W3C

– DRAFT –
Web Authentication WG

09 February 2022

Attendees

Present
agl, akshay, andre, davidturner, davidwaite, dveditz, jdeng, jeffh, jfontana, johnbradley, jpascoe, kenbuchanan, martinkreichgauer, nadalin, nicksteele, nsteele, raerivera, sbweeden, timcappalli, wseltzer
Regrets
-
Chair
-
Scribe
jfontana

Meeting minutes

* thank you

<wseltzer> nadalin: Week of June 6, RSA, considering a Thurs/Fri F2F

tony: Tim will take on finding meeting spot in SF

tony: nothing new on the charter
… open pull requests.

https://github.com/w3c/webauthn/pull/1663

JeffH: added verficaiton, need some responses

https://github.com/w3c/webauthn/pull/1576

jeffH: conditional UI. nominally ready to go. waiting on review. it is feature complete
… still waiting for reviews, but comments are pending.

tony: have some triage

https://github.com/w3c/webauthn/pull/1695

timC: this addresses some feedback coming in on multi-device credentials
… some assurance the cred is backed up
… bit 3 is static, bit 4 is can change

akshay: we will soon have back-up.

timC: soon flag still requires a check for backup

agl: if msft does not need that 4th state can it go?

shane: less states is better

jBradley: we need to be specific in the validation.

shane 7.2 is where we should look.

timC: this is meant to guide users
… it is not designating what the key can do

jbradely: we could have another approach
… could force autheticators to have multiple AAGUIDs.

shane: an appealing patten for RPs, provides options

akshay: looking if it can be backed up or not

timc: should we be more explict that its not a security property

shane: it is something RPs want to know

timC: not just on the RP side, the user can decide what to do

jeffH: what is the model we are talking about
… bit 3 is it capable of being a single stautus
… yes

jbradley: imagine RPs will make diff decisions depending on what comes back.

nickS: think vast majority of RPs will not evaluate this.

jbradley: that could be true

timC: I will work on this in the section
… 7.2

tony: we have a couple of un-triaged issues.

https://github.com/w3c/webauthn/issues/1694

jeffH: M.jones said he would do this.

https://github.com/w3c/webauthn/issues/1697

jbradley: appears safari on ios and OSx there are attestation issues
… outcome is you don't get a credential made.
… cold be issues for RPs

agl: that reflects my understanding. is this an apple bug

?

jbradley: were we too vague in the spec

?
… need to be explicit
… in CTAP can't ask for an attestation type

<dwaite> WebKit issue @ https://bugs.webkit.org/show_bug.cgi?id=224042

agl: perhaps add wording.

jbradley: OK. I will work on this and talk to apple

jbradley: I will write the PR

https://bugs.webkit.org/show_bug.cgi?id=224042

jbradley: not planning on sync for non-discoverable creds

agl: that is right

agl: our guidance will be to request discoverable creds.

https://github.com/w3c/webauthn/issues/1696

agl: this is deleting

tony: yes

jbradley: we have said no to this in the past

tony: any new insight?

agl: new API for deletion; prescriptive

agl: having channel at RP to send something back.

timC: concepts are there, but we should be open to some of this

agl: don't want users to get locked out.

Minutes manually created (not a transcript), formatted by scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).

Diagnostics

Succeeded: s/still//

No scribenick or scribe found. Guessed: jfontana

Maybe present: jbradely, jBradley, nickS, shane, timC, tony