Automotive Working Group Teleconference

Rudi: still 10 issues that need to be addressed for VISS
... plenty of discussion but some lacking resolution
... we need our editors to see if they can wrap them up so we can bring the doc to CR
... Urata, from your testing can you say that from your perspective it works?

Urata: basically the remaining issues are very detailed and not major obstacles

Rudi: we want to publish VIAS as FPWD. there are not too many issues

Urata: I have been focused on VIAS the last two weeks, solved 5 github issues
... you can see them from the closed issues page

https://github.com/w3c/automotive/issues?q=is%3Aissue+is%3Aclosed

Urata: there are two remaining issues for VIAS at present
... one is rather big, write WebIDL for VIAS
... I haven't received review for my change but merged it myself

<urata_access> https://w3c.github.io/automotive/vehicle_information_api/vehicle_information_api_specification.html

Urata: I added WebIDL definition on this document
... Hashimoto commented on the security and privacy considerations
... there are some other minor issues that have been addressed
... another rather big github issue is 156

https://github.com/w3c/automotive/issues/156

[[Whether VIAS is high level API or Websocket based APIs?]]

Urata: if this as a high level API then it should be more abstract
... Rudi's comment was that VIAS is a counterpart to VISS and should be described concretely, using web socket

Gunnar: agree we do not need to mention explicitly the Javascript library and that is consistent with your WebIDL definitions

Urata: if it is a library definition then it would be considered an application

Peter: I do not see how it would be defined as an application
... a library is not an application
... I agree it should not be defined as a JS library

Urata: yes, it can be implemented as a JS library but the spec does not define how it should be implemented
... we should remove the phrases in the spec that call it a JS library

Gunnar: isn't it possible to check what is normally done in W3C?

Peter: usually it would be WebIDL at W3C but we are also not the usual group

Rudi: correct, it would be a definition of an extension or to be implemented directly in a browser. we have not received support from browser vendors and took a different route

(service/sockets)

Rudi: agree we should remove JS library and I am OK with removing the direct correlation with VISS and sockets
... a VIAS compliant client would most likely need to implement against VISS but it does not need to be explicit
... presently we only have 1 server definition, VISS

Urata: in the current VIAS document there are connect and disconnect methods
... if the underlying connection is something other than web sockets it might be a stateless service without a persistent connection

Rudi: we can clarify connection-less protocol option

Urata: that is a good solution
... I will address 156
... we also have issue 164

https://github.com/w3c/automotive/issues/164

Urata: VIAS already has this in mind
... publishing FPWD should be possible after addressing the remaining issues

[discussion on VIAS implementations. most likely will be a JS library but can also be done directly in a browser or browser extension]

Urata: my concern is whether VIAS will be accepted as a W3C specification

Rudi: we will find out when we get to FPWD what the feedback will be from the broader community, receive their comments and pushback
... I wanted to touch base on demos at AMM

Ted: we have a table and demos planned by Peter and Wonsuk
... I would like to see Urata demo too if he is interested
... I will have a W3C banner of sorts and some printed materials on W3C. if you are doing demos you may want to have some materials as well for possible follow up

Urata: I have a demo but the UI is a bit rough and the data is not realistic
... if it is necessary I can bring it but think Peter-san's demo will be better

Ted: do not feel pressured but encouraged :)

Peter: I think the test suite is very useful as it is
... we will be opening OpenDS for signals
... hopefully I can have that available in time, some other priorities are competing for resources

Gunnar: keep in touch, we might be able to find some resources to help fix it up

Peter: hope is to run this on Genivi development platform and not on a PC
... on our github there is a report in progress

Gunnar: the normal showcase is only one day, Genivi development platform will be open the entire week