W3C

- DRAFT -

Automotive WG

04 Oct 2016

See also: IRC log

Attendees

Present
Kaz_Ashimura, Junichi_Hashimoto, Rudi_Streif, Shinjiro_Urata, Wonsuk_Lee, Tatsuhiko_Hirabayashi
Regrets
Chair
Rudi
Scribe
kaz

Contents


WG Charter status

-> https://www.w3.org/2014/automotive/charter-2016.html proposed Charter

kaz: got approval of W3M
... will work with the W3C Comm Team to make the AC announcement
... now generating some text on the history about the changes of the deliverables
... would like to send this out today

rudi: questions?

(none)

TPAC 2016

-> https://www.w3.org/2016/09/19-20-auto-minutes.html TPAC 2016 minutes

rudi: any update/question?

junichi: we fixed a lot of topics on the service spec

<hira> Could you tell me how to join conf.call?

rudi: also made lot of updates
... updated the repository

wonsuk: regarding VSS, we need to describe attributes for body, chasis, etc.
... need to know what the exact format/style

<hira> I did it!

wonsuk: how to delete children based on the event exchange between the client and the server

rudi: if you use the wild card feature, you need to respond all the tree structure

wonsuk: like body.*

rudi: you should respond everything

wonsuk: we need to clarify what to do
... body.* includes all the information under the tree

rudi: "." notation should be straight forward

urata: we can't tell under what the specified "body" belongs

<rstreif> It would be Attribute.Body.* and not Attribute.children.Body.*

wonduk: how can we make doneNotification?

urata: we can omit children

rudi: "children" is a keyword for JSON
... but not a keyword for the "." notation

wonsuk: ok
... I raised an issue related to WebIDL during the f2f as well
... subscribe VSS
... WebIDL is used to define JavaScript API
... not JSON messages or data itself

rudi: right
... but WebIDL could be used for simple definition of JS methods
... it's rather simple APIs like get/set/subscribe
... WebIDL could be used to define data types

junichi: do you have any candidate other than WebIDL?

wonsuk: JSON Schema
... for JSON messaging
... syntax definition
... we can define JSON message structure
... would be right direction, I think
... what is confusing is that WebIDL has its own data structure

rudi: true
... in VSS, we use simple data types like uint16 which could be simply converted to any programming language

wonsuk: also looking at OCF's work
... they're defining some data structure based on JSON Schema

rudi: maybe would be a right direction

kaz: WoT IG has been also seeking possible data type definition mechanism
... we might want to survey existing examples

urata: there are examples like JSON Schema and Schema.org

wonsuk: some of the W3C specs just use some tables

kaz: some limited vocabulary?

wonsuk: right
... one possible benefit of JSON Schema is possible synchronization
... validation tool for JSON Schema could be applied to our messaging mechanism

rudi: I'm supportive

wonsuk: I can make some examples based on JSON Schema

<scribe> ACTION: wonsuk to generate some examples of JSON Schema usage [recorded in http://www.w3.org/2016/10/05-auto-minutes.html#action01]

<trackbot> Created ACTION-26 - Generate some examples of json schema usage [on WonSuk Lee - due 2016-10-12].

urata: there was discussion on the RESTful APIs
... in the new Charter document, the word "RESTful" is replaced by "advanced protocol"
... as the starting point, WebSocket is enough, I think

rudi: we're focusing on Web browser/runtime
... so this is sufficient for the moment
... and eventually JS library
... applications could be going to use RESTful interface, etc., in the future

urata: it's ok by me to mention only WebSocket and concentrate on it at the moment

wonsuk: yes
... we're concentrating on WebSocket
... but we're open about the possible protocols
... if there is any good approach, we can extend the capability

urata: one of my concerns is security
... when we use other protocols, e.g., HTTP, the security model would be different very much

kaz: so the text on the Charter itself is OK?

urata: ok

kaz: btw, theoretically, WebSocket simply provides a TCP/IP socket capability and we could use HTTP/HTTPS or whatever on that transport mechanism

urata: ok

GENIVI AMM Schedule

urata: btw, another question of Hashimoto-san was the schedule of the Genivi AMM in 2 weeks

junichi: security meeting on Tuesday
... want to confirm the schedule

-> https://www.w3.org/auto/wg/wiki/Main_Page#Oct_19_-_21_.2C_2016_.2CGenivi_AMM.2C_Burlingame GENIVI AMM schedule

kaz: basic schedule:

[[

10/19 - Security Arch

10/20 - Morning - W3C Status to GENIVI

10/20 - WG/BG Meeting - Agenda TBD

10/21 - WB/BG Meeting - Agenda TBD

]]

wonsuk: we need to clarify the details esp. Thu/Fri

kaz: let's continue the discussion by email

wonsuk: btw, who is attending the AMM?

kaz: me

rudi: me

junichi: me

wonsuk: me

hira: me

urata: me
... arriving on Tue.

wonsuk: good :)

urata: wondering about Kevin/Adam

rudi: don't know...

junichi: there was a doodle poll

-> http://doodle.com/poll/g249mvntn26xsdu5 doodle poll on f2f attendance

kaz: the doodle poll doesn't include Kevin or Adam

rudi: will check with them directly

kaz: tx!
... let's continue the discussion by email!

[ adjourned ]

Summary of Action Items

[NEW] ACTION: wonsuk to generate some examples of JSON Schema usage [recorded in http://www.w3.org/2016/10/05-auto-minutes.html#action01]
 

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.144 (CVS log)
$Date: 2016/10/05 01:24:00 $