The Verifiable Claims Task Force

A Task Force of the Web Payments Interest Group

Verifiable Claims Telecon

Minutes for 2016-12-13

Richard Varn is scribing.
Matt Stone: WG vote has been submitted to the AC, here's our agenda for today.
Matt Stone: 1. Vote for Verifiable Claims WG is open - Vote!
Matt Stone: 2. Use of JSON/JSON-LD in spec[2]
Matt Stone: 3. Use of WebIDL in spec[3]
Matt Stone: 4. Privacy Section Discussion[1]
Matt Stone: 5. Brief intro to github pull requests
Matt Stone: 6. Reference library discussion (javascript/python)
Matt Stone: 7. The role of digital claims (Tim Holborn)
Matt Stone: Let's move reference library up in agenda to now
Matt Stone: Move point 6. reference library to point 2. in the agenda

Topic: Vote for Verifiable Claims WG is Open

Manu Sporny: Must be a w3c member to vote. Still time to join.
Manu Sporny: I've been encouraging people to vote via email especially to those who expressed interest and support. manu will send you instructions on how to join and vote
Manu Sporny: If you are a member get you AC rep to follow the link and vote
Jonathan Holt: Is there a secret hand-shake if I join?
Manu Sporny: Of course there is a secret handshake. :)
Manu Sporny: Will send out a call to action on the vote. feel free to reuse it to encourage others you know.
Manu Sporny: Don't forget to vote mail as well as recruiting emails will go out
Matt Stone: +1 For manu's persistence
Dan Burnett: +1 Thanks Manu
Jonathan Holt: Is making appeals to people to join the W3C and vote
Manu Sporny: We are looking to garner 30-50 votes

Topic: Reference Libraries

Manu Sporny: Useful links:
Christopher Allen: Looking for signature compatibility with bitcoin/koblitz [scribe assist by Matt Stone]
Manu Sporny: More specifically - We're talking about implementations for this:
Jonathan Holt: For the notes: Specifically, I am working on getting the American Board of Medical Specialties (ABMS) the group that verifies board certification for physicians in the US to join the W3C and sponsor me to represent them. I am an advisor to the Database and Information Technology and during our last meeting I gave them a briefing regarding the importance of this work.
Manu Sporny: +1 Jonnycrunch! That would be great!
Christopher Allen: Looking for python libraries to support signatures
Manu Sporny: We're trying to do the Python implementations here:
Christopher Allen: Making libraries in JavaScript and Python in order to show interoperability [scribe assist by Matt Stone]
Christopher Allen: 1) Where should these be hosted, 2) what other requirements to move these forward, 3) small honarium available to move it forward [scribe assist by Matt Stone]
Christopher Allen: Javascript and python to be used as they have supporters and by using both we show interoperability. who else interested. where to host. what other requirements to move forward. he has a small discretionary honoria type budget to help.
Glen Willen: Part of rebooting reputation poc
Christopher Allen: The Proof of Concept from #RebootingWebOfTrust of a Portable Reputation Toolkit:
Manu Sporny: Need two interopberable representations to be a valid standard so this work on python and js linked data signatures would meet the bar for exiting the standards world but more would be welcome like ruby
Jonathan Holt: Please add me @jonnycrunch on github
Manu Sporny: Can use this repository. posted above. just ask.
Christopher Allen: Time stamps and open time stamps need work and a home
Christopher Allen: There is a slack so far at #RebootingWebOfTrust
Christopher Allen: It doesn't have to be there.
Christopher Allen: Weboftrust.slack.comn
Manu Sporny: Stone what irc and server will we use and manu answered that we will get a vcwg irc channel
Gregg Kellogg: The WG may also choose to create a mailing list specifically for developers.
Christopher Allen: Ooops,
Gregg Kellogg: Or, a vcwg-comments mailing list.
Christopher Allen: Yes, koblitz and timestamping are not likely WG work.
Manu Sporny: Will use the irc channel. wot slack channel will do out of scope discussion and other communications forums will support the work we do on official channels. like work on koblitz.
Christopher Allen: Would like to see a library of to the spec ideas for further definition and work
Jonathan Holt: UPort from Consensys, not sure if they conform to standard
Christopher Allen: I'd like to see a verifiable claims repo of examples of verifiable claims
John Tibbetts: That extended transcript work related to competency based ed from IMS rebooted this morning to be redone and put into the current vctf structure and format to be shared as done to this group
Christopher Allen: And maybe a wishlist (an easy one: I am a person"
Matthew Larson: On Acclaim team from pearson and he is working on a signed OBI vc. wants to be able to verify that his signature is correct
Matthew Larson: That way we will have a vc from Acclaim
Christopher Allen: Yeah @stonematt!
Matt Stone: :)
Manu Sporny: Key management in scope for wot but not vcwg
Manu Sporny: Christopher reply--we should collect our various examples into one place on GH.
Dan Burnett: +1 Manu, the playground would be great
Matt Stone: +1 Playground
Christopher Allen: This playground does have ability to sign with a koblitz key (
Manu Sporny: A number of us have been talking about a vc playground to create one, digital key signed, and seeing if the are formatted correctly and valid. Digital bazaar will work on creating the playground separate from the digital sig and LD sig playground. if you are interested in working on that let us know.
Christopher Allen: A bitcoin private key starts with L)
Manu Sporny: Will put a repository on open creds for now for vc examples
Drummond Reed: Manu, please add me as well
Drummond Reed: Note that I'm not on the call yet - just monitoring the chat - had a conflict - will join shortly
Christopher Allen: What is the nature of a conforming task. would like to take X and run it through various tools like JSON-LD, VC, Python one, Ruby one, and validate it. likely a topic for test suites discussion.

Topic: Use of JSON/JSON-LD in spec

Manu Sporny: Question is do we put JSON-LD up front or keep it hidden. thinks we keep in hidden and not require it as some will work better with JSON and others would be expecting JSON LD so better to not require it and would be unhappily surprised
John Tibbetts: +1 On Dave's description
Dave Longley: Start off simply with JSON and add other more complex and rich claims can say then how LD is helpful but not required for less complex forms
Dave Longley: To be clear, they should always be JSON-LD, but for simpler claims, most people should be able to write software that doesn't need to do any processing.
Dave Longley: So i'm not suggesting we don't use JSON-LD for simple claims, but that we don't require processing for many basic, useful applications
Dave Longley: So it behaves like JSON, but keeps its semantics.
Jonathan Holt: Need address meaning/semantics in the model
Matt Stone: Data model independent of implementation technology has confounds and overlaps in real world
Jonathan Holt: +1 For semantic interoperability
John Tibbetts: Dave longley expressed what he was trying to get as well. the underlying data model is a graph not a hierarchy for vc. certain simply reps can be thought of in a JSON POV but other implementations require a graph way of thinking and implementing
Matt Stone: +1 For "it's a graph, not a hierarchy"
Manu Sporny: Answering matt's q--we have to talk in the wg to address the spec in the abstract and still address the realities of real implementations. the data model is a graph. we want semantic interop. need a way to not shun those who do not know some of the more up to date current languages and methods. make it so they do not have to know the new to participate.
Drummond Reed: So, simple question for Manu: is the ultimate spec going to use JSON-LD or just plain JSON?
Manu Sporny: Also note we prefer to use JLD, but many are alergic to it among browser makers for sure and think there is no need for JSON-LD. need to meet the criticism and do it in a way to make both sides happy as has been show or discussed lately
Matt Stone: +1 To maintain support for the more complex and composed claims
Dave Longley: Agree with manu in gen. take lesson from need compatibility between simple and complex so that can work with LD without explicit knowledge of using/doing that. but still need explicit tacit knowledge to do the complex work. LD will be needed and be LD compatible. scale and decentralization issue
Drummond Reed: That's the answer I expected - I just wanted to know if you thought the W3C folks were going to force it to become just JSON
Dave Longley: For people that aren't familiar with JSON-LD ... it is a subset of JSON, so everything you write in JSON-LD is valid JSON.
Manu Sporny: +1 For it being a persistent resource/URL.
Jonathan Holt: Need a good implementation guide to help in adoption which encourages use of LD up front were warranted. need a persistent place to store hash. looking at IPFS.
Matt Stone: Can we take this off the agenda and move on to repo?
Jonathan Holt: Issue with ipfs with that is needs to be a deterministic json document. line breaks ( pretty print) in json changes the hash of the document.
Dave Longley:
Drummond Reed: Who runs Is it W3C?
Jonathan Holt: +1 For purl
Manu Sporny: How do we make it more palatable to other nonLD people. managing the url is part of the work for longlived decentralized contexts. shared link. IPFS is also a option. no final context one way to do it.
Manu Sporny: Drummond, it's a number of companies - Spec Ops, Digital Bazaar, Botatsu... working in a way like Sovrin Foundation.
Drummond Reed: It's run by: [scribe assist by Dave Longley]
Drummond Reed: Cool, thanks
Manu Sporny: Specifically, it's NOT run by W3C.
Jonathan Holt: Can cache some contexts
Drummond Reed: Gott run now to the XDI TC meeting
Manu Sporny: is run by a consortium of companies that have committed to keepign the site alive for 100+ years.
Manu Sporny: +1 For not meeting on 27th

Topic: Next Meeting

Matt Stone: Proposal to not meet on the 27th.
Drummond Reed: +1 To not meeting on the 27th
Gregg Kellogg: There are other uses (e.g., WoT), where having a context that is known not to change can be important.
Matt Stone: We will meet next week but not the 27th.
Matt Stone: Burn are you still there?