Automotive Working Group Teleconference

Paul reviews agenda

VISS comments/issue

Ted: no new issues, objects nor comments to call for consensus by Friday deadline
... I am preparing document for pubication and will keep group appraised
... originally I hoped to get it published tomorrow

Paul: where would we see comments?

Ted: in the spec we steer people towards github issues

Gunnar: although Rudi's CfC accepts silence I wanted to give verbal support
... there is reference to VIAS and whether that should be removed as a result of upcoming decision

Rudi: good point, any of the spec editors on the call?

[none]

Ted: it could remain and a link to it as a Note, perhaps deemphasized

Paul: it reads like it is a spec

Hira: we should perhaps discuss VIAS survey before

Gunnar: the github issue I am working on only says to check the wording

draft charter review

[deferred to complete VIAS decision and continuing on github]

VIAS survey and decision

survey results

Paul reads off stats and comments

Ted: the three implementation records are for the same (ACCESS) implementation

Paul: Adam mentions intention to integrate in a web runtime
... is that built into browser or as a JS library?

Rudi: a JS library being used inside a web runtime

Ted: people seem pretty evenly split on how they saw it, as a JS library or browser API

Gunnar: next question is how your thinking has evolved

Paul: conversation was a bit split

Gunnar: if RSI is backwardly compatible with VISS then VIAS should similarly

Paul: we moved away from WebIDL to a service given what the auto industry needs
... these can diverge since the API and is somewhat already

Peter: the big question is would VIAS be a widely adopted standard or should be published as a Note
... I don't see it getting adopted

Paul: neither do I, people will write their own libraries

Peter: in JS world it works like that
... it is still very useful as a Note

Urata: I wrote a comment that I would like to have an explanation on why TAG wants HTTP+WS

Gunnar: REST mapping is easily done and should

Paul: should VIAS support HTTP is one question but whether the API should support any protocol is another question

Gunnar: whether we should work on HTTP is more a question for the next charter

Urata: intention of adding MQTT and others is to indicate VIAS could be implemented over any overlying protocol
... if it is cause of misunderstanding, I can remove them

Paul: VISS is defining a protocol and TAG told us to use HTTP REST
... that is a bottom up perspective and have anything sit on top of it
... question to the group is whether they support a totally separate specification that could have anything underneath

Urata: JS library is one type of implementation
... what VIAS is defining is an API which is kind of a service thing and not talking about implementation
... it is one way to realize VIAS

Joakim: VIAS could be written as a WebIDL?

Gunnar: we are going in circles, how would people feel about deferring on this for six months?

Paul: we have the charter to work out

Gunnar: we have useful work we need to do first

Peter: I should think we should decide and more forward and don't want to repeat this

Joakim: I can see the advantage of having a JS library

Paul: some are ambivalent as their companies are not invested

PatrickL: I have some minor interest in a JS library but have been staying out of it

Ulf: I think what is being displayed right now is CR or Note, question 7
... of the responders only 25% are in favor of CR

Ted: question 7 also states the challenges (existing and pending issues list, TAG review and appeal Director decision) of bringing VIAS to CR. Subsequent questions inquire about non-JS platforms being used by others and level of effort and commitment WG participants are willing to make toward VIAS at this point

Hira: if VIAS has more time, I will contact more browsers, tier 1s etc
... if needed having HTTP in VISS

Peter: I agree with Ulf, question 7 is the core and decides it clearly
... we should focus on RSI

RESOLUTION: Publish VIAS as a Note, the group will focus on VISS+RSI in charter

Hira: please summarize

Paul: per the survey VIAS will become a Note. VISS and RSI will go into the new charter

Hira: can we keep the timeline for VIAS?

Paul: the Note can be maintained and people believe it would be useful (as a JS library)
... people would like to focus work on a fuller service
... when that becomes more stable and useful later we should consider other software libraries in a general IDL, not just JS

Hira: this is premature, can you tell me how to get it back from Note on to REC quickly?

Ted: it would have to be part of a chartering process

Hira: I want more time to recruit browser vendors

Gunnar: you can seek more implementations and then come back to the group but there is no guarantee we can provide up front

Hira: I want wording that welcomes invitation to implement VIAS within the Note

Paul: the core of the group's interest is the service specification
... the two additional OEMs who joined this group have done so for the service

Urata: I understand we cannot continue this circuling discussion and vote is a good way out of the loop
... I understand it is not just a question of the number of implementations and this could return to REC track
... my concern is fewer would implement VIAS as a Note
... the majority were in favor of it becoming a Note and that is understandable to me

Hira: we need to add wording to VIAS showing the group decision and encouragement of implementing

Ted: we can work that out as an issue in github before we publish it as a Note