W3C

Automotive Working Group Teleconference

12 Jun 2018

Attendees

Present
Paul, Ted, Hira, Guru, Ulf, Magnus, Gunnar, PatrickB, Adam
Regrets
PatrickL
Chair
Paul
Scribe
Ted

Contents


Paul does roll call

Paul: we were discussing whether to continue with the topics PatrickL proposed
... we have a proposal from Ulf for consideration at a wider attended call to introduce
... BG summary
... AGL did a review of VISS/ViWi

Magnus: that is fine by me

BG task force summaries

Ted: when we moved to this rotating call schedule we agreed to have the BG call second Tuesday of the month shared with the WG call
... in practice we have only had a couple such calls. Wonsuk and I have been discussing by email whether that works and if we want to make adjustments
... today's call in the rotation isn't conducive to Asia and he is not here
... he, our other BG chair Paul and I will discuss and see how we may want to modify it
... there are two task forces that fall under the purview of the BG, Automotive Payments and Big Data
... since the BG acts as an incubator it is possible to propose additional topics you think should be explored and see if there is common interest
... we certainly have plenty going on with the resources we have but want to remind people of that and we are interested in hearing ideas
... the auto payments tf is formative, we have been hearing from some stakeholders and collecting insights in these calls
... IFSF, NACS, WEX, Mastercard, JLR and others have been sharing their views
... Ian Jacobs (W3C Web Commerce Lead) and I have been speaking with other stakeholders and trying to determine the viability of W3C Web Payments and other standards applying to automotive use cases
... we are hearing different architectures where it is not clear how it would be applicable, perhaps in a hybrid solution
... I agree for consistency and to avoid user (driver) confusion you do not necessarily want head unit interacting with different merchant sites directly but going through a portal to enforce consistent UI and therefore intuitive UX
... certainly our signals work is useful for payments, the data has many uses including to help reduce fraud
... we are talking with the FIDO alliance on our collaborative authentication work and its application here
... we will be trying to figure out the direction this task force should take or if we should consider pausing it for the time being
... on the data tf we have only had a couple calls so far, next biweekly call is on Thursday but shifted an hour later
... first call was to go over potential areas of focus, deeming some out of scope
... we delved a bit more into data contracts and ontology on the next call and hope to continue the consent capture conversation started at the F2F

https://github.com/w3c/automotive-pay/wiki

https://www.w3.org/community/autowebplatform/wiki/Data_tf

Ulf's proposal

https://lists.w3.org/Archives/Member/member-automotive/2018Jun/0000.html

Ulf: as mentioned I think we should base ourselves going forward on VISS, however there are quite a few benefits from RSI to take into consideration
... from the data perspective VSS seems preferable. in VSS we may want to allow for service model resource elements as seen in RSI (ViWi)
... VSS currently cannot support it but it should be possible to have zero or more leaf nodes under a resource node
... in ViWi there is a registry of element nodes and in VSS there is no place for that
... my proposal is to invent a new branch node type rbranch in which the definition of its children, element nodes is found
... rnode
... first needed a generic node definition. these are the main additions and from a logic point of view, this addition to VSS would make it possible to support similar model to how things are done in ViWi
... you can have multiple element nodes under a resource node
... we can expand this further and create an access paradigm as in ViWi and support its query mechanism as well
... this way we can support both models and port resource data models
... this would be a hybrid in some way
... I have also in parallel to thinking about this have been working on a proof of concept implementation on top of VSS
... beta level solution is ready and willing to share

Ulf: just a quick introduction for now

Ted: thank you Ulf for investing the time into understanding the divide and working to bridge it
... we will have a more substantive conversation next week and encourage people to read your proposal and the proof of concept code you will be sharing in preparation

Magnus: I spent some time investigating and was initially a bit skeptical but have also spoken with Ulf on the side
... it is a great effort to try to bridge the two and have a same model

Ted: as discussed at the F2F, settling on a single data model alone is significant
... ideal we have a single method for accessing that information as well
... I was asked by someone what happens if we cannot converge VISS and RSI
... we are not at point to make a decision but the group can propose one, the other or even both
... VISS has a head start on the standards track, RSI might get more momentum from market share and impossible to predict
... I still want to try to merge them and see strengths each have
... there is a precedent at W3C to have competing specs published and see which prevails in the long run
... for entertainment we have VTT and TTML for handling captions and subtitles for video objects
... VTT predated but it is looking like TTML has more uptake and what W3C won it's Emmy award for
... having two solutions is not easy process at W3C and would take some time. ideal is clearly a single solution
... I will start a mail thread on this topic at some point. I do not want to have a serious discussion on it yet but make sure the group is aware

Ulf: I will provide my patch and instructions on how to make it run

AGL spec review

http://docs.automotivelinux.org/docs/apis_services/en/dev/reference/signaling/architecture.html#query-and-filtering-language

http://docs.automotivelinux.org/docs/apis_services/en/dev/reference/signaling/high-viwi-architecture.html

Paul: some links were brought to my attention about analysis AGL did
... they apparently also did an implementation
... a perspective from outside the group is useful. we should try to get the people involved to share their view
... we do not want to turn this into a political debate
... mostly an introduction for people to read and we can have a conversation later

Ted: I spoke with Fulup at the last Genivi AMM, mostly on how to engage Japanese OEMs and Tier 1s better and not about this.
... he joined some past calls on VISS/RSI when the group had some meaningful calls comparing them.
... I will invite him and his colleagues to join a call to go over them and encourage the group to read in advance.
... I agree a somewhat objective opinion would be worth hearing and second Paul's call to avoid getting mired into politics
... many have formed their opinions already and need to represent their organization's interests

[adjourned]

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2018/06/12 14:17:16 $