DID Working Group Telco — Minutes
Date: 2020-05-12
See also the Agenda and the IRC Log
Attendees
Present: Daniel Burnett, Ivan Herman, Yancy Ribbens, Justin Richer, Amy Guy, Markus Sabadello, Chris Winczewski, Eugeniu Rusu, Paul Madsen, Jonathan Holt, Orie Steele, Kyle Den Hartog, Drummond Reed, Michael Jones, Tim Cappalli, Kaliya Young, Manu Sporny, Dave Longley, Rouven Heck, Ganesh Annan, Joel Hartshorn, Dmitri Zagidulin, oliver terbu, Kim Duffy, Brent Zundel, Thomas Schwarz
Regrets:
Guests: Adrian Gropper,
Chair: Brent Zundel, Daniel Burnett
Scribe(s): Yancy Ribbens
Content:
- 1. Agenda Review, Introductions, Re-introductions
- 2. Next Topic Call
- 3. Working Draft vs CR
- 4. Working Sessions - Poll
- 5. Status of the spec
- 6. Core issue status check
1. Agenda Review, Introductions, Re-introductions
Daniel Burnett: check on working sessions
… majority of call time will be on status check
… anyone not yet introdcued themwelves
2. Next Topic Call
Justin Richer: looking forward to seeing what this tach can do
Daniel Burnett: https://github.com/w3c/did-core-registries/issues
Daniel Burnett: the plan for this call is the core registries
Amy Guy: hey burn, I juuuust sent an email about some topics for that to the list
Daniel Burnett: we thought it PRs time to dedicate a call to the registries
… there is a special topic call on did core registries
… have not seen that myself
Rouven Heck: works with consensus
3. Working Draft vs CR
Daniel Burnett: working draft vs recommendation
… we are in the working draft stage
… every document that goes out has boiler plate at the top
… we are not at the point of consensus
… everything doesn’t need to be perfect yet
… there will be a time when we need completeness
… just bringing this up because we have seen some PRs that could go in
Amy Guy: +1 incremental improvements
Daniel Burnett: if you see that suggestion from manu and the others that’s why
… any question about working draft vs CR
Daniel Burnett: if you have examples for JSON and CBOR formats, we need to get it in
… something important is that these formats are not there
… if they aren’t in they are going out
Jonathan Holt: just put a PR in
… is it master or github pages to base against?
Daniel Burnett: master
Kyle Den Hartog: does the assertion need to go in before that point
… like key agreement and all the key buckets and all the authentication
Daniel Buchner: He’s talking about all the random strings that mean nothing in the spec
Daniel Burnett: this is getting to what must the spec have for it to be functional
Manu Sporny: short answer is yes
… if it’s not defined anywhere..
… there’s an extension that allows us to add it after the fact
… we still need to have that discussion about key agreement and that stuff
… this is a call for getting anything in
Daniel Burnett: there are a large number of issues to work through
Daniel Buchner: Can getting things in be proposals to take most out?
Orie Steele: keyAgreement != only way to handle encryption to date….. if you care about encryption, consider assisting with the spec
Daniel Buchner: half serious
Daniel Burnett: in general arguing to remove sections of the spec that have support to date is not a good approach to get the spec done
… that looks like an attempt to delay the spec
4. Working Sessions - Poll
Daniel Burnett: there are a number of things that need to be discussed
… we might benefit from a number of topic calls
Daniel Buchner: Keys and endpoints, keys and endpoints, my kingdom for a simple JSON doc with keys and endpoints
Daniel Burnett: we want to let you know that we need to send out a doodle poll about what week will be a good one
Drummond Reed: +1 to some extended calls. A virtual F2F.
Manu Sporny: I wanted to try to limit the calls to 90 minutes each
… attention and arguments go up after that
… and then leaving a little bit of time
… between each one
… back to back calls doesn’t give people enough time to internalize things
Daniel Burnett: I know what you mean about the 90 minutes
… this is not f2f but instead remote
Manu Sporny: I’m fine w/ burn’s counter-arguments – defer to Chairs.
Daniel Burnett: as you discovered with face to face in particular, sometimes you need to alternate topics
… that’s the intent is to find time when we can do this
… we’re just getting nervous that if things stack up it could be hard to make progress with 1 hour per week
… expect a doodle poll
5. Status of the spec
Daniel Burnett: we want another update on the status of the spec
… thought it would be good to have another update to see if we’re close to horizontal review
Manu Sporny: the only thing we’re waiting on is section 3
… we’ve noticed a number of new PRs coming in
… saw Mike Jones put in 3
… and I think Amy has done some work
Drummond Reed: Amy has done some fine work
… trying to make up for it this weekend
… amy’s approach and my approach is different and need to produce a draft of the two
… and then put into PR’s
Michael Jones: there’s still a bunch of stuff that’s not self-contained
… for instance there’s a bunch of identifiers who’s syntax are not defined
Daniel Burnett: it is important for a horizontal review to not wait
… for the first one absolutely do not wait
Ivan Herman: I agree
… internationalization might have some issues
… some of the issues are relevant that mike is raising
… for others the document should be done as soon as it’s done and reasonable
Daniel Burnett: all we are waiting for is a document that is reasonably complete
… we are not ignoring it at all
6. Core issue status check
Daniel Burnett: https://github.com/w3c/did-core/issues?q=is%3Aissue+is%3Aopen+sort%3Aupdated-asc
Daniel Burnett: for those of you joining us for the first time, we are reviewing the issues in order
Daniel Burnett: https://github.com/w3c/did-core/issues/170
Daniel Burnett: the first issue I see is 170 assigned to Mike
Michael Jones: waiting on getting JWKs into the spec
… where do we stand with that?
Daniel Burnett: who is writing JWK examples?
Orie Steele: Ive been writing most JWK examples
… this issue is about the difference between key expressions
… I think we have determined that if you want to express formats not in JWK you can’t use JWLs
… we need more examples
… . this one is about the id and type property not represented as JWKS
Manu Sporny: to answer mikes question, there are things that need to be reconciled
Manu Sporny: https://github.com/w3c/did-core/issues/240#issuecomment-627357903
Manu Sporny: I added a comment here two hours ago about considerations around JWKs to make the security attack surface smaller
… this is coming back to we are going to have a special call
… and we said we’d deal with them in the issues
… so we need to have another special topic call
… and once we have alignment there it will be easy
… writing the examples will be come easier
Michael Jones: I also thought there would be a special call
… I noticed there was a pub key hex
… and there is some dangling stuff we need to remove
Daniel Burnett: we plan to continue the topic calls
… there are a number of these issues that are still hanging around
… this one then is waiting on a special topic call to review
Orie Steele: bitcoin and ethereum community use publicKeyHex… its likely to get added… because choosing not to will be seen as extremely excluding to those communities, especially to consensys
Daniel Burnett: https://github.com/w3c/did-core/issues/36
Drummond Reed: this is waiting on Markus and I to update the abnf
… it’s still an open issue
… I still think we need to close on that
… once we update the abnf
… we need a proposal on what that prefix will be
Kyle Den Hartog: i’ve realized the extensions will handle these types of things
… lets figure out how to prefix it
Daniel Burnett: https://github.com/w3c/did-core/issues/95
Daniel Burnett: we haven’t discussed this
… kyle you pointed out that it’s not necessary
… justin can you look at Kyles comment
Kyle Den Hartog: +1 to justin
Daniel Burnett: when you put this in in November there where pieces missing
Kyle Den Hartog: around if we get the contract resolution in place then we can close
Daniel Burnett: we can leave it open until the resolution contract stuff goes in
Michael Jones: I think that breaking this into two specs would make it harder for developers to find things they need
Justin Richer: we are not going to split it
Manu Sporny: +1 to what selfissued just said.
Manu Sporny: +1 to not splitting it.
Dave Longley: +1
Orie Steele: +1 to selfissued comment
Daniel Burnett: I think it’s a bad idea to split
Daniel Burnett: https://github.com/w3c/did-core/issues/33
Dmitri Zagidulin: there’s a few things
… i’m not sure we have full group consensus
… I think some people have proposed same as property
… we would like the ability to say
… that the subject has several dids on several methods
… this DID document is deprecated or expired
Kyle Den Hartog: I think this sis an example of how we can use the registries
Drummond Reed: sorry its ironic that this is the synonym issue
… its a powerful one but it’s complex
… I think there’s increasing need for this feature but if we are to tackle it it will need some work
Daniel Burnett: if you’re opposed to defining the function please put that comment in
… will schedule time if we need to
… we need people working offline as well
… we only have a few people giving comments here
… . they need to do it in github before we schedule another call
Daniel Burnett: https://github.com/w3c/did-core/issues/55
Oliver Terbu: about allowing eth address
Orie Steele: https://w3c.github.io/did-core-registries/#ethereumaddress
Oliver Terbu: if someone could please summarize what needs to be added to get this solved
Orie Steele: I added eth address to did core registries
… there is a place where it’s a subject of the special topics call
… I agree with Oliver
… its been frustrating
… it feels like we can’t get properties registered in any way
Dave Longley: +1 to being inclusive
Orie Steele: we will address this on a special topic call
Manu Sporny: want to support what Orie said
… unless someone objects it will be easy to address
Michael Jones: what is being proposed is it’s requesting to register a hash of a key
… it doesn’t contain the key
Oliver Terbu: it is not a key representation
… my understanding is the did spec allows for other things
… how we are going to use tat for authentication
Kyle Den Hartog: +1 to allowing verificationMethods other than keys
Oliver Terbu: is for getting a signed message
… its something that the eth community does and lives with
Orie Steele: See https://identity.foundation/EcdsaSecp256k1RecoverySignature2020/#ES256K-R
selfissed: verification method is different than a public key format
Orie Steele: its registered as “EcdsaSecp256k1RecoveryMethod2020”
Michael Jones: i’ll add that comment to the issue
Daniel Burnett: i’d like for that to occur on Github
Daniel Burnett: is there anything else you want to add for this item Oliver?
… next item
Oliver Terbu: no, thanks
Daniel Burnett: https://github.com/w3c/did-core/issues/9
Michael Jones: if its true there are no references to RSA signatures we can close this
Daniel Burnett: editors will propose it be closed
Manu Sporny: just to be clear 2018 still exists in the spec
… we may transition to ed25519 verification key
Daniel Burnett: https://github.com/w3c/did-core/issues/163
Drummond Reed: this is one of those issues about how to divide up how to do the hyperlinking
… how we should format property names
Amy Guy: burn, drummond, feel free to assign me to do it when we are closer to CR
Daniel Burnett: this is just an editorial thing
… if anyone has opinions they wish to share
Daniel Burnett: https://github.com/w3c/did-core/issues/137
markus sbadello: the issue is independent of syntax
… even if we don’t use matrix params for representing DID params
… if those are defined normatively in the spec
… I thknk what I would suggest is to rename the issue
Manu Sporny: +1 to Markus’ suggestion.
markus sbadello: then continue discussion and make a decision on that
Drummond Reed: +1
Manu Sporny: It really is about DID Parameters.
Manu Sporny: and what we define there.
Daniel Burnett: if you want to change the title go ahead and make a suggestion in github about why you want to keep it open
Orie Steele: Im in favor of removing things that are not well defined…. and only adding definitions for things that have strong consensus.
Manu Sporny: +1 Editors can discuss on next call.
Daniel Burnett: I would like to ask the spec editors and to decide if there’s a merging or collapsing of the issue
… we are out of time for issues today