The Verifiable Claims Task Force

A Task Force of the Web Payments Interest Group

Verifiable Claims Telecon

Minutes for 2016-12-20

Agenda
https://lists.w3.org/Archives/Public/public-webpayments-ig/2016Dec/0027.html
Topics
  1. Introduction to Pari, from Intuit
  2. Agenda Review
  3. Verifiable Claims WG Vote
  4. Discussion on Github
  5. Intro to Github Pull Requests
  6. Use of WebIDL in the Spec
Organizer
Dan Burnett
Scribe
Matt Stone
Present
Dan Burnett, Matt Stone, Pari Lingampally, Manu Sporny, Richard Varn, Shane McCarron, Nate Otto, Jonathan Holt, John Tibbetts, Dave Longley, Nathan George, Derek Callaway, Gregg Kellogg, David I. Lehn, Bill DeLorenzo, David Ezell, Joe Andrieu, Matthew Larson
Audio Log
Joining
Matt Stone is scribing.

Topic: Introduction to Pari, from Intuit

Pari Lingampally: I'm an engineer from Intuit. We are currently working on a project w/ blockchains and identity. mostly here to listen and learn.

Topic: Agenda Review

Dan Burnett: Any changes to the Agenda?
No changes requested.

Topic: Verifiable Claims WG Vote

Manu Sporny: Above threshold and have some positive votes. need some larger companies to weigh in,
Richard Varn: When does the vote close?
Manu Sporny: W3C Member-only vote link for Verifiable Claims: https://www.w3.org/2002/09/wbs/33280/VCWG/
Manu Sporny: Most of the responders are smaller companies, but work has generated some new w3c members.
Manu Sporny: Looks good so far
Manu Sporny: Need to respond to MSFT and Mozilla on the AC Forum
Manu Sporny: Of the 4 that initially pushed back, 2 have declared they won't block. working w/ the other 2

Topic: Discussion on Github

Manu Sporny: Github issue tracker for spec: https://github.com/opencreds/vc-data-model/issues
Dan Burnett: Chairs would like the group to have discussions on GitHub and make sure time in our weekly calls is used for topics that require live discussion or decision
Dan Burnett: Review and comment on issues that are logged at GitHub
Dan Burnett: Will use call to discuss issues that seem to be "stuck"
Dan Burnett: Will drive to have proposals to review in the call, so we have a clear decision to debate.
Nate Otto: This GitHub repo looks good.
Manu Sporny: Pari - this group is about to become a w3c member group. feel free to participate via issue tracker
Manu Sporny: Would love to have Intuit's input and collaboration
Jonathan Holt: Yes, Intuit is a member
Jonathan Holt: What's the process to consolidate and prioritize
Dan Burnett: As w3c we must attempt to address every issue. must be careful w/ consolidation so we don't miss any issues
Dan Burnett: Prioritization is based on getting discussion. editorial issue will start getting addressed naturally
Jonathan Holt: Seems like several that overlap and need discussion
Dan Burnett: We will be careful about combining issues as "duplicates"
Pari Lingampally: Thanks for the link manu, I will forward this to the appropriate people and make progress on our vote.Btw just looked, Intuit is a W3C member.
Matt Stone: We're adding a more coarse-grained prioritization. We may add another tag called "requires discussion", to make sure those items are discussed on these calls. [scribe assist by Manu Sporny]
Matt Stone: We're not looking for ordinal prioritization, more coarse grained stuff. [scribe assist by Manu Sporny]

Topic: Intro to Github Pull Requests

Dan Burnett: Do we need to have this discussion today?
Dan Burnett: Anyone who would particularly like me to walk through this process?
Manu Sporny: Some may be shy?
Dan Burnett: Anyone who doesn't want us to go through it
Dan Burnett: We have several participants that aren't developers. we'll go quickly if there are no questions
Nate Otto: I understand pull requests in GitHub. We've implemented this procedure for Open Badges and have a small amount of participation via pull request. We've even had a couple non-developers level up and learn how to use PRs.
Matt Stone: Stop scribing the "how to use gitHub" presentation
Manu Sporny: Maybe too deep. go fast :)
Manu Sporny: 3 Ways to be heard - 1) fork/edit/PR in gitHub or 2) comment on issues directly 3) send an email.
Manu Sporny: Option 2 is the simplest and best if you don't wan to be an editor
John Tibbetts: Regarding issues v. pull requests. I've been using issues until have better understanding of the group consensus? if an issue become concrete, will it be bounced back to " us" to make a PR?
Dan Burnett: Depends...
Dave Longley: You can always submit a PR on your own and say it's a proposal for people to take a look at
Dan Burnett: If there's clear concrete agreement, chairs may ask you to make a PR, if you don't know how or resist, chairs will find a way
Dave Longley: It doesn't mean the group will accept it, it's a conversation starter
Dan Burnett: Chairs may ask editors to make a change and ask for comments from the group

Topic: Use of WebIDL in the Spec

Manu Sporny: WebIDL is used in specifications. VC datamodel is syntax agnostic. we're trying to decide what syntax to put in the core spec
Manu Sporny: Consensus we'll include JSON, JSON-LD (probably),
Manu Sporny: Do we need others? like XML, WebIDL, etc. or put them in another document?
Dan Burnett: Asks why this one to go to a separate doc?
Manu Sporny: WebIDL will come in at some point - when it's integrated into the browsers. in the meantime, it's an odd expression to those who aren't browser vendors
Manu Sporny: Include expressions that are more accessible and common for these participants
Manu Sporny: Move webIDL out until we come a real working group
Nathan George: +1 On separate document. I agree with manu's point that his is overly browser-specific, and we should address it later if we can.
Jonathan Holt: Had to rejoin, got kicked off.
Dave Longley: We've prohibited ourselves from doing browser specification of this work
Derek Callaway: Occasionally some application servers will use IDL for CORBA IOR (Inter-Object Request) brokers or Java RMI (Remote Method Invocation) .. no doubt it's a bit obscure, but not unheard of in enterprise-class software
Shane McCarron: Also hard to express this in webIDL, since it's a w3c recommendation, we should try, but we can do it in another doc
Dan Burnett: Generally agree w/ discussion. goal: generally define a datamodel, and show that it can be expressed in >1 manner. we do ourselves a disservice if we don't do it in more than 1
John Tibbetts: Disagree: we should do it one way and it should be JSON-LD
Derek Callaway: You could write a WebIDL->YAML translator (or whatever syntax description language you want)
John Tibbetts: In the core specification - move the alternatives to separate documents w/ many difference representations
Matt Stone: +1 JohnTib
Dave Longley: +1 To John's suggestion of a separate doc with many different representations as examples
Dan Burnett: Please continue the discussion in the issue on gitHub
Shane McCarron: I think it would be worth our while to help fix WebIDL so you can express things like these data structures