Automotive WG

15 Nov 2016

See also: IRC log


Kaz, Paul, Peter, Shinjiro, Adam, Patrick, Song, Kevin, Ted
Paul, Peter


Agenda and Welcoming VW

paul: don't have concrete agenda for today but can talk about the updates
... spec issues, reviews, etc.

kaz: we can briefly check the status and confirm our policy

paul: also Kevin and Adam for spec
... basically, VW has just joined W3C
... the question in Burlingame was that we were working on our Charter
... we have discussed with VW guys
... the Automotive BG is working on new proposals
... creates reports and donates ideas to the WG
... so the approach is working with the BG first
... happy to have another OEM
... that's my summary

ted: VIWI includes not only vehicle information but media information, etc.
... VW was interested in volunteering for the vehicle signal side as well

patrick: from our side, this is a good path to take
... don't want to influence the spec itself at the moment
... VIWI is something the BG could be interested
... we need feedback
... just vehicle information is not enough
... feedback from the BG and the community would make perfect sense

kevin: welcome VW's participation
... tx for joining us
... the RESTful interface is interesting
... working within the BG first would be a good approach

patrick: tx!

peter: really great to have VW on board!

paul: great!

WG Charter Update

paul: btw, I saw comments for the proposed new Charter
... what's the update?

ted: responded to the commenter
... initial misunderstanding was using cloud services from the vehicle using the interface

peter: Rudi made some statement
... found it's kind of defensive
... need to explain the misunderstood point
... didn't see any more responses other than Rudi and Wonsuk

ted: we should be fine

paul: only one objection?

ted: yes

paul: conversation with them?

ted: know the commenter
... will work for the objection
... we should incorporate feedback
... very busy this week but will talk with them

Possible Member Submission for VIWI Proposal

ted: another point is VW's submitting their VIWI proposal to W3C as a Member submission
... so that we can get feedback from the community

patrick: how to do that?

ted: there is a specific procedure
... write a template for the submission
... can help you

patrick: we can do that but is that the most common way?
... would be easy to use the common path
... if it's not only putting on GitHub, we may need some kind of extra agreement within the company

ted: there are two paths: 1. BG report (discussion within the BG) and 2. Member submission

patrick: ok

paul: anything else on this topic?


Issues with the Spec

paul: the next topic is issues with the spec
... JSON schema and WebIDL?

adam: proposal on equivalent way
... machine readable
... maybe we should go with the approach with WebIDL and think about how to apply JSON Schema
... JSON Schema is quite good way

kevin: pros and cons with the both

<Paul> https://github.com/w3c/automotive/issues/99

patrick: there is always object definition
... JSON Schema on our side is the foundation
... could generate human readable WebIDL based on that
... we're generating that ourselves

adam: that's encouraging

paul: makes sense to me

kevin: good idea to change something obvious and see it

paul: what about the spec actions?
... any comments?
... simple implementations?
... testing?
... that's the way to go

paul: have not got many comments for the FPWD yet

kevin: what about the client spec?

paul: Powell is busy for a while

<Paul> https://github.com/w3c/automotive/issues/91

adam: can take an action item for issue 91

paul: there are quite a few action items
... should go through during the next call on Dec. 6
... very late for Adam and Patrick, though
... and we have another action for the client spec and need to ping Powell
... high-level API for Web developers


kaz: regarding testing, Hira-san mentioned he and Urata-san were interested in testing

shinjiro: right
... I'm creating a server prototype for initial testing
... can make contribution for our testing

paul: great

shinjiro: one question is that the W3C testing environment is for usual Web browsers
... on the other hand, the vehicle spec is not for usual browsers

paul: we have a framework for testing servers
... maybe I can provide that
... let me ask Jeff about that
... might help since there is same pattern

shinjiro: great
... also there is test mechanism for Node.js
... maybe that would be useful

paul: Patrick, do you have any mechanism?

patrick: regular Web tools
... we use REST

paul: right
... will talk with our QA team
... may have example tests

patrick: postman automated tests
... continuous integration
... UI tool
... we're doing HTTP and WebSocket is extension with our implementation, though

paul: the basic framework is similar
... would be helpful

peter: what would we test?
... implementation or spec?

paul: basically testing the sections of the spec

peter: ok. we're not testing the implementations

paul: implementations meet the spec

peter: an issue on testing I created on GitHub

paul: issue 75

<Paul> https://github.com/w3c/automotive/issues/75

peter: Urata-san, can you describe what are you planning to do?

shinjiro: about the testing?

peter: yes

shinjiro: creating a prototype implementation first
... and then create test suite
... that's what I want to try

paul: one of the deliverables
... we need two implementations
... the test need to be doable

kaz: we need to be able to point to two implementations of the spec to see implementability and interoperability of the spec, though the implementations themselves are not W3C deliverables

kevin: one thing to see is VSS implementation could be a reference one

paul: test against mach server and test against data generator
... what are you doing with VSS, Urata-san?
... how do you implement it?

shinjiro: I'm creating a VSS server using Node.js
... need some data source as an alternative of the actual vehicle
... can test the implementation using the mocked-up data
... so I have a VSS server and an emulation server

paul: great
... sounds like ACCESS is doing an implementation
... Melco as well, Peter
... VSS as a data model
... people can contribute to implementations
... I would do my best as well
... next meeting on Dec. 6

kaz: regarding testing, I'll ask TV guys about their knowledge about testing environment as well

paul: ok

[ adjourned ]

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.148 (CVS log)
$Date: 2016/11/15 18:35:48 $