Automotive Working Group Teleconference

08 Mar 2017


See also: IRC log


Ted, Mike, Hira, Rudi, SongLi, Urata, Paul, Junichi
Powell, Peter


Status updates

Rudi does roll call

Rudi: Ted can you please provide an update on staff resources concerning the group

Ted: Kaz is taking on new responsibilities as part of W3C's reorg plus in a number of other Working Groups so is resigning from Automotive

Rudi: Hira you voiced some concerns about Kaz' participation
... I hope KDDI will continue in this capacity

Hira: this is upsetting to Japanese members but understand the decision
... we would have preferred this be deferred until the spec reaches CR
... Japanese OEM are looking at Smart Device Link and we want to catch their attention as well

Rudi: SDLink is entirely different and understand Toyota is considering that route
... that is more of a competition to other similar approaches for pairing with smartphones than what we are doing

Hira: SDLink has already been published in public and covers vehicle data and media, including security
... this specification may be designed for automotive, smart phone and tablet but we have to notice these activities for our specification

Rudi: we have to monitor what is going in the industry

Hira: we want to compete in this specification race

Rudi: I do not think we are competing with it directly
... we need to work on other domains (media etc) as we have been discussing for next generation/after CR

Urata: can I discuss about the Kaz topic
... Kaz was really a great help for Japanese members as English is a big problem
... I know he is very busy but wondering if we could ask him to help through the F2F
... I understand there are resource allocation matters

<Zakim> ted, you wanted to mention AGL and Japanese OEM

Rudi: that is more a question to W3C

Ted: we cannot provide native speakers for every Working Group for every nationality we have participants. I encourage anyone who has difficulty understanding any comment to ask by speaking up or in irc

Ted: we only have KDDI, Melco and ACCESS as participants. It would be great to have Toyota and other Japanese OEM (and Tier 1s) involved. Hopefully AGL's involvement will help with Toyota and others in time

Rudi: agree to Ted's point, if things are unclear please ask for clarification

Hira: can we ask him to join for WoT concerns?

Ted: for joint meetings with WoT like TPAC then yes we would reach out to him, the WoT lead Dave Raggett and their chairs

Urata: I have made some changes to my prototype implementation in github
... the access control portion is currently empty and cannot test the authorization piece
... I can implement the authorize method with a dummy function
... this is a mock implementation but it is achievable. I will continue this week and next
... the test cases will be complete by next week
... I have only used my test suite against my prototype implementation and ask for others to run it against theirs

Rudi: I really appreciate your effort and moving this forward swiftly
... I agree for the purpose of testing and especially since we deferred on the authorization piece you approach is reasonable
... perhaps in the next round (next gen) it would be more explicit but it is left to implementers at this point

Urata: I tried to put the test case on public server so people can run the framework themselves
... from the description of the test suite it is suppose to execute locally and it is not possible to run online
... I ask people download and run themselves and run themselves

Rudi: I will ask Peter Winzell from Melco to run the test suite against his implementation that is running on the Genivi Development Platform
... SongLi perhaps you are interested in running it as well

SongLi: absolutely

<urata_access> https://github.com/aShinjiroUrata/web-platform-tests/tree/dev-urata-vsss-test

Ted: also worth asking Powell to run it since he demoed VinLi's platform with VISS at Genivi showcase during CES
... if it is at all possible to run online instead of locally with /etc/hosts to have our static hostname, then I would be willing to work with you (Urata) to run an instance at MIT

SongLi: having an online tester would be beneficial
... we can skip certificates as well if they are a problem

Ted: benefit of an online test serveris it provides and opportunity for soliciting implementation reports
... I will also encourage those in the group to do short write ups of their experiences in implementing

Rudi: Powell has been working on the client side specification but there is still a gap there
... he is not here but perhaps others may be able to comment

Urata: I am in a discussion with Rudi-san on authorization
... the authorize method according to the specification should be done for data path and actions
... we may want to specify actions as a parameter to authorization
... this piece is not clear enough to me yet
... leaving some things to be implementation dependent is ok but we should confirm we can realize a system to authorize properly with json web token or oauth

Issue 140

Hira: I have raised a question on 137

Issue 137

Hira: I want to add normative or non-normative by section on his generated document

Rudi: that could make sense, please suggest which you feel should be normative in that issue would help and we can then discuss further
... it will help to learn where we should be more specific

Hira: we may need this clearly defined in order to be able to make test assertions

Rudi: yes or it could be left to the implementer

Urata: the VIAS seems agnostic in the manner it communicates
... it would be worth saying web sockets or http
... it has an assumption of using web sockets which is different from our initial intent
... in the future it should be modified

Hira: next time we modify specification for high level api we should not touch the draft spec

Urata: changing the draft VIAS now isn't necessary

Hira: we would need to clarify normative and non-normative

Paul: is the concern about the VIAS spec being written to a socket?

Urata: if the group thinks VIAS should be written in a more agnostic manner we should remove socket from the spec

Paul: agree it should be informative and not normative
... we said earlier that the developer shouldn't care how VIAS talks to the service
... we do want the two specs to be consistent and compatible
... the client spec is meant to stand on its own so it should be non-normative on how it communicates

Planning for Specification Teleconferences in March

Rudi: we had agreed we wanted to have some longer teleconferences to complete this work on the specs and test suite
... the question is to start this week or next?
... there will be timezone conflicts as we are well aware

Ted: for me next week would be better as the following I will be in Chicago for coordinating Automotive and Web Payments at their F2F

Hira: can we have a doodle?

Rudi: sure

Urata: I would like to continue the discussion on VIAS and hope we can have a test implementation

Rudi: yes that would be necessary and hope something that Powell is working on

Urata: tests are necessary for CR?

Rudi: yes

Hira: I will generate test assertions in March

Planning for GENIVI AMM in May

Ted: Genivi has confirmed space during their All Member Meeting for the 10th and 11th
... I'll send email to the group 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: 2017/03/08 02:38:30 $