Automotive Working Group Teleconference

04 Apr 2017


See also: IRC log


Wonsuk, Paul, Ted, Urata, Mike, Rudi, Junichi


<scribe> scribenick: ted

Status update

Paul: Urata can you catch us up on VIAS?

Urata: Wonsuk and Adam have taken care of most of the VISS issues

Wonsuk: i introduced and issue and a pull request

issue 171

Wonsuk: i wanted to remove optional part of the filter property

related pull request

<paul> https://github.com/w3c/automotive/pull/173

Urata: the change you made has path optional
... in getVSS path was mandatory

Rudi: does this mean we would return the whole tree, null...

<urata_access> https://w3c.github.io/automotive/vehicle_data/vehicle_information_service.html#idl-def-authorizesuccessresponse

Wonsuk: we want to keep the filters in getVSS as mandatory instead of optional

Urata: in the existing webIDL definition it already has optional path

Rudi: if it is optional and you call getVSS without then what? we simply need to explain what the behavior would be in that case

<paul> Tried to go in as host, but it looks like you need MIT creds

Wonsuk: if client side wants to use path property the can
... I'll handle that and make clear on the issues list

Urata: there are still a number of open issues to review

<urata_access> https://github.com/w3c/automotive/issues/146

Urata: this one concerning tokens isn't resolved. we should be consistent with "authenticate" to "authorize"

(in VIAS & VISS)

<paul> Access tokens as proof of authentication

<paul> https://oauth.net/articles/authentication/

Urata: Rudi responded with access token

Paul: that is what oauth says

Rudi: we should just decide and move on

Urata: in some cases the VISS server won't be doing the authorization but another server

Ted: how about a straw poll and record in minutes and on issue as a comment and consider it a proposal. people not on call have a period to object or it carries

Paul: there is very nuanced semantics as Urata's use case demonstrates

Urata: the opinions on the thread are very diversified

Results: leaning on straw poll is to go with "authorize"

<urata_access> https://github.com/w3c/automotive/issues/156

Urata: there is an opinion that VIAS could abstract itself away from web socket

Paul: my understanding was initially we wanted this higher level API to resemble the earlier WebIDL spec
... that has shifted
... we can move it out later after CR
... and defer on decision for now. it can be moved into an appendix
... that was Hira's suggestion, to have VIAS as an appendix later

Ted: it is already spelled out as a separate entity in the charter so probably should stay that way

Wonsuk: we can have two variations or really try to figure out how to make a more convenient api for developers
... one way to simplify is to poll the WG participants to see which they are planning on implementing

Rudi: we should try to deliver both at the same time

Paul: they can move as separate entities, having the service available sooner than the high level

Ted: let's keep VISS on schedule, these minutes and email from one of the chairs for people to review the issues they raised or respond to try to wrap them up with final suggestions


Ted: I'll reply to Rudi's email on this with WG meeting suggestions, additional details

Next process to CR

Ted: basically encouraging closure on issues, then one of the chairs will send a last call

Genivi AMM Demos

Paul: send suggestions for demos to the list

Urata: should we decide some concrete date to go to CR?

Ted suggests deadline for issue

Urata suggestions the 18th call

Paul: I'll send to the list

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2017/04/05 01:07:16 $