Verifiable Claims Telecon Minutes for 2016-05-10

Thanks to Nate Otto for scribing this week! The minutes
for this week's Verifiable Claims telecon are now available:

http://w3c.github.io/vctf/meetings/2016-05-10/

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

----------------------------------------------------------------
Verifiable Claims Telecon Minutes for 2016-05-10

Agenda:
  https://lists.w3.org/Archives/Public/public-webpayments-ig/2016May/0009.html
Topics:
  1. Introduction to Alok from Cambridge Blockchain
  2. Review of Survey Status
  3. Work on Existing Docs
  4. Verifiable Claims Data Model Specification
  5. Survey Results/Overview
  6. One page W3C AC Rep Summary
  7. Architecture Summary
  8. Implementers Summary
Organizer:
  Manu Sporny
Scribe:
  Nate Otto
Present:
  Nate Otto, Manu Sporny, Alok Bhargava, Jason Weaver, Dan Burnett, 
  Eric Korb, Richard Varn, Dave Longley, Brian Sletten, David 
  Ezell, Stuart Sutton, Rob Trainer, David I. Lehn, Colleen 
  Kennedy, Rebecca Simmons, Matt Stone
Audio:
  http://w3c.github.io/vctf/meetings/2016-05-10/audio.ogg

Nate Otto is scribing.
Manu Sporny:  Introduction to new participant, then go through 
  deliverables and their status.
Manu Sporny:  As some of you may have heard, we're currently 
  talking with the credentials transparency initiative about some 
  naming conflicts, but we'll delay putting that on the agenda 
  until next week.
Manu Sporny:  Any other items for the agenda to discuss today?

Topic: Introduction to Alok from Cambridge Blockchain

Alok Bhargava:  I work for a company called Cambridge Blockchain; 
  young company in the identity space. Our CTO has a lot of 
  experience in blockchain as well as identity. He's not able to 
  participate. I'm looking to participate in this primarily to 
  learn about the push to standardize how identity is used on the 
  Internet. I have a very limited background in identity and 
  crypto. I have about 12 yrs experience 
  designing/building/marketing telecom systems. Role at Cambridge 
  Blockchain is VP of Product.
Manu Sporny:  Welcome to the group, Alok

Topic: Review of Survey Status

Manu Sporny: 
  http://manu.sporny.org/tmp/vctf/VerifiableClaimsResponses-2016-05-06.png
Manu Sporny:  Next up on the agenda is a review of the survey 
  status. Here is an image capture of the survey results.
Manu Sporny:  We'll clean this up and put it into an HTML page at 
  some point.
Manu Sporny:  We have a very healthy number of responses at this 
  point. We asked about 83 people to respond. That number of people 
  responding is a good sign. We have responses from the financial 
  sector, the education sector, healthcare, government, national 
  institutions, NGOs, entertainment industry, mobile vendors. A 
  very healthy set of people across multiple market vertcals.
Manu Sporny:  I'm going to start going down from the top.
Manu Sporny:  Very good support for the problem statement.
Manu Sporny:  Up to around 96% positively responding that the 
  goals are good to pursue.
Manu Sporny:  This is really good, because it means we have the 
  language on the problem we're trying to address in the goals and 
  scope of work down really well.
Manu Sporny:  We asked if people's verifiable claims problems 
  would be addressed if the use cases proposed were expressed.
Manu Sporny:  We really limited, narrowed the use cases to ensure 
  we had a narrow scope. Most people thought that more needed to be 
  done beyond these to fully meet their needs.
Manu Sporny:  We limited use cases because large scopes make 
  people nervous, that you won't be able to achieve, or they would 
  have too broad reaching grant of patent rights.
Manu Sporny:  We asked if people would join and participate, and 
  44% of respondants said that they would join and partcipate in 
  the work.
Manu Sporny:  34% Put themselves as other in this space. They 
  would mostly like to participate but don't know if they have the 
  bandwidth right now.
Manu Sporny:  We definitely have 16, and potentially 4-5 more if 
  they find the time.
Manu Sporny:  And typically as these things get officially 
  started more orgs join up.
Manu Sporny:  We did get some feedback that they would like to 
  see more incubation of the work.
Manu Sporny:  Some because they didn't see a specification, they 
  weren't sure about the architecture...
Manu Sporny:  We did speak with the Web Payments IG. We're 
  getting close to the question of whether the WPIG would support 
  the work going forward. One of the big questions is if the 
  specification is created, who is going to implement it into their 
  products.
Manu Sporny:  We're going to go around calling people to see if 
  we can put them down as implementers.
Manu Sporny:  We're doing good with the survey. We found some 
  other items that we have to create to make sure the vote goes 
  well.
Jason Weaver: Are more responses desired?
Manu Sporny:  Any questions on survey status or where we are with 
  the survey?
Manu Sporny:  Are more responses desired? Yeah. If you haven't 
  taken the time to fill out the survey, please do it. If you know 
  an organization who really should have filled out the survey, get 
  them to fill it out as soon as possible.
Manu Sporny:  We did a soft close on the survey last friday 
  (started reporting results), but still accepting responses.

Topic: Work on Existing Docs

Manu Sporny:  Went into the work with Charter, Use Cases, and 
  FAQ. Now we're going to try to rev those documents in response to 
  survey. Not much work is needed, we'll try to get it done in the 
  next couple weeks.
Manu Sporny:  Any questions on Charter, Use Cases, and FAQ?

Topic: Verifiable Claims Data Model Specification

Manu Sporny:  Dan Burnett, I'm going to hope that you give a 
  quick overview of the Verifiable Claims data model.
Manu Sporny:  When we got the survey, we got a number of people 
  saying "that's great that you're going to do the work", but if we 
  don't see a proposed technical solution that we have moderate 
  buy-in on, there's nothing to standardize, and we can't see if we 
  support it.
Dan Burnett: Very drafty spec proposal:  
  http://opencreds.org/specs/source/claims-data-model/
Manu Sporny:  This is kind of a new dynamic at w3c. You used to 
  not need this (often counterproductive), but some big companies 
  are pushing for this
Dan Burnett:  There is a draft document, includes some reSpec 
  errors to correct; includes some about identity.
Dan Burnett:  We have a data model for how verifiable claims 
  work, but there are different languages that different users of 
  this technology would like to use to define the syntax for this 
  data mode.
Dan Burnett:  Challenge: to show how data model can be 
  represented in the three syntaxes people want to use
Dan Burnett:  Using the terms "claim" or "verifiable claim" to 
  mean what we used to call "credentials".
Dan Burnett:  One of the things that wes important would be to 
  distinguish between claims, and "verifiable claims". The latter 
  has a signature.
Dan Burnett:  In the data model, you'll see a section for claims, 
  and for verifiable claims.
Dan Burnett:  You'll see examples of how to express these
Dan Burnett:  One thing I wanted to write up, but no set location 
  to put it: how to write about identities. We need to have an 
  ability when we're talking about ids in claims, the ids are 
  suppose to reference identity @ids.
Dan Burnett:  Until we figure out how to describe that more 
  generally, it's here.
Dan Burnett:  A lot of text came from the original identities and 
  credentials document is here in appendices, pulling from as we 
  try to expand and write introduction/abstract, etc.
Dan Burnett:  We have a bunch of reSpec errors. Shane and I are 
  fiddling with them for now.
Dan Burnett:  Obviously drafty. If there are problems with the 
  webIDL, don't worry about that yet. If there's something 
  obviously wrong with the rest, let us know.
Eric Korb: Korb 908
Richard Varn:  Did you start with the Open Badge Initiative for 
  what you'd expect to find in a spec, or is that something that 
  needs to be brought in still?
Richard Varn:  I'm glad you're representing identity. It's 
  tricky, but it has to be done.
Dan Burnett: This started from:  
  http://opencreds.org/specs/source/identity-credentials/#expressing-identities-and-credentials
Dan Burnett:  As far as what I started from, posting a link...
Dan Burnett:  My history doesn't go far enough back to know where 
  all that came from. If I've put in something inappropriate or 
  left out something appropriate, let me know.
Dan Burnett:  My aim was to show the data model, separate from a 
  syntax, and then show implementations in syntaxes.
Dan Burnett:  We need to avoid problems with critics right away 
  who say something like "It needs to only be represented in 
  webIDL"
Manu Sporny:  Richard, the specification we have right now in 
  this group, we believe can fit the Open Badges stuff in in its 
  entirety. Open Badge is just another claim as far as the system 
  is concerned. Whatever that claim is, we can just stick it in the 
  claim and see how that works. We haven't discussed this a lot in 
  this group recently. Would be good to hear from Nate and maybe 
  Kerri on where OBI fits in.
Manu Sporny:  We should be good to go as far as open badges go.
Manu Sporny:  Let me jump back. burn, thanks for putting this 
  into this new form. Spot on, I think it's great to explain what 
  the data model is in prose and then talk about how you express it 
  in JSON-LD, WebIDl, etc. Will work for a lot of communities. It 
  increases our ability to have this stuff reused.
Manu Sporny:  For example, there has been a lot of interest in 
  blockchain communities to inject these credentials in sidechains 
  and other structures.
Manu Sporny:  All in all, I think the spec is a step in the right 
  direction, Dan.
Manu Sporny:  Two comments; One of them is the removal of the 
  word credential. We have decision to make on that if a credential 
  is a set of verifiable claims, for instance. The word identity as 
  well may be the other most controversial.
Dave Longley: +1 For using credential but burying it in the 
  document as just a container for claims
Dan Burnett:  Actually debated for quite a while whether I wanted 
  to remove the word credential. If we started with the word 
  credential right at the top, I thought we'd have trouble based on 
  the conversation at the last month. But I wasn't trying to change 
  semantics (for what I'm calling "claims data set"), so if we need 
  to go back, we can.
Brian Sletten:  I just want to get a sense because you said this 
  is a new thing being asked by some of the bigger members. Seems 
  like fluctuation on a sine wave -- we can't proceed without 
  technical details... we can't define technical details at this 
  stage.
Manu Sporny:  There is disagreement. Some members think that this 
  is not appropriate to do at this stage. Others think it should 
  produced right now.
Manu Sporny:  The reason we're producing this spec right now is 
  to answer the people who say "you don't have a technical 
  proposal".
Dan Burnett: Dlongley, the original doc had credential as a 
  container for claims, which means the main section titles would 
  talk about "credentials".  I'd want to hear more about how you 
  think we could bury it while still having it be the outer 
  container name
Manu Sporny:  For the organizations who say "you shouldn't enter 
  with a technical proposal", we say "we were asked to do a 
  technical proposal. None of this is set in stone; it's the job of 
  a working group to decide on the technical proposal". We're not 
  trying to make each side 100% happy but meet the minimum bar that 
  each side has.
Manu Sporny:  This is one of the things we think we need to stave 
  off formal objection from some large organizations.
Manu Sporny:  Does that make sense?
Brian Sletten:  "Ish"
Jason Weaver:  Richard asked about the model working with Open 
  Badges. My question stems from that. A lot of what we call these 
  things today are being exchanged in different regional standards. 
  How to PESC, __ align with the architecture?
Dan Burnett:  I think by "bury" all we need to do is make the 
  main subject about "claims" ... start the document by talking 
  about what claims are and how this spec will detail how to model 
  them. then we can get into how a credential is a "set of claims" 
  ... i just think we need to avoid having it front and center is 
  all. (i don't think we're too far off from that now) [scribe 
  assist by Dave Longley]
Manu Sporny:  We are aware of PESC. We're trying to create a 
  generalized way of expressing claims. Anything those other 
  standards can express should be able to be expressed in this data 
  model. We do want strong semantics to go along with these claims. 
  It will probably be up to those standard-setting bodies to decide 
  how they want to express their claims in this data model. We have 
  not seen a case where a standard couldn't use this model to 
  express claims.
Jason Weaver: EMREX is the EU data standard
Dan Burnett: Dlongley, would love to talk with you offline after 
  the call.  I will email you.
Manu Sporny:  Key point is to make those groups understand that 
  we can build a way for our model to map to their data model. For 
  example, working with Credential Transparency Initiative to see 
  how we can express their model in our data model. Haven't had to 
  change the model. Just a matter of figuring out what the 
  vocabulary is.
Nate Otto:  From the open badges perspective, we expect to be 
  able to express the open badges data model via a verifiable 
  claim. Given that it builds off of other specs - we'd specify how 
   you do an open badge using verifiable claims. [scribe assist by 
  Manu Sporny]
Dave Longley: +1 To nate
Manu Sporny:  Until we have ratified specification, it'll be a 
  constant discussion on how best to express this stuff, but as 
  NateOtto says, there's no reason why we can't express many types 
  of data in this model. If you find a case where you can't, this 
  means we screwed up and it needs to be fixed.
Manu Sporny:  If there's no more questions, burn hopefully you 
  have time over the coming month to keep whittling the spec down 
  to something we want to include in the package to the w3c for 
  voting.

Topic: Survey Results/Overview

Manu Sporny:  Another document we need to generate is a summary 
  of the survey results.
Manu Sporny:  It's pretty compelling to see all these 
  organizations agreeing with the work. Need a 1-2 pager overview 
  of that.

Topic: One page W3C AC Rep Summary

Manu Sporny:  Another thing ACs asked for was a 1 pager specific 
  to advisory committee members. These people are usually asked to 
  vote on many things per month, and they're usually very busy. 
  Throwing them 20 pages will usually result in them not reading it 
  and not voting on it.
Manu Sporny:  One of the AC reps asked us to come up with a good 
  1 pager overview to help them on this. This work has not been 
  started yet, needs to be done.
Manu Sporny:  There were a list of questions that person included 
  in their response, we'll go over that. FAQ: Is there a clear 
  problem statement? Do you have buy-in? Is the scope narrow 
  enough? is there a spec? Who's going to implement?
Manu Sporny:  Those are the general things AC Reps want answered. 
  (also: Why would this initiative need a standard in the space?)

Topic: Architecture Summary

Manu Sporny:  Once we get all those answers onto a piece of 
  paper, that'll be the 1 pager opener.
Manu Sporny: 
  https://github.com/WebOfTrustInfo/ID2020DesignWorkshop/blob/master/topics-and-advance-readings/a-self-sovereign-identity-architecture.pdf
Manu Sporny:  Next up is a summary of the architecture that we're 
  building.
Manu Sporny:  There's a diagram on the first page of this pdf 
  where we're talking about issuers, holders, inspectors, 
  repositories.
Manu Sporny:  We've started thinking about that architecture 
  through this UN rebooting trust workshop paper draft. We're 
  probably not going to include things like decentralized ledgers, 
  decentralized hash tables. We don't want to imply that that's 
  what we're working on in phase 1. Clearly all of us understand 
  that blockchains & DHTs will play a part at some point in the 
  future of this work.

Topic: Implementers Summary

Manu Sporny:  Next document we're going to work on is this 
  implementers summary. We need to know who is going to implement 
  this work. If it's only 2-3, clearly we're not going to be able 
  to do this work. We need a bunch of orgs to say that "we will put 
  this into production"
Manu Sporny:  We need to gather a list of those organizations.
Manu Sporny:  That's all the items that came out of the survey 
  requests.
Manu Sporny:  Charter, Use Cases, FAQ, Data Model, Architecture 
  Summary, Implementers Summary is what makes up the pack we'll 
  distribute for a vote. (did I miss one?)
Manu Sporny: +Survey results overview
David Ezell:  July 1 Web Payments F2F has Verifiable Claims on 
  the agenda. Was going to send manu a note to coordinate
Manu Sporny:  There's a question of whether the Web Payments IG 
  will support this work. We'll have these documents ready to go in 
  a month and a half before that face to face. That'll go up for an 
  advisory committee vote. Do people feel this timeline is too 
  long?
Richard Varn:  I don't object to the timeline for this 
  standards-based work. I just see us like a squirrel in a cage 
  going and going forever. If this continues to stretch out 
  forever, do we have a plan B? We could have done a shorter 
  timeline approach and be already out in the world for 
  implementation
Manu Sporny:  I'm becoming more and more convinced that this will 
  go forward at the W3C
David Ezell:  The timeline is kind of artificial, mainly because 
  we have these face to face schedules. If the f2f worked two weeks 
  from now, we'd be doing it then.
David Ezell:  I haven't heard a lot of concern raised at the IG 
  level
David Ezell:  It's just a process to get people to vet their 
  concerns. I think this architecture document will help a lot. I 
  also think there is a related develoment at w3c that I want to 
  mention -- the browser extensions community group. It may turn 
  out that browser extensions are a way to get all this stuff 
  implemented. Which gets us out of the way of requiring a browser 
  vendor to write the code.
Manu Sporny:  To be clear phase 1 and phase 2 don't require 
  browser vendor support. Clearly we want their support at one 
  point, but we don't want to put two mega-companies in the way of 
  deploying this ecosystem that we want to see.
David Ezell:  The main point is that the IG is not intentionally 
  slowing this work down, it's just our schedule of face to faces 
  that sets times we can talk
Manu Sporny:  Either the IG's going to get behind this and 
  support it or say No. At that point, either we go to Plan B or we 
  go straight to the general W3C for a vote.
Manu Sporny:  By July 15 at the very latest July 30 we can start 
  votes. By August we'll know whether we have a WG or not, in time 
  for TPAC in portugal.
Manu Sporny:  That's it for the call today. We'll meet again next 
  week to go over some of the concerns with the Credential 
  Transparency Inititative.
Manu Sporny:  Thanks everybody for joining. Chat again next 
  Tuesday at the same time.

Received on Tuesday, 10 May 2016 16:25:47 UTC