Mini f2f of Automotive WG/BG at CES

05 Jan 2017


See also: IRC log


Ted, Kaz, Hira, Mike, Paul, PatrickL, SongLi, Rudi, Wonsuk, Powell, Urata
Peter, PatrickB
Rudi, Paul
ted, kaz


<ted> scribenick: ted

<scribe> scribenick: ted

Plan for next specification round including VW ViWi

PatrickL: I will let you know what PatrickB has been up to. He has an open source test/mock server people can use
... we have not been able to do extensive testing yet
... we are not sure what to use for testing the web socket subscription part

-> https://github.com/wzr1337/viwiServer Mock server code repo

-> https://lists.w3.org/Archives/Member/member-automotive/2017Jan/0021.html PatrickB's note

Ted: Peter or Kevin might have some ideas for testing the socket piece

Paul: we have some code (at OpenCar) for testing web sockets and think Urata-san said W3C's test framework can do some rudimentary web socket testing

PatrickL: we haven't found anything that would let us put in a test suite to go against web sockets

Paul: I'll strip out the proprietary pieces and make our socket tester available on github and have received permission to do so

Kaz: We also need such a testing framework for VISS as well and think Urata is working on such a module

Paul: that is my recollection as well

<inserted> [ Note that the title of the server spec has been changed to "Vehicle Information Service Spec" (VISS). See also: https://github.com/w3c/automotive/pull/114 and https://www.w3.org/2016/12/20-auto-minutes.html#item02 ]

Hira: Urata has already made a test module prototype

PatrickL: We have spent some more time looking at VSS as well
... when comparing both we have to first have a clear picture of the use cases. Both have their advantages but first be clear on needs and what we are trying to solve

<kaz> ted: @@@explanation on TAG, HTTP2, etc@@@

Ted asks if VW has looked at VSS at Genivi that VISS refers to

<kaz> VISS

<kaz> VSS

PatrickL: it is a good model for signals data but maybe trying to get a full view of state of vehicle might be a bit counter intuitive
... my opinion does not really matter and it would be best to have input from developers

OCF Demo

<kaz> demo video

Paul: Wonsuk was telling me on the elevator about the OCF demo

Ted: Rudi gave the youtube link and broadcast it on Tuesday's call

Wonsuk: demo will be running the next 4 days, provides a digital dashboard
... we have complimentary apps for iOS and android. we have a OCF server providing VISS to complient devices
... we have a smart watch device running on Tizen
... the android app can track a vehicle real time
... Sanjeev has a newer video which I'll run during the showcase in a couple hours

<kaz> scribenick: kaz

ted: a task force working on use cases, comparisons, seek developer and architect input are good steps. i'll start a wiki of notes and send mail next week

<kaz> scribenick: ted

Paul: one thing that came up in the BG as a result of ViWi being made public was it peaked AGL interest
... we will be meeting with them here tonight
... including them getting more involved

Ted: we spoke with them in the past, hoping to engage them. all were in agreement but it didn't get started

Paul: @@a is trying to go both Android + HTML5. @@b/@@c have similar approaches and would be good to get them on board
... @@d has a program that people can use a live vehicle and their app provided they are a registered developer and have a vin number

Ted: I also heard but cannot confirm that SDLink might use VSS


Paul: I know they are web sockets, Google isn't turning up anything on them and VSS

Kaz: as mentioned in Burlingame, Web of Things WG might be worth talking to

Ted: they are doing things fairly differently than either of our approaches. I can see their work sitting on top of ours much like Sanjeev is already doing in OCF

Rudi: Sanjeev has been doing this within OCF+Genivi
... the OCF bridge provides VSS, and makes the vehicle an OCF object
... that brings web expertise, IoT and automotive (Genivi) together

Ted: Alan and Dave are going to be meeting with OCF this week as there is interest from both sides in better coordination
... I think we should do our own approach and let WoT do the same as OCF and we can focus on bridging

Kaz: I'm not suggesting the Automotive group and VIWI need to use the WoT approach.
... However, WoT is discussing interface between server and client for IoT, and there is some overlap.
... Also they've been working with OCF.
... So discussion with WoT guys would be useful.
... I think I should join the meeting with OCF this week as well.

Paul: new charter isn't on wiki, shouldn't we update it?

Ted: yep

Updated Charter

WG Wiki

Discussion on Chaigning the spec name

[discussion on changing the spec name]

<kaz> previous discussion

<kaz> Server Spec

<kaz> Kevin's issue

Rudi: wonders if we need to chang the spec names on the Charter as well

Kaz: we can change the title [on the spec]
... we don't have to update the charter itself

Paul: back to the timeline, we hit our initial milestone for VSSS FPWD, next is CR in April

Rudi reads definition of Candidate Rec from Process document

<kaz> Process Doc

Client spec

Paul: what about the client spec?

Ted: I suggested Powell pause as I believe Visteon is joining and interested in that piece

Paul: Powell and Vin.li will be demoing based on their approach and hopefully we will learn more about others' interest soon

Powell: server spec changes will dramatically impact the client one. I had to implement a test server in order to work on the JS library

Paul: should we start cataloging issues?

<kaz> +1

Powell: I am in my notes. Request ID will make sense
... I'll throw some up on issues list

Songli: mentions running into some similar problems

<kaz> scribenick: kaz

paul: do we want to have a companion document?

powell: someone should go into the spec
... how to handle ID
... we don't need to add anything but should clarify how to deal with the server spec

<ted> scribenick: ted

Songli: we can make either a simple rudimentary server or a more complicated nuanced one
... we may come up with some best practices for implementers based on our experiences

Paul: I encourage you both to start submitting issues based on what you are seeing

<kaz> +1 to start with some guideline/best practice/primer kind of document

Powell: everything pretty much works

Paul: if we can keep to the timeline for CR in April then we can announce it at the next Genivi AMM
... next milestone after that is Proposed Recommendation
... three months of handling comments

<kaz> scribenick: kaz

paul: Viston is joining and interested in the spec
... so we should hold on the work on VIAS (client spec) for a while
... we clearly will have the CR in April

powell: two different approaches on notation, JS vs WebIDL

paul: VIAS has very specific scope and we should stuck with that

rudi: agree
... let's get moving forward
... don't see mutual exclucivity between JS and WebIDL

powell: what is the expectation for implementations?
... mapping messages defined by the server spec to JS?

rudi: good as the starting point

paul: guessing what Visteon is doing is different from our definition but we'll see

Updated Charter

VIAS Guidelines

paul: I'm fine with WebIDL
... we should put our proposals on GitHub

powell: will do

paul: anybody who can help Powell, please speak up

Testing framework

paul: next, Urata-san, do you want to share information about Testing?

urata: not much progress since the last meeting

-> https://www.w3.org/2017/01/03-auto-minutes.html Jan 3 minutes

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

<urata_access> this is the starting point of creating test framework

<urata_access> have some more test cases in my local environment

<urata_access> going to add test cases according to the test assertion list

Meeting plan

paul: anything else?

rudi: meeting plan?

-> https://www.genivi.org/ GENIVI AMM on 9-12 May 2017 in Birmingham UK

wonsuk: we need to send emails to the group

kaz: sorry but need clarification for the minutes
... are we talking about the collocated meeting with GENIVI AMM?

paul: two topics here, (1) collocated meeting and (2) VIWI as a big topic for the BG

GitHub isssues on Testing

hira: one proposal
... I've made a proposal on Implementation Report Plan
... would like to have several issues on GitHub
... want to have separate issues for 4 topics

paul: yes, you should

kaz: yes, if it's easier to handle your issues, you should create those separate issues

paul: +1

rudi: have good press conference and demos at CES!

[ 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: 2017/01/06 17:56:58 $