The W3C Credentials Community Group

Verifiable Claims and Digital Verification

Go Back

Credentials CG Telecon

Minutes for 2020-07-07

Tyler Adams: Hello everyone! :)
Kyle Kemper: Great to be here! Hoping everyone is having a terrific day
Heather Vescent: Great to have you here Kyle!
Gregory Rocco is scribing.
Kim Hamilton Duffy: Gregory Rocco works with Wayne on the Spruce team
Kim Hamilton Duffy: Anyone else new to the call that wants to introduce themselves
Tyler Adams: I work on moonlight and vivid SSI project
Kyle Kemper: Presenting a little bit - great to be here, founder of Swisskey, crypto since 2013. Technology will help us unlock golden age.
Heather Vescent: Love it Kyle! New golden age, I am with you!
Kim Hamilton Duffy: We'll have another opportunity for intros next week - this is also where we ask for reintroductions. Will jump over the list. Joe Andrieu, Adrian Gropper are you here?
Adrian Gropper: Volunteer cto of a non-profit, I lead a project for many years trying to demonstrate self sov. technology, not just ID but how we use other standards to let people run their own authorization servers where we process VCs and DIDs

Topic: Announcements & Reminders

Kim Hamilton Duffy: Next is Anil -
Anil John: Save it for the Q&A
Kaliya Young: John jordan was in a bad accident, I created a virtual card for folks to sign - will put the link in the chat
Kim Hamilton Duffy: Thank you, sorry to hear about that accident. For the other announcements and reminders, I'm including a link where we track - not a whole lot going on but we do have Identiverse which turned into a virtual conference and that series began in June and is going through the 6th of August. So in case you are still interested in participating, joining the link is here.
Kim Hamilton Duffy: Our other ongoing reminders: usual CCG timeslot, we do have a couple of taskforces associated but the first one is the VC for Education taskforce which is at mondays 8am PST, 11 EST, DID resolution calls are at 1PM PST, 4PM EST.

Topic: Progress on Action Items

Kim Hamilton Duffy: That's it for the announcements and reminders - next in the agenda, we have a bit on products and action items.
Kim Hamilton Duffy: So progress on action items - a couple we're working through. I'll include a link to it in IRC - there are two items and we're hoping to move through these quickly. Basically we don't need to resolve these, just identify blockers and figure out any steps.
Manu Sporny: +1 That item belongs to DID WG.
Kim Hamilton Duffy: Issue #109: closed out, issue there is that the CCG believes this is going to be passed over to the DID working group and so it looks like we just want confirmation that it belongs to that group and we can close it.
Manu Sporny: It's confirmed that it's on Dan B's plate from the DID working group and we'll be handing that over to the wg so they can finish that off to get the tech/architecture group input.
Kim Hamilton Duffy: I've added a comment and will follow up. Next item: complete update to RSA signature suite. We had a bit of a discussion, what we wanted to do - no traction on the work item but we identified a related issue to create a new issue. Effectively what manu proposed: specific concerns around JWS 2020 so the current proposal is to close this one and create a new one to track those concerns. Thoughts or objections?
Kaliya Young: I was on q for announcements for jj
Orie Steele: It covers the NIST curves and ed25519 and secp256k1
Manu Sporny: Right so the RSA sig suite discussion resulted in Orie proposing a JWS signature suite like a jose signature suite, that discussion will likely happen in a DID special topic call - we need to figure out how we're going to integrate that with --- used for expressing legacy signature suite, RSA, things like that b26r1, if anyone's interested in participating -- based on the outcome of that, we may want to discuss the outcome of that in the communit[CUT]
Manu Sporny: Not this week but next week we'll have a special topic announced - outcome of call will be made public, CCG will decide if they want to discuss the outcome.
Kim Hamilton Duffy: I will add an update to this issue and we will follow up.
Kim Hamilton Duffy: Okay great - that is it for action items
Wayne Chang: Call for objections sent almost 2 weeks ago
Kim Hamilton Duffy: Before we move into the meat of the call, want to talk about method registry - we moved it over to the DID working group and that means we're retiring it from CCG - so Wayne sent a call for objections two weeks ago. Wayne - what was the deadline?
Wayne Chang: I think it was almost two weeks - if there's no objections, we're clear.
Kim Hamilton Duffy: Last call for objections - we have not received anything so far so we can proceed with closing it out so we'll move forward.
Kim Hamilton Duffy: If you heard some choppy audio, in addition to our communication issues we had a gremlin muting speakers. We should remind everyone if you can speak clearly and slowly, that would help a lot. Now we are ready for the roundtable
Kim Hamilton Duffy: I'm going to ask Wayne to set up expectations for discussion

Topic: Roundtable: “What’s in a wallet?” Part 1

Wayne Chang: Basically, we are growing as an ecosystem, whenever that happens you run the risk of people talking over each other - just going on guidelines will help us have productive discussions to get to outcomes we like instead of flame wars.
Wayne Chang: There's a bunch of emotional energy spent - how do we keep the conversation on track. We opened this communication culture issue and we captured 4 links so far - just some reading material, just some good social guidelines to follow rather than people showing off deep intellect and to be kind to one another.
Wayne Chang: Just to cherry pick: some social rules: no "well actually" - when you try to make a point and someone wants to nit pick that's tangential and serves no purpose. That doesn't move anything forward. No surprise - especially for newcomers, also subtleisms - avoid - if you're trying to make an example, think carefully and make sure the analogies are extensible and welcoming. These are some examples of behaviors to avoid as a community - no hard and[CUT]
Wayne Chang: Last of all, let's try to listen to listen - I hope that sets a pretty good headspace for everyone to be in this week.
Kim Hamilton Duffy: Thank you Wayne - we have 4 speakers today, Manu, Chris, Dan, Kyle K - each are scheduled for 5 minutes with the remainder for Q&A. So I'm going to kick off Manu - i'll be aggressively booting people off after 5 min.
Manu Sporny: Thanks Kim - and thanks to the chairs for the invitation. I apologize, the diagram I put together is overly technical so I'm going to ground the discussion a bit.
Manu Sporny: There are many different views - what it is, what it does, some people think of wallets like cryptocurrency and they deal with cryptographic material, others think of them in an analog to your physical wallet. You keep credit cards, loyalty cards, you know, pieces of paper that are important for you when interacting with the world.
Manu Sporny: Many different variations in between - there's no clear understanding about what is, what isn't, and what a digital wallet should and shouldn't do. I'm going to outline standards that in pre-standards work that we've been doing here as well as work in the DIF storage working group in a tech architecture kind of way. Apologies to the non-technical or those not following the work.
Manu Sporny: What I'm getting to is a simple wallet architecture: from my perspective: a wallet is something that you use that is digital, on mobile phone/in web browser, holds digital things (pictures, your ID docs, your dependents healthcare certs,), it stores that for you, it's yours, you control access to the information, and basically you use it to manage critical parts of digital life.
Manu Sporny: At a high-level - one definition we can use - I'm sharing the link in the chat to a diagram I sent yesterday around what these things that make up a wallet look like. These are the technical bits and pieces that we've been working on and how they're used together to provide this digital wallet experience. To interact with issuers, verifiers, etc.
Manu Sporny: There are three components that we've identified - number of companies in this group that have done work around encrypted data vaults that have wallet front-end services and have some kind of key management - there are three pieces.
Manu Sporny: First - the one we all interact with - the wallet service. It's a website, an app, has buttons you click on and basically interact-with. This is the GUI to manage the digital things in your wallet. That's the lowest purple box - we use something called a credential handler API to move things in and out of your wallet -
Dave Longley: Credential Handler API == "CHAPI"
Manu Sporny: The wallet is backed by two other things - the wallet facade is in the front, but behind the scenes: 1. key service - this is what deals with your pub priv keys and gives you the security you need around the wallet. Public private cryptography backed by hardware modules that are difficult to compromise so there's a digital signature aspect of the wallet and you get access via web kms service.
Dmitri Zagidulin: For those interested in CHAPI - this repo has link to the CHAPI spec, as well as demo sites:
Orie Steele: I wish WebAuthN supported raw signatures... and could be used for things other than authentication.
Dmitri Zagidulin: Encrypted Data Vault spec (as part of the Secure Data Store spec) -
Manu Sporny: And then the other thing behind the scenes is the encrypted data vault: the thing that stores and protects the data from prying eyes and ensures consent. Those encrypted data vaults hold aspects of your life -credentials that you need for work, music that you listen to but you keep in another vault, and maybe one for family photos in another. We use the vault api to get to it
Manu Sporny: In it's most simplistic view, that is the fundamental technical part that goes into a wallet and more-or-less how we view the most basic form of wallet.
Kim Hamilton Duffy: Thank you so much - chris you're up
Christopher Allen: So i'm sharing the general paper that this came from and I'm now sharing the very specific diagram in there.
Christopher Allen: Basically I wanted to walk folks through this diagram to explain what we're trying to do and say what we felt was missing - Joe and I initially took a look at a simplified diagram around digital ID wallets which is included in the paper, and then also we tried to do the same for digital currency wallets and basically developed a more advanced look at both of those because a lot of the things are common but areas where they're different.
Christopher Allen: What you see here: a lot of nodes and different aspects of what is in your wallet, clearly toward the center you see what people perceive as the wallet (going down - human interfaces, trusted interfaces to hardware - need to trust we're getting the right information) - upward we have various kinds of storage and relationships that the wallet needs to do
Christopher Allen: Then we have this red line: the trust line where the wallet has to trust cryptographic surfaces through API, airgapped, but that line needed to be marked as important that leads to cryptographic services under control of the wallet and then of course going right as we go toward network interface, we have a trust surface of the wallet's network interface to the broader network.
Ryan Grant: Sorry, the dottet lines represent what? the gremlin cut you off.
Christopher Allen: Dotted lines represent ---- API that there was state information or other information being conveyed whether or not it was a proof - or things that weren't strictly an API. Had hoped to take this to the next level, felt that a number of these nodes were named differently by different platforms in the self-sov community, some of these nodes should be 2-3 stacked items with different names as a stack.
Christopher Allen: I am also fairly certain and know that we're missing nodes and anything in here about recovery/revocation/delegation/reputation/onboarding/TOU, various kinds of evaluative components, and updates and pricing - if you're going to have a cryptocurrency thing in here. We had hoped we can get 1-each from the different wallet vendors on different SSI platforms to help us take this to the next level but also start naming the arrows. We didn't [CUT]
Christopher Allen: Was a mistake - we needed to name the arrows as well as defining the nodes - that's where we left the project.
Kim Hamilton Duffy: Thanks so much - now we have dan b
Kim Hamilton Duffy: Need to come back to him - kyle would you be ready to go
Kyle Kemper: Alright - thank you kim for having me and first off, want to say big thank-you to this community who is developing the core infrastructure on which this tech will play out. My name is Kyle Kemper - been in crypto since 2013, in 2015 I had my wallet vision download around what will be truly possible when we have SSI and have self-custody wallets that are unified.
Kyle Kemper: I wrote a book called "unified wallet" - it's a vision on how we can unlock seamless experiences. Personally I'm not a deeply technical person, I am a developer - my journey has brought me to the point where I'm building a brand called SwissKey - problem in the ecosystem of trust.
Kyle Kemper: Closely aligned (digtial currency and ID), currency is the cash slip and the card slots are the verified claims, assets, objects in our super wallets. Been building swisskey for the last year being in switzerland and being impressed with their values and sensibilities - hallmarks of neutrality, and their strong and subliminal messages -
Kim Hamilton Duffy: Daniel said he was muted fyi. Have we muted him here?
Kyle Kemper: It represents safety at the subliminal level - I'm seeking to build out a suite of wallet products to enable secure/simply operate in this new digital paradigm. We are going to be holding keys for different solutions be it money, identity, loyalty. I am grateful to be here to express this project that I'm working on and make myself available to everyone on the call. I feel we all share this transformational purpose.
Kyle Kemper: Make this all accessible to the masses.
Wayne Chang: Hey kyle - really quickly, the topic is what is a wallet - wanted to leave some room for that.
Daniel Buchner: Hey Kim
Daniel Buchner: On IRC now
Kim Hamilton Duffy: Hi Daniel, we found the culprit. You shouldn't have to do anything to unmute. We unblocked you
Kyle Kemper: In terms of the technical path, I'm working with edge wallet, and then we are working with Tangem who are the makers of a hardware wallet, NFC card, they're exploring into the ID space so we can have a physical card as well - the digital wallet is important but there's a need for hard cards. These represent the "DVDs" for netflix.
Kyle Kemper: That's basically kind of it from me in terms of this - thank you - I yield back my time.
Kim Hamilton Duffy: Thank you so much, now we figured out the root of the issue - we muted daniel - can you try to speak.
Daniel Buchner: Dan Buchner, work at Microsoft. We have a wallet in the market doing legacy identity things like logins, etc. 30M users on that current app and that's what we're delivering some of our stuff with (DID) - so users have keys
Daniel Buchner: Part of the aspiration is to create portability specs so you can re-instantiate it on other wallets. 3 components: 1. encrypted outer envelope: data you need to encrypt needs multiple types of encryption strategies, shamir, whatever you want to do to secure it. 2. There's a lot of data over time in this payload - need to make it something that isn't 100s of MBs of JSON. Something that can be inflated and understood.
Daniel Buchner: What goes into that payload may include things like keys but context as well. Context to use your DIDs in - many people might think "well, good enough to have keys for DID and re-instantiate" but context is a sensitive security thing to protect, but if I had 1000 DIDs and didn't know what I used them for, I would be lost.
Daniel Buchner: If I just use random DIDs with people, I would slowly be leaking sensitive context and relationships with people. Whatever we do should be able to handle that goal. Those are the major pieces we hope to standardize and we're here to work with folks on that.
Manu Sporny: +1 To what Daniel said! :)
Kim Hamilton Duffy: Fantastic, thank you so much -we've gone through our speakers and it's good time for Q&A
Kyle Kemper:
Kim Hamilton Duffy: Any questions - just add yourself
Kyle Kemper:
Kyle Kemper: My mission: make crypto and DID accessible to the masses
Wayne Chang: So I wanted to summarize a bit and people feel free to type in the chat. We heard "what is a wallet" - we heard a bunch of different perspectives - interfaces to capabilities and mapping - what's that. Philosophically, what does a wallet make someone feel or how is it used by a person - and what are the technical requirements we have on scalability and security.
Wayne Chang: Common thread: interfaces and the ability to make components for these interfaces. Also mention what prompted this conversation was the recent presentation of what was called the universal wallet spec by Ori at Transmute - as we begin to see these emergent standards - that's all the comments I have.
Kim Hamilton Duffy: We have adrian
Orie Steele: Good question :)
Adrian Gropper: Quick comment - I am certainly a user not a developer, I'm trying to figure out the business model as what are essential vs. optional. My question: how is a SSI wallet going to differ from my 1Password wallet which right now is the closest in terms of business models and functionality as far as I can tell - in other words, has a lot of characteristics plus APIs and it's self sov to a large extent because it doesn't share data.
Kyle Kemper: If interested go to to download Edge Wallet; we are working with Edge to rebrand their wallet...
Heather Vescent: Great question Adrian
Kim Hamilton Duffy: Did any speakers have a response to that? If more than one we'll let the first person go.
Kyle Kemper: In terms of the business model in the self sovereign wallet they all speak to our project and we are making revenues when people are buying and selling crypto, trading, redeeming - we're selling the physical cards as well - we're going to be utilizing the physical cards as recovery methods - comparing a 1Pass or a LastPass to the wallet, those are excellent examples of keychains but they're dumb. The assets can't speak to each other
Daniel Buchner: We hope wallets are as free and open as browsers, and should cost users $0 to run (possibly with sane constraints for consumption, if there are any major cost incurrences)
Kim Hamilton Duffy: I'm talking with the chairs about key-management was going to call out Christopher and Manu or if they wanted to respond.
Kyle Kemper: Wallets definitely need to be free
Christopher Allen: I'm actually fairly concerned about wallets - I feel that a lot of the business models for them are subtly corrupting in the cryptocurrency field or selling some hardware or you're making your money from accepting a particular cryptocurrency so you'll get 100-200k if you do some oddball currency.
Manu Sporny: +1 To the corrupting influence of cryptocurrency wallets :)
Dave Longley: "Free" has a cost :)
Dave Longley: "Facebook" is "free"
Daniel Buchner: Wallets, DID Networks, and other core infra/UA stuff should be so mind-numbingly commoditized as public utilities, it should cost you more to monetize them than you would make in doing so
Christopher Allen: Or you'll get a bunch of tokens from a company that you can speculate with and turn into a return. I worry about that trend happening in our environment - right now every platform developer has their own wallet to incentivize their own DIDs and VCs scenarios but people aren't going to want to have 8 of these wallets - people want 1 wallet that works with Sov, Ion, DIDWeb and whatever else that their vendors/issuers want them to use.
Christopher Allen: That makes for a difficult business model for the wallet vendor at the bottom - if you look at my diagram, there's a lot of pieces so it's not like a wallet is an easy thing to build.
Moses Ma: Can someone compile a list of DID-friendly open source, templates and reference wallets that developers can review?
Daniel Buchner: Just so freaking, absurdly open and cheap that it's not even something we consider trying to squeeze $ from
Christopher Allen: You're building 3 layers - trustable layer for security implications, --- i'm concerned.
Kyle Kemper: Agree on all points...listing fees are definitely an issue and creates do we allow tokens and projects to freely list themselves vs paying to play
Kim Hamilton Duffy: So we can work through questions 30 second replies is the going-amount. What we'll do - we'll work through the q in order.
Phil Long: Thank you kim - curious about what the community thinks or what the presenter thinks are the minimum viable functions, and the distinction between OSS and value added services layer that take advantage of the contents of wallets with the consent of the owner.
Dave Longley: Dbuc, just want to make sure we're wary of wallets becoming like facebook wrt data model (that would represent the antithesis of the work here, IMO)
Kyle Kemper: Send, Receive, Request, Transaction History, Recoverability...
Kim Hamilton Duffy: Thank you phil -
Daniel Buchner: Sure, definitely not surveillance wallets
Daniel Buchner: That's a given
Daniel Buchner: I just mean that it should be like Firefox
Kerri Lemoie: I'm doing tech research - dan can answer first as most active wallet. What are the most compelling reasons individuals will adopt wallets.
Kim Hamilton Duffy: Dan if you add to q
Orie Steele: Wanted to respond to adrian's question around password managers - they're the most familiar analogy for storing credentials (in this case logins) - I think structurally if you look at their business models, a lot of them have similar structures to Manu's diagrams with encrypted data vault - they rely on client-side encrypt, they have various different data model structures for storage and that's a point of differentiation - some of them support [CUT]
Orie Steele: Some structure data. A lot of them charge for multi-device replication, I think password managers are an excellent thing to consider when we think about what wallets are and how they integrate with other systems
Manu Sporny: +1 To Orie's point on a good analog.
Anil John: My organization is investing in these things - standardized APIs, goodness of a wallet but what I didn't hear about is the user experience of a wallet. One of the things - we like putting time and pressure behind consent, we are having a competition for user interfaces for wallets. How do you interact with issuer, selective disclosure, presentations - so from UI perspective, focus is on that. Is that something that is going to be of value t[CUT]
Anil John: Confidence in transaction and user-comfort is important. Is there value in putting money where our mouth is.
Kim Hamilton Duffy: I'm excited that you're doing that, if you have comments add to q
Orie Steele: Interesting to note that password manager have "autofill" feature... and CHAPI has a similar functionality for send credentials from the wallet to the websites.
Christopher Allen: I'd like to see a formal information information lifecycle model created for wallets to start a UX requirements
Moses Ma: Again, can the chairs arrange to have someone compile a list of DID-friendly open source, templates and reference wallets that developers can review? My belief aligns with Daniel's... a wallet is a feature, not a product.
Daniel Buchner: I think there's a variety of reasons re: use cases that identity is still a powerful - we don't charge for wallets or snoop on your stuff - we own a few properties like LinkedIn that would benefit from people creating credentials and exchanging them. Adding that level of trust helps those properties - imagine if recruiter tools said for $1 more we can scan trusted profiles that have the right credentials.
Daniel Buchner: That's valuable to microsoft - we don't care about monetize, we just want stuff that happens on top of it. This is why we do this for free and not do the thing that other companies do.
Kim Hamilton Duffy: Thanks daniel -
Manu Sporny: Wanted to add to dan and adrians question - what's the nearest analog? Password managers are very good analogies and that's good, the thing that makes what we're doing here different is the standards. No interop between password managers - can't move information between them. One key thing is the interop standards where you can move contents from one wallet to another without it being a disruption in your work life.
Manu Sporny: So again, not surprisingly standards are the key binding characteristic that makes this different.
Orie Steele: +1 To standards and interoperability
Drummond Reed: +1 To full portability between wallets from different vendors / projects
Manu Sporny: +1 To rocco
Kim Hamilton Duffy: And I have a special callout - we want to thank Rocco for his excellent note-taking -
Ryan Grant: +1
Wayne Chang: +1 Rocco
Heather Vescent: +1 Rocco thanks for scribing!
Markus Sabadello: +1
Mike Prorock: What is the best way to address wallet usage by non technical users? e.g. we work extensively with users that have limited access to tech and sometimes limited literacy rates, but they have smart phones.... have thoughts been put into those types of usability issues?
Drummond Reed: +1
Dave Longley: +1
Christopher Allen: +1 To scribe
Anil John: +1
Phil Long: +1 Rocco
Dmitri Zagidulin: +1 Thanks rocco!
Kim Hamilton Duffy: Thank you for all the speakers for this excellent conversation, thank you and ahve a good day.
Kyle Kemper: +1 Rocco post your doge address
Tim Bouma: +1 Rocco
Moses Ma: Bye peeps
Heather Vescent: Thanks Manu.
Heather Vescent: We have another set of speakers for next week.
Manu Sporny: +1 To that, expect turnout will be about the same, feels like we're just scratching the surface.