TL: Last time we discussed this we ran out of time.

DKA: We need to review their issue 54? All of their issues still open?

TL: Yes. Let me dig out the minutes of the discussion. I believe we should look at 54. We had some feedback on how this report should be formatted - especially on JSON serialization. Mike addressed those questions in 53 - but I need to look at this.

DKA: Was this related to something IETF did?

TL: I believe so

AR: Structured headers should be something that IETF does, no? Is this structured? This has it's own type system.

DKA: You mean IETF's work?

AR: Yes. We should wait for Mark to comment on this.

TL: I think the point we should be looking into is if the IETF spec is good enough for the reporting requirement.

AR: (Something about JSON, choppy audio.)

Action: Leave note on their issue, potentially invite to future call? Revisit 5/22.

Issue 185 - WebVR - Peter, Travis, Hadley, Alex

TL: Haven't looked at it yet. Should invite them to a call.

AR: Should we reach out?

TL: Yes.

DKA: How about the 29th?

TL: SGTM.

DKA: Can someone send an invite?

TL: I can take care of this.

Action: Send WebVR folks an invite for a call on 5/29.

Issue 134 - AOM - Travis, Sangwhan

Action: Sangwhan to send Alice invitation for future call

Issue 225 - Permission Delegation - Lukasz, Dan, David

AR: Status - this is delayed - time to sort out - in Chromium for certain users, changing whether users see a paired requests - warmings about subframe requests. I discussedd (with the proposer) what happens in lower levels.

DKA: Any other implementors working on similar features?

AR: No idea.

DB: Some people at Moz who saw this proposal liked it. One piece of feedback - list of permissions in the API spec is the list of permissions that are requested. There are also permissions that are implicitly granted throuhg a user action. So a question is - if you are talking about ability to delegate permissions do we need a list of those permissions as well?

TL: in some of those cases are they all tied to user action? you'd be delegating without the need for user interaciton?

DB: The reason the spec should be able to control that is that the user might e confused [otherwise]. We have a model where you see a file picker or media picker and there is an association with a tab, but not with an origin. If we're moving to a model where we are associating witha top level page than maybe the top level should be able to restrict.

AR: we don't have a way to restrict. I can have a [chrome] conversation about what data they do have to inform this...

Issue 231 - Payment Handler - Hadley, David, Alex

DKA: What is the status on this? David left a comment.

DB: I think they opened a bunch of issues on their side, should check what the current progress is.

DKA: We should review the current work.

DB: They linked 5 issues, of which one has been closed.

DKA: We need to review the response and see if this can be closed.

DB: I'll get to it.

Action: Revisit 5/22.

AOB

DKA: Next week is the AC meeting. Should we have a call?

PL: I can chair.

(Others expressing availability. Will have call.)

@torgo: Lots of issues were opened up. All still open. @travisleithead what is your current view on this? A bit of comments on some of the issues

@travisleithead: We want to grab these people at the F2F or invite them.

@torgo: Who can bring these people together, should we have a focused call

@travisleithead: will be hard to bring all together time zone wise

@slightlylate there is currently an intent to ship, but no permission yet, and open privacy issues etc - TAG is still waiting for resolution on issues.

@travisleithead: Dropping a note in the issue.

@torgo: Bumped to 29th to see how we want to proceed. Is this time sensitive?

@slightlylate: will give feedback to team that they should act on the feedback when they are doing an intent to ship. I will reach out to them.

@travisleithead: I posted to blink-dev, might need to be approved

TL: Closed. To use JSON.

DKA: Close and move to F2F?

TL: Sounds good. It didn't look like a great fit for structured headers, so not much we can do here