solution is will then show you a little bit of 
>  code before you show you the demo of what we've done.
> Kayode Ezike:  Granger status management has been around for some 
>  time and one of the earliest conceptions of it really came from 
>  certificate authorities with the notion of certificate revocation 
>  lists which is a list of certificates that an issuer has revoked 
>  prior to its from its expiry and that is sort of the the main 
>  model that we think about in this space when we think about 
>  country status management.
> Kayode Ezike:  Generally we think about a resource that is 
>  managed by an issuer that informed the public about the state of 
>  outstanding credentials and there's been a lot of work done in 
>  this space over the years in terms of Standards development and 
>  developer tools for how to manage the status of these credentials 
>  as well as how to verify the information on them such as 
>  revocation or suspension however storage and authorization have 
>  kind of.
> Kayode Ezike:   Been left out of scope largely and.
> Kayode Ezike:  Good that it would be great if we can simplify 
>  that for issuers and their deployments.
> Kayode Ezike:  Thank you so next slide great so I want to give a 
>  quick definition of printer status according to the v-spec mental 
>  status is an object value property that enables the discovery of 
>  information about the current status of the verifiable credential 
>  such as whether it is suspended or revoked.
> Kayode Ezike:  So in other words now before we go on another 
>  words it gives consumers of this credential and idea of the 
>  issuer's view of the current validity of a credential next please 
>  so here on the right we have an example of a verifiable 
>  credential that has a special status property on it but one thing 
>  to know is that it is an optional value so you not every verify.
> Kayode Ezike:  Krista have this field on it.
> Kayode Ezike:  If it does have the field on it it must Define two 
>  main properties which is ID and type and it must also valid like 
>  the remaining properties are specified by the type field so in 
>  this example we know that we have a status festering 21 entry the 
>  type and the three types of rather Fields below it status purpose 
>  statuses index in status credential relate to or rather are 
>  defined in the stands for students.
> Kayode Ezike:   T10 tree.
> Dmitri Zagidulin:  Toyota quick question I do we want to wait 
>  till the end for questions or do you want to encourage people to.
> Dmitri Zagidulin:  If they encounter something or when I ask 
>  questions about each slide to raise their hand.
> Kayode Ezike:  I'm happy to take questions as they come up.
> Kayode Ezike:  Is it just the two Fields so the state-of-the-art.
> Kayode Ezike:  There have been a lot of great work and done in 
>  sequential status space and one of the early conceptions of this 
>  or examples of this came in 2018 from hyper legit Indy via their 
>  HIV ew1 one penetration spec and they use a bunch of cool 
>  technology cryptographic graphic accumulators to determine which 
>  credentials have been revoked and interestingly enough I learned 
>  in the suspect that it seems that this actually is.
> Kayode Ezike:  Predates to know now have medication.
> Kayode Ezike:  Like I found to be interesting and a couple years 
>  later the ccg would develop a speck of their own via the 
>  revocation list 2020 spec and soon after be a refined version and 
>  the status was 20 21 spec which sort of granted a more General 
>  notion of status that goes beyond just a vacation but also 
>  support system suspension and other forms of criminal status.
> Kayode Ezike:  Next please and the most for all these 
>  specifications all right one more back sorry.
> Kayode Ezike:  So this is the this point on hosting so for any 
>  all these specifications is important for verifiers to be able to 
>  to check this the status of the credential somehow right and the 
>  most obvious way to do this is to host the resources and in a 
>  public location namely a distributed Ledger or a short controlled 
>  website.
> Kayode Ezike:  Any questions here.
> Kayode Ezike:  You can move on next slide please.
> Kerri Lemoie:  No questions yet sure.
> Kayode Ezike:  Great so I want to talk to you all about status 
>  list 20 21 this is the suspect that we use in our design and the 
>  it's one of the more advanced specs out there and in my opinion 
>  and this is examples of the left what you see is a credential 
>  from a couple slides ago so nothing new here yet we have 
>  verifiable credential had the credential status on it and it has.
> Kayode Ezike:   As all the fields that we discussed earlier.
> Kayode Ezike:  Now if you pay attention to the key in on the 
>  status purpose status index and Status credential properties 
>  these are the properties that are introduced by the status least 
>  2021 spec the first of these properties is the most self 
>  explanatory this is saying that.
> Kayode Ezike:   The issuer.
> Kayode Ezike:  Acting this type of status for this credential so 
>  for verifier when they want to verify the status of the 
>  credential this is the information that they will learn now the 
>  other two properties are more specific to sort of like are sort 
>  of the meat and potatoes of how this specification works and it 
>  relates to a new resource at the introduced so if you can just go 
>  forward just a couple steps very until you see a new object in 
>  the right.
> Kayode Ezike:  So this on the right is a status list 2021 
>  credential and the best way to think about this credential is it 
>  is the credential that manages the status of a batch of multiple 
>  credentials it's not philosophy that one more time but this is a 
>  credential that the issuer manages that tracks a batch of several 
>  different credentials and so what this means if you could just 
>  click one more time Carrie.
> Kayode Ezike:   So we so.
> Kayode Ezike:  This was the most important information on this 
>  country is this encoded list value right here and what it is in 
>  the simplest representation you can think of it as a sequence of 
>  characters let's let's say X's and O's that represents the the 
>  status of a credential at that position right and so if the value 
>  let's say the value of that was X it would mean that it is 
>  invalid respect to the status purpose so in other words it is 
>  revoked.
> Kayode Ezike:   Value is 0 it means that it is valid with.
> Kayode Ezike:  The words is not revoked and if we go back to the 
>  if you focus again on the left side you will see that the status 
>  list credential property that is just the way for verifiers to 
>  retrieve this credential on the right so it's a public link that 
>  they can access and the status index tells them which position in 
>  that encoded list is represented a represents this credential.
> Dmitri Zagidulin:  So I want to I want to pause here before we 
>  move on to the next slide I want to make sure everybody absorbs 
>  this so.
> Dmitri Zagidulin:  I want to make sure everybody understands what 
>  that encoded list is for right because at its at its simplest a 
>  credential status is binary for a given status purpose so like 
>  for a vacation rental is either revoked or it's not so the very 
>  first sort of naive implementations of replication status where 
>  to host a.
> Dmitri Zagidulin:   Some kind of object.
> Dmitri Zagidulin:  Either cover fabric credential or something 
>  else that's a revoked or not for each individual credential right 
>  easy so or even before that I think what open badges did is just 
>  publish a plane list of all revoked credential IDs on their 
>  website yeah so you get you get the credential you can go look up 
>  that list to see if it's revoked.
> Dmitri Zagidulin:  So very easy and we wouldn't be having this 
>  problem I rather this this conversation here except there's a 
>  couple of major major problems there one is privacy that if you 
>  publish a list of remote credentials well you can go see 
>  everybody whose credentials are revoked but an even more 
>  important one is what's known as the phone home problem it's one 
>  of the downsides of Hosting.
> Dmitri Zagidulin:   An individual status.
> Dmitri Zagidulin:  Nation for each credential is that whoever is 
>  hosting it covers running that website can track.
> Dmitri Zagidulin:  Requests can can track how many times and from 
>  where from what IP address is and what time stamp some verifier 
>  is is checking.
> Dmitri Zagidulin:  Revocation and that that is generally 
>  regarding the community has too much information that that's like 
>  that's not necessary that is a threat so then the reason the 
>  status list spec arose is that okay so if we don't if we don't 
>  want the host whoever's controlling the web whole of the web 
>  server to know each time each individual credential is looked up 
>  why don't we batch it why don't we rely on.
> Dmitri Zagidulin:   On herd immunity on herd privacy.
> Dmitri Zagidulin:  And batch a whole bunch of credentials at 
>  random.
> Dmitri Zagidulin:  And then the verifier can request this whole 
>  batch this this entire credential that has the encoded list that 
>  contains the status of 20 or 100 I forget how many it is.
> Dmitri Zagidulin:  Potential statuses and that way the request 
>  will be spread out over that entire cohort of credentials and 
>  that way we get her privacy I see David is on the queue.
> Kayode Ezike:  When you credential is revoked right so when your 
>  credential is revoked this this this French on the right will be 
>  updated the encoded list will be up to such that the bit at that 
>  position I 4 5 6 10 is now 1 or Annex discussed earlier.
> Kayode Ezike:  Yes yes that comes up I think some question time 
>  so that really I think bows down to the I think we leave that to 
>  the implementers of the their system because that really comes 
>  down to how often how what is it catching sort of system like do 
>  you check every day do you check every hour and the you'd have to 
>  give an essay to the individual using your system to let them 
>  know that this is just going to be valid.
> Kayode Ezike:  For a given day or for.
> Kayode Ezike:  Etcetera but I don't think that's something that 
>  this back tries to address or to solve.
> Kerri Lemoie:  Give me three.
> Dmitri Zagidulin:  If I may add to that so the cache control is 
>  left to the individual protocol meaning because this particular 
>  credential is held over http.
> Dmitri Zagidulin:  His publishing the status list can rely on 
>  HTTP cache control directives meaning each time the verifier 
>  sends an HTTP get for this status credential one of the headers 
>  in the response says only cash this for an hour meaning we're 
>  going to we're going to renew this thing every hour and that way 
>  the verifier knows how long to cash it right so we get that the 
>  caching mechanism for free with HTTP so we don't need to put in.
> Dmitri Zagidulin:   An explicit.
> Dmitri Zagidulin:  Potential now if we were if the URL of the 
>  status of protection was not HTTP was what some other was ipfs or 
>  some other method that doesn't come with cache-control metadata 
>  then you're right David we should include an expires field in the 
>  statuses credential.
> Kayode Ezike:  Okay thank you.
> Kerri Lemoie:  No more questions for now Kayode if you want to go 
>  ahead.
> Kayode Ezike:  Thank you for question is David and I'm great 
>  thanks for getting ahead of me Michelle's going to get into the 
>  herd privacy notion which is I think one of the main benefits of 
>  the specification which is just to reiterate when a verifier is 
>  interested in learning about the status of a credential they are 
>  only communicating to the issuer that they're interested in a 
>  subset of credentials that they manage not in a specific one so 
>  it keeps the issuer away from fine-grained details about how a 
>  specific.
> Kayode Ezike:  Potential is being used and if you just.
> Kayode Ezike:  Or just one.
> Kayode Ezike:  Thank you this is a digital representation just 
>  the visual Learners out there the green slots represent the valid 
>  credentials with respect to the purpose so I'm revoked and events 
>  Lots represent the invalid by evoked credentials.
> Kayode Ezike:  Next slide please.
> Kayode Ezike:  So we have a number of goals that we were 
>  considering what we designed our country stars infrastructure the 
>  main governing goal is that we wanted to simplify the deployment 
>  for issuers of the printer status infrastructure and we do this 
>  by leveraging third-party services for the storage and 
>  optimization of grantor status resources and the great thing 
>  about this is that we're kind of meeting is adjourned.
> Kayode Ezike:  Is worth where they are right.
> Kayode Ezike:  Allowing them to use a familiar hosting service 
>  and and also providing potential path to switch between services.
> Kayode Ezike:  So if we can step ahead just one slide.
> Kayode Ezike:  We'll get into the solution.
> Kayode Ezike:  So we again decided to use a static list 2021 
>  specification and feel free to consult the design doc at that 
>  link and the whole conclusion I think this is really where we 
>  innovate and we decided to use Source control services such as 
>  GitHub get lab and code Berg and we also support issuer hosted 
>  websites traditionally and so what this allows us to do again is 
>  we.
> Kayode Ezike:   It's a biscuit.
> Kayode Ezike:  Like that organizations already use these services 
>  to host their code and and also these Services often provide 
>  developers with oauth tokens that they can use to access apis of 
>  these services to update resources and so why don't we just use 
>  this these services to help issuers manage their current status 
>  lists and metadata associated with it.
> Kayode Ezike:  Any questions here.
> Kayode Ezike:  To reload it.
> Kerri Lemoie:  Any chance has an adversity.
> Dmitri Zagidulin:  Yeah if I could just add a couple more words 
>  so I want you I want you to everybody to picture the the very 
>  specific conversation the very specific problem the solves.
> Dmitri Zagidulin:  As okay we're issuing credential great can we 
>  add revocation sure why not what's a good spec what we've got 
>  this status list 2021 spec okay then engineering how hard of a 
>  lift is this to add this to our system how many hours before you 
>  can add revocation to to our issuer and that's where the real 
>  problem starts because hosting a file on a website that b is easy 
>  the really difficult part which is what makes okay.
> Dmitri Zagidulin:   It is work so.
> Dmitri Zagidulin:  Difficult Park is part is adding the user 
>  interface and specifically adding permissions to who is allowed 
>  to edit who's allowed to revoke the the credential right so the 
>  spec gives us the data model gives us the protocol the mechanisms 
>  very easy what's really not easy and what shoots up the 
>  implementation time in too many weeks not to mention like really 
>  hard requirement Gathering is.
> Dmitri Zagidulin:  Delegation the chain of command of.
> <xander> I don’t think you can set http cache control headers on 
>  GitHub-hosted files.
> Dmitri Zagidulin:  Okay so you know how do we model the trust and 
>  permission hierarchy of a particular University down to the 
>  registrar and then how does it register our delegate individual 
>  officers to be able to hit the button to flip that bit for for 
>  the file to be updated so the the the main Innovation here is the 
>  realization that.
> Dmitri Zagidulin:  Because permission control and and login 
>  systems and graphical user interface is the hardest part about 
>  this can we Outsource it to somewhere else and the realization 
>  was made oh get hosting organizations like GitHub and get lab and 
>  code Berg already provide all of that they provide the permission 
>  systems the login systems they produce institutions are familiar 
>  with setting up Gap GitHub organizations all of that.
> Dmitri Zagidulin:   Stuff is worked out for us why don't we reuse 
>  it.
> Dmitri Zagidulin:  Geico I'll go ahead Gary.
> Kerri Lemoie:  One question I'm asking these are all my be 
>  hosting Services could someone just use get for this on its own 
>  without using GitHub get lab and or code Berg.
> Dmitri Zagidulin:  So the the question with get is always which 
>  protocol will get use right because get has SSH it has HTTP and 
>  it has a number of other protocols words hosted so some server on 
>  the cloud has to be running it gets server.
> Dmitri Zagidulin:  So can you use get a loan sort of it requires 
>  a get server in a good client.
> Kerri Lemoie:  I think so Xander.
> Xander: I guess I have to have you spoken before so I'll just 
>  quickly I'm the security lead for the pocket team just wanted to 
>  follow up on David's question real quick so the idea there was 
>  that you were going to rely on HTTP cache control headers to set 
>  the expiry time for a different certificate right but you're also 
>  talking about using services like GitHub to do the hosting I 
>  don't believe that you can set the expired header value if you're 
>  using a service like that.
> Xander: So that they.
> Dmitri Zagidulin:  And that's that's a very good point.
> Xander: You may need to put that you may need to put the time 
>  value on the file done.
> Kayode Ezike:  Right yeah that's a good description that we can 
>  we start to use these third party dresses I think there's 
>  different levels of access that they are each provide and I'm 
>  sure maybe like if you have a paid account that makes a 
>  difference as to how much you could control but I think largely 
>  you make a good point and.
> Xander: I don't think so I got used the Enterprise version before 
>  I think basically if you're using GitHub to host files like get 
>  on pages or whatever they really handle everything and you don't 
>  get access to that level.
> Xander: It may be worth looking into.
> <dmitri_zagidulin> that seems like a good motivation to clarify 
>  (in the status list spec) the recommendation of always having 
>  expiration timestamps
> Kerri Lemoie:  Yeah time in the queue tell jump in real quick 
>  that's an excellent point and I know you haven't actually I'm let 
>  you get into like how you doing this so maybe I'm jumping the gun 
>  here Harry I'm sorry if I am but um are you considering using 
>  just as part of this because just get versioned which is 
>  something to consider okay that's the comment there but yeah.
> Kayode Ezike:  This is this is for like the accident the location 
>  of us for example.
> Kayode Ezike:  Yes it's something like this came up actually you 
>  kind of hear made great recommendation at some point about 
>  something like that where you would because get help for example 
>  has a way for you to use a link that it's post using GitHub Pages 
>  you can also use a link to the file directly that doesn't rely on 
>  GitHub pages but I would imagine that the.
> Kayode Ezike:  Might be a rat.
> Kerri Lemoie: 
>  https://docs.github.com/en/get-started/writing-on-github/editing-and-sharing-content-with-gists/creating-gists
> Kayode Ezike:  He got distracted the same way in that case as 
>  well in terms of controlling what the cache mechanism is like so 
>  yeah that would be something to explore but the good thing I'll 
>  show you a little bit later how some of these things can be 
>  customized for different status clients of different services and 
>  there's a way to add flexibility for for that.
> <kerri_lemoie> Link above to info about gists.
> Kayode Ezike:   Again to that little bit.
> Kayode Ezike:  There are no other questions can move on to the 
>  next slide.
> Kayode Ezike:  Great so this is a sort of a visual representation 
>  of the architecture that we have issue admin who presumably would 
>  already have the permissions that they need inside of a example 
>  with GitHub organization so they already have access to create 
>  read and write data to Repose in the organization and if you 
>  click ahead.
> Kayode Ezike:   So this.
> Kayode Ezike:  This actor would be able to hit the credential 
>  status and point of a VC API instance which allows them to update 
>  the status EG revoke a credential.
> Kayode Ezike:  More step and then ultimately that enables them to 
>  create read and update data in these different services.
> Kayode Ezike:  So before we move on to a demo wanted to show a 
>  little bit of code and so this is the main sort of class that we 
>  Implement to with this code and so there's this tells subclasses 
>  different methods that they need to implement for them to be 
>  valid printer status clients and so for example you need to 
>  provide a base URL that can be used to reference the printer 
>  status information you need to.
> Kayode Ezike:  Boo the client too.
> Kayode Ezike:  That is people from from the code need to enable 
>  them to read and write status list and log data in config data so 
>  as long as the subclasses can Implement these abstract methods 
>  and they're valid client and so if you can step forward this one 
>  step please carry will see that in this example GitHub a cultural 
>  status client for their get credentialed status URL you notice 
>  that there is a GitHub that I owe their meaning that.
> Kayode Ezike:   Using GitHub pages but this is also the place 
>  where.
> Kayode Ezike:  I could use a distance that or URL that points 
>  directly to the file and there's also we also are using the 
>  octave kit Library a popular GitHub SDK to update and read 
>  information from the repo which obviously would look different 
>  for different services.
> Kayode Ezike:  Any questions about this.
> Kerri Lemoie:  I think we're good once you go ahead.
> Kayode Ezike:  So we're just going to get into a demo now just 
>  step forward one and taxes and maybe that's going to come to my 
>  email or something.
> Kayode Ezike:  Probably going to email about that 17.
> Dmitri Zagidulin:  Karen thank you have to hit request again on 
>  the subsequent page.
> Kerri Lemoie:  Let me know when it's all set Katie I'll refresh 
>  it.
> Kayode Ezike:  Yeah it is I just shared it.
> Kayode Ezike:  The volume on.
> Kerri Lemoie:  Were you able to hear the volume when I get out.
> Kerri Lemoie:  Dimitri do you know how to make that work.
> Kerri Lemoie:  Yep that's share audio.
> Dmitri Zagidulin:  In the settings so if you go to the 3 3. Menu 
>  at the bottom settings there is yeah share out the others share 
>  audio checkbox.
> Kerri Lemoie:  Thanks your patience everybody but I'm getting 
>  there.
> Kerri Lemoie:  All right let me try that again it doesn't work 
>  Katie you could just walk us through it.
> <kerri_lemoie> Can you hear?
> Dmitri Zagidulin:  Now that doesn't seem to be coming through so 
>  we can encourage people to watch since we we pasted the link to 
>  the slide deck everybody can watch the video on the demo on their 
>  own.
> Dmitri Zagidulin:  But you can you can describe what when it's 
>  done.
> Kerri Lemoie:  We're going to start over and you can just sort of 
>  walk us through what's happening.
> Kayode Ezike:  Sure I mean it's a recorded a demo but effectively 
>  what we were demonstrating is requesting a credential from our 
>  version of V Capi importing that into DC learner wallet and then 
>  from there we kind of show you throughout the whole process 
>  called the repo is updating and so the GitHub repos I'm using a 
>  demo and and then we revoke the credential.
> Kayode Ezike:  Again from the VCA Pi you show you the.
> Kayode Ezike:  Two then rebuild me also show you.
> Kayode Ezike:  That in the wallet and now shows that the 
>  credential is revoked.
> Kayode Ezike:  Again it's you should be able to view that video 
>  in the presentation but that's effectively what's going on.
> Kayode Ezike:  So that was the last of it actually so if there's 
>  any questions or feedback I'm happy to take in you David go 
>  ahead.
> Kayode Ezike:  Oh so so the so the credential continues to have 
>  the same the credential never changes per se it's the information 
>  at at least the credential that the holder controls is the 
>  credential that the status could ensure that managers know that 
>  has been coded list value that manages multiple credentials is 
>  that one that will be shuffled and change as a credentials are 
>  revoked suspended.
> Dmitri Zagidulin:  Can I can I jump in here as well.
> <nate_otto_(another_device)> Nah. credential.id is optional in VC 
>  Data model anyway. Reissuing the same one doesn't violate it.
> Dmitri Zagidulin:  So David you're proposing or what you're 
>  saying is essentially if we required verifiable credentials to be 
>  content addressable so that every time the content of a 
>  verifiable potential change the ID has to change but the v-spec 
>  does not actually have that requirement you are you are in this 
>  very frequently done allowed to publish credentials with the same 
>  ID while their contents changes for example if you reuse.
> Dmitri Zagidulin:   Issue it and then.
> Dmitri Zagidulin:  Timestamp or a different signature so that is 
>  that is not a requirement in this back.
> Kerri Lemoie:  I'm to add to that question sort of who may be 
>  briefly talk about the difference between credential status list 
>  and also credential refresh carrier to meet you would you mind 
>  explaining that are both of you.
> Dmitri Zagidulin:  Shark a candidate doing on go ahead.
> Kayode Ezike:  How can I just ask this versus credential 
>  refreshing scent.
> Kerri Lemoie: CredentialRefresh: 
>  https://w3c-ccg.github.io/vc-refresh-2021/
> Kayode Ezike:  Yeah so my understanding generally is that 
>  subconscious has what we discussed here which is that the issuer 
>  manages a publicly accessible resource that allows verifiers to 
>  check the status of the repo or rather the other credential for 
>  the route for the refresh service my understanding is that 
>  whenever if that is not provided the issuer exposes an endpoint 
>  that.
> Kayode Ezike:  Allows it that allows.
> Kayode Ezike:  Verifiers to to refresh the credentials whenever 
>  it has rather just to be able to refresh it whenever I would ever 
>  Cadence in the seems reasonable that's generally how I understand 
>  it but I'm happy to allow for any other discussion on that as 
>  well.
> Dmitri Zagidulin:  Yeah so the refresh spec.
> Dmitri Zagidulin:  Complementary to the expiration mechanism so I 
>  have my driver's license is good for four years what happens when 
>  it expires up to four years the refresh spec basically describes 
>  the credential version of oh I have to take the old one into the 
>  drive the Bureau of Motor Vehicles on the issue me a new one so.
> Dmitri Zagidulin:  Both the status and the refresh exist 
>  alongside each other and in fact some of the some of our example 
>  specs have both or if you look at examples in open badges version 
>  3 spec a lot of the verify the credentials their have both a 
>  status list section and a refresh section.
> Kerri Lemoie:  Thanks anybody else have any questions about this.
> Kerri Lemoie:  Or any other points they'd like to make about the 
>  status and how this works.
> Kerri Lemoie:  I bet I have a question for all of you is there 
>  anybody here who has implemented credential status and like to 
>  tell us about how they're using it and how they've implemented 
>  it.
> Kerri Lemoie:  Okay Patti really appreciated that I see so many Q 
>  Jonathan the floor.
> Dmitri Zagidulin:  Anthony might be muted.
> Kerri Lemoie:  Yeah you can't hear your Jonathan if you're 
>  speaking.
> Jonathan_Bethune: Okay is that.
> Dmitri Zagidulin:  If you have like an additional Hardware mute 
>  on your mic maybe yes better yeah.
> Kerri Lemoie:  Yep they are to thank you.
> <kerri_lemoie> Link to presentation: 
>  https://docs.google.com/presentation/d/1UYFcVLYaz8jhmmYM8l43cBg-mtINlxlFMESCbYupjwo/edit#slide=id.g143e60161fc_0_35
> Jonathan_Bethune: Better I was using one earpod and it's the one 
>  that's dead so okay sorry about that so real quick by way of 
>  introduction I think I've spoken much my name is Jonathan and the 
>  engineering manager for pocket colleague of Xander who spoke a 
>  little while ago just real quick I was wondering if there was a 
>  way to get a link to the presentation and actually had a lot of 
>  discussion internally about the X and we're just wanted to look 
>  into a little.
> Kayode Ezike:  Sure yeah definitely can share that.
> <phil_l_(p1)> Nate's Comment:
> Jonathan_Bethune: There we go oh that's fast okay let me well I 
>  actually didn't get the click on it we just grabbed it's in the 
>  chat right there we go.
> Kerri Lemoie:  Yeah it's going up a little bit the chat moves 
>  pretty fast because of the transcription.
> <phil_l_(p1)> Good point Nate.
> Jonathan_Bethune: Yeah got it alright thank you.
> Kerri Lemoie:  Okay if no one has anything else okay did you have 
>  anything else you would like to add.
> Kayode Ezike:  That was that was everything I just thank you all 
>  for your time and yeah we're continuing to refine this this work 
>  and we primarily are supporting GitHub and get lab the moment and 
>  working on some of the other services as well so just happy to 
>  have this opportunity and thank you all again for there for 
>  discussion.
> Kerri Lemoie:  Yeah thank you very much.
> Dmitri Zagidulin:  And if I can add so if.
> Dmitri Zagidulin:  If anybody has questions about this work 
>  please send an email to the VC edu mailing list or if you're free 
>  to contact KO myself directly the code code is free it references 
>  an open spec we encourage everybody to join in the conversation.
> Kerri Lemoie:  Hip and actually seen my hush you have killed 
>  yourself up.
> https://imsglobal.github.io/openbadges-specification/ob_v3p0.html
> Mahesh_Balan_-_pocketcred.com: Yes I had a question slightly 
>  unrelated here but more on the open badges 3.0 spec I've been 
>  trying to get hold of it and I seem to be pay bald by this IMS 
>  Global if somebody has that and can make the open badges 30 spec 
>  which is publicly visible I would appreciate it thank you.
> https://www.imsglobal.org/spec/ob/latest/main/
> Kerri Lemoie:  I can grab that link for you give me one second 
>  here tell Ted you're in the queue your the floor.
> <kerri_lemoie> Open Badges 3.0 spec: 
>  https://imsglobal.github.io/openbadges-specification/ob_v3p0.html
> TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): I just 
>  following up on responding to David's question about the unique 
>  identifier for each credential in the same way that a URI is 
>  expected to be stable that is that is the stability that we mean 
>  for the ID in these verifiable credentials URI is always meant to 
>  name the same thing which doesn't mean that it the thing it names 
>  stays the same forever.
> TallTed_//_Ted_Thibodeau_(he/him)_(OpenLinkSw.com): Much like 
>  your name always means you, but your contents don't stay the same 
>  forever.
> <dmitri_zagidulin> thx Ted!
> Kerri Lemoie:  Hey Alec it is empty and listen no one has 
>  anything else we have a few minutes left so we could end the call 
>  here give another Med see if anybody has anything otherwise I 
>  hope you all have a great week and I will see you next week.
> <phil_l_(p1)> Nice preso!
> Kayode Ezike:  Cheers thank you.
> Dmitri Zagidulin:  Thanks KO day thanks everyone.
> <elizabeth_miller> Thank you!
> Kerri Lemoie:  Thank you headed.
> <john_kuo> Thanks, great discussion
> 
>