- From: Rudolf J Streif <rudolf.streif@ibeeto.com>
- Date: Thu, 3 Oct 2019 14:15:55 -0700
- To: ted@w3.org, public-automotive <public-automotive@w3.org>
- Message-ID: <6e73b4a4-814e-bb28-d264-a4856c294b13@ibeeto.com>
Thank you, Ted. Albeit I was not able to attend the F2F, I followed up on a discussion we had on this mailing list about REST and the nesting depth of the signal tree. My suggestion was to use a UUID that can be derived from the tree branch. This week I submitted a pull request (https://github.com/GENIVI/vehicle_signal_specification/pull/125) that changes the current UUID to one that is computed from the branch path using UUID5 (SHA-1 hash). The namespace is computed as a UUID5 using 'vehicle_signal_specification' as the object identifier. I think that should make it pretty simple to implement a REST server using the UUID as identifier. I think that the UUID will also help with a registry as it can greatly speed up lookups of what is available and who provides it. :rjs On 10/3/19 12:45 PM, Ted Guild wrote: > As the minutes are scant, I'll add the following summary. > > We discussed a number of proposed, accepted and early exploration of > other standards efforts (ISO TC22, TC204, TC211, JTC-1; AutoSAR) to > align on around vehicle signals. Additional are stemming from the > transportation data workshop we held later in the week and I provided > a summary of those developments on the latest WG call. We had a guest > who offered to discuss with contacts at VDA. We discussed continued > attempts to establish something with SAE. > > Adnan shared his and Daniel's analysis of data structures of VSS and > ViWi, how tree structure is more conducive to REST and Graph. We > discussed the various options around VISS/ViWi/Gen2 the group could > take to get us moving forward again. Patrick joined remotely for part > of the second day and we agreed to continue on Gen2, hopefully drawing > more in readily from ViWi such as registry as that will expedite spec > development but not the data depth limitation. In order for this to > happen we need to revitalize the group, issued a call for editors > including asking individuals in room to reach up to management to get > an influx of resources. We are receiving affirmative responses in > subsequent WG calls. > > We did various preparation for the Meetup Monday night and debriefed > after on Tuesday. We received a good turnout, 50+, lots of questions > and interest. I will be sending them a workshop report and other > materials about our activity. > > We discussed various industry trends. GM and Android news sparked > discussion including how there will likely be two or more data streams > on vehicles and VISS/Gen2 still useful in places like TSU. > > We prepped for the workshop at the end of the week. > > We reviewed a proposed overhaul of the W3C Auto landing page > http://w3.org/auto and made a number of revisions based on feedback. > > We discussed some features for certain use cases like accommodating > aftermarket equipment signals. > > We discussed the various major use cases and when Gen2 will provide > sockets (eg on-board data sampling), REST (on-board and in-cloud where > sockets make less sense as you have sampled stored data) and Graph > (cloud as primary but also scenarios on-board for AI). > > We agreed we want to start prototyping a Graph representation of > VSS/VSSo. > > https://www.w3.org/2019/09/09-auto-minutes > https://www.w3.org/2019/09/10-auto-minutes > -- ----- Rudolf J Streif CEO/CTO ibeeto +1.855.442.3386 x700
Received on Thursday, 3 October 2019 21:16:21 UTC