Credentials Community Group Telecon

Minutes for 2014-12-09

Agenda
http://lists.w3.org/Archives/Public/public-credentials/2014Dec/0029.html
Topics
  1. Introduction to elf Pavlik
  2. Signatures Update
  3. Context Work Stream Update
  4. Roadmap Work Stream update
  5. Hosted vs. Signed
Organizer
Manu Sporny
Scribe
Manu Sporny and Dave Longley
Present
Manu Sporny, elf Pavlik, Eric Korb, Brian Sletten, Dave Longley, Victoriano Giralt, Mark Leuba, Mary Bold
Audio Log
Manu Sporny is scribing.
Manu Sporny: Any changes or updates to the agenda?
elf Pavlik: Agenda sounds good to me!

Topic: Introduction to elf Pavlik

elf Pavlik: Glad to be here, I've been involved in W3C Working Groups and a variety of open source/culture communities for a long time now. Most of my introduction to the group can be found here: http://lists.w3.org/Archives/Public/public-credentials/2014Dec/0034.html
Eric Korb: Documentary about elf Pavlik: http://vimeo.com/61478618

Topic: Signatures Update

Brian Sletten: I'm happy to help out on this work stream.
Dave Longley: Basically, we're about getting the RDF dataset algorithm up to date... we have implementations in JavaScript, PHP, Python, and Java.
Dave Longley: Those implementations are in sync w/ one another - so the spec needs to be brought up to date.
Dave Longley: We just put out the JSON-LD signatures library.
Manu Sporny: Here's the library release for JSON-LD signatures: https://github.com/digitalbazaar/jsonld-signatures
Dave Longley is scribing.
Manu Sporny: At some point we're going to need many more organizations to start using the digital signature mechanism for JSON-LD than just the people in this group. We need to make sure we provide the tooling to make integration easy. The good news is that we've had these libraries for multiple years now, we've done bug fixing and patching here and there, but they are generally stable. W3C depends on outside groups to do implementations, good news is we have 4 implementations (in different languages). We're ahead of the game in that respect.
Manu Sporny: We are in experimental stage way before we need it [scribe assist by elf Pavlik]
Manu Sporny: We can move along at a fairly rapid clip.
Manu Sporny: What we are trying to do with the JSON-LD signatures library is to put something out there to point to, so we can point to working code where that happens. We target node.js and the browser with that library, kill two birds with one stone. We want to integrate this with the JSON-LD playground so people can play with it.
Manu Sporny: There's been some back and forth with Melvin over adding some new features, etc. So, just a heads up to the group we've got this library out there under a BSD license, use for commercial or non-commercial purposes freely.
Brian Sletten: I'm assuming there is interest in implementing in other languages like Java, etc.
Dave Longley: We do have a C++ version out there on Github [scribe assist by Manu Sporny]
Manu Sporny: Would it be difficult to bring that up to date? [scribe assist by Manu Sporny]
Manu Sporny: Yes, the biggest part is getting the normalization algorithm done, once Dave gets the spec updated we can ask for more implementations.
Dave Longley: Getting normalization working is not too difficult by looking at other code... we need to get new JSON library. [scribe assist by Manu Sporny]
Brian Sletten: Is there a priority/roadmap? [scribe assist by Manu Sporny]
Manu Sporny: Yes, basically this is what we want to do: Create Roadmap and technical specs. Roadmap should outline how we go from where we are today to full blown credentials ecosystem that we want. Basically, we create roadmap, convince Web Payments IG (or W3C membership) that roadmap on credentials is worth doing. Roadmap outlines the ecosystem - schools, universities, trade schools, recruiters, corporations/universities/non-profits that hire people, expiration of credentials, etc. We move the specs we create into W3C WGs, put them through the W3C process. Then in parallel focus on implementations/helper libraries that make the ecosystem easier to implement for everyone in the ecosystem.

Topic: Context Work Stream Update

Manu Sporny: Since Nate and Sunny aren't here, we can't make much progress on it. Hopefully they can give us an update via mailing list.

Topic: Roadmap Work Stream update

Mark Leuba: One area I'm interested in getting feedback fromt he group about - emphasis of the roadmap. One part of the process is getting through w/ technical specifications to hopeful ramping of the standards status.
Mark Leuba: Another aspect is demonstration of the concept.
Mark Leuba: My interest is in the latter.
Manu explains what the roadmap is trying to do (vision, macro level exchanges, detail how standards align w/ the ecosystem we're trying to create).
Manu Sporny: Roadmap should outline how we go from where we are today to full blown credentials ecosystem that we want. Basically, we create roadmap, convince Web Payments IG (or W3C membership) that roadmap on credentials is worth doing. Roadmap outlines the ecosystem - schools, universities, trade schools, recruiters, corporations/universities/non-profits that hire people, expiration of credentials, etc. We move the specs we create into W3C WGs, put them through the W3C process. Then in parallel focus on implementations/helper libraries that make the ecosystem easier to implement for everyone in the ecosystem.
Mark Leuba: I'd really like to also get involved in the latter - appreciate the explanation. [scribe assist by Manu Sporny]
Manu Sporny: So that addresses the first question, "What do we want to deliver in this document"
Manu Sporny: After weeks of discussion that's kind of the approach we're settling on (from web payments group), we're just reusing the thinking from that group and take this and feed it into the Web Payments IG.
Manu Sporny: It may be more digestible by them with that approach.
Manu Sporny: General thoughts from the gruop?
Eric Korb: I'm wondering what we have to do regarding the concept of integration/support of other standards or proposed standards. For example, Open Badge Interface specification are different from Identity Credentials. Do we need to think about interoperability or not? [scribe assist by Manu Sporny]
Manu Sporny: In general, W3C frowns on standards that don't reuse existing standards, the official stamped by W3C 400 member vote, etc. The space that we're in with the credential stuff, nothing really exists. As a secondary thing, no standards body wants to see two standards that do the same thing out there.
Manu Sporny: That's why we want to align with BA and don't create two standards that do the same thing in the space.
Manu Sporny: To specifically answer your question we absolutely have to think about interop with open badges, but that doesn't mean historical interop, what we have to say is here is the vision for the future for the next year, and say the BA and what we're doing is interoperable. Our roadmap needs to say something to that effect.
Eric Korb: Ok, I agree... have another point, but will pass until I get my thoughts together. [scribe assist by Manu Sporny]
Eric Korb: Context and vocabularies - is that in the roadmap? [scribe assist by Manu Sporny]
Manu Sporny: Yes, without those, we don't have interoperability. [scribe assist by Manu Sporny]
Manu Sporny: Yes, if we don't have JSON-LD contexts and vocabs our systems won't interop.
Dave Longley: What we can do is look at most basic credential that we've been working with... based on that, these are the minimum things required in a credential. [scribe assist by Manu Sporny]
Eric Korb: What about a validator? [scribe assist by Manu Sporny]
Manu Sporny: Out of scope typically.
Manu Sporny: The group can decide to create a validator... let me break it down, things for the standardization process and things for being successful. We need technical specification, 2 interop implementations and a test suite. That gets us through the standards process.
Manu Sporny: That doesn't get you a market adopting what you build, that just gets you a technology to be reviewed by companies, etc.
Manu Sporny: Next we need things like helper libraries like the JSON-LD signatures lib, commercial implementations, companies saying they back the work, etc. these are things that are needed to make sure the technology can be successful in the market.
Manu Sporny: If we can get some hot new startup to say they are using the credentials spec/OBI stuff/badges to express things that's a big win. If we can get a c-sharp JSON-Ld signatures implementation and a credentials creation library that's another big win.
Manu Sporny: To help make this stuff successful getting off the ground.
Brian Sletten: Are there any good roadmaps for this stuff to use as examples?
Manu Sporny: No, a good question, but no, traditionally W3C hasn't had those sorts of things this early on.
Manu Sporny: Let me try and think a bit on that though.
Manu Sporny: Might be able to find a good example, but not off the top of my head.
Eric Korb: I think there was a conversation we had w/ lessons learned w/ JSON-LD. [scribe assist by Manu Sporny]
Eric Korb: Many of you did JSON-LD, but we have new people. [scribe assist by Manu Sporny]
Manu Sporny: Prepping all these docs in a CG so a WG has a set of docs to pick up and run with. That's one of the biggest things you can do, it cuts 1-2 years off WG time just by doing that. And we have much more leeway in how we create those docs ... we can move quickly, they can't move as fast (in WG).
Manu Sporny: Having a roadmap doc is also good.
Manu Sporny: The thing that made the JSON-LD work different is that one of the WG has already decided that they wanted to solve the problem. W3C companies said they wanted to solve the problem and we had a solution ready to go. The W3C has not to date said they wanted to solve credentials on the Web. That's one reason we need to have a roadmap to say why we want to solve this on the Web and how we will solve it so we can propose it to the IG.
Manu Sporny: W3C has stopped creating wikis for CGs, i don't know if you have to request one or not, if we want to work in wikis we can do that, typically we've just been working in respec docs.
Manu Sporny: The editors can use whatever editors they want, just in general W3C wants people to work in collaborative environments, and non-proprietary, like wiki/respec, google docs is next, ms docs are probably last.
Manu Sporny: Mark and Mary are the task leaders for the roadmap.
Manu Sporny: Every check in to the opencreds repo goes live immediately.
Manu Sporny: Any modifications made on the spec show up on the website within 15 seconds.
Manu Sporny: Elf, if you want to work on any of this stuff we'd be happy to grant you access and do PRs, things of that nature.
Manu Sporny: How do we go from the roadmap to the future we want. The strategy is that we're going to work on technical specs on what we believe the ecosystem should look like, we're going to demonstrate how various players interop, propose to the IG, IG will make WGs to work on the work items. The WGs start up and standardize the work we're doing here while we work on parallel on the tooling necessary for success of the standards in the market.
Manu Sporny: That's the general strategy we've come to consensus on in the last couple of months, thoughts?
Brian Sletten: Sounds good.
Brian Sletten: That sounds like a reasonable strategy. [scribe assist by Manu Sporny]
Dave Longley: +1
Eric Korb: Does it make sense to setup a timeline for goals? [scribe assist by Manu Sporny]
elf Pavlik: Example of optimistic timeline: https://www.w3.org/wiki/Socialwg#sched
Manu Sporny: All working groups have to set up a timeline and say when they have to be done, and it always takes way longer than they think it will take. But it may be a useful exercise to set some goals and say by when, etc. If we want to give people the idea that we could be done in 2-4 years, we don't want people to think it will take 15 years to do, they are going off the wheels, etc. A timeline would be good with the understanding it will take way longer than we think it will.,
Manu Sporny: If someone wants to take a hack at the timeline, please do, it would be a good discussion for the CG to have.
Manu Sporny: The roadmap doc might not be done for another 3 months, hopefully sooner, but it might take that long to get reviews, comments on it, etc.
Manu Sporny: Any other thoughts on the roadmap doc?
Manu Sporny: If possible, mark and mary, do you think both of you working together, could you get a draft of this document out by the second week of january?
Mark Leuba: I believe we can.
Mary Bold: Agreement from me, I'll back him up.
Manu Sporny: This will be our last call for a while to focus on getting documents in shape and because of the holidays.
Manu Sporny: When we come back in January we should have something to review as a group, that's the goal.
Manu Sporny: The goal I'd like the group to set today.
Dave Longley: +1
Manu Sporny: Mary, Mark - remind me to show you the docs

Topic: Hosted vs. Signed

elf Pavlik: I noticed that Open Badges supports both hosted/signed, but the stuff this group is working on is largely signed. Is there still support for hosted badges?
Manu Sporny: There are still hosted vs. signed. The vast majority of the open badges (in the previous minutes) is hosted, not signed. I think there are a number of people in this group that would like to see this trend change; so the default will be signed and if an issuer site goes away you can still get to the badge. We want people to have the ability to self-host these badges and we want the badges to be valid when they self-host it.
Manu Sporny: Hopefully that kind of answers that question, both are supported (hosted and signed) in open badges, mostly hosted today, many members here want to do signed as the default, and we want people to be able to host their own badges.
elf Pavlik: So hosting a badge would be possible but verification would be linked off elsewhere, etc.
Manu Sporny: Wrt. working on documents, please try to take advantage of the mailing list, so community can follow conversations. [scribe assist by elf Pavlik]
elf Pavlik: +1