See also: IRC log
<kaz> scribenick: McCool
<mkovatsc_> https://www.w3.org/WoT/IG/wiki/IG_WebConf#Agenda_of_next_WoT_IG_WebConf:_14_December_2016
<inserted> scribenick: McCool_
Yongjing mentions the msg from oneM2M to W3C regarding liaison has been sent to the W3C liaison Team
<kaz> Yongjing will forward that message directly to Dave and Kaz
OCF Liaison... call this Friday
currently skype, can make another arrangement if necessary
kaz: got approval for WG from director
so, we are done... working with Wendy for the next step
Michael McCool mentions APA issue has been resolved
should accept the collab pull request
seem to be some conflicts
is a conflict in the pull request... number 286
Matthias will attempt the merge
<kaz> WG Charter at the official location
Kaz will also check details. the point here is that there should not be any difference between Wendy's version and our repo version.
all other WG charter pull requests to be closed
can run diff tool on version in repo with version used by Wendy to make sure there are not any differences
also, please accept 271, presentation to OCF; Mozilla issue is resolved
<mkovatsc_> Wendy, could you send me your version? I will commit it to our repo.
Wendy made some edits, Matthias to copy into repo
starting date is date of W3M: Dec 7, 2016
Kaz working with Wendy for announcement
Kaz will contact chairs to start planning meetings for WG
McCool: F2F funding has been obtained
will book Crowne Plaza
Feb 5 set-up in the afternoon
feb 6-9 for event
invitational day-- the 9th?
regarding visa, intel should provide invitation letters
please contact Ryan Ware for visa letters
a form would be useful...
<kaz> the procedure depends on countries, so people are encouraged to check with US embassy in their countries
<kaz> Yongjing asks McCool to put the info on VISA/invitation on the f2f wiki
<kaz> f2f wiki
<inserted> scribenick: McCool
oneM2M agreement being reviewed, will be distributed when ready
will discuss next week
<inserted> Kaz has just found the liaison message from the oneM2M contact on the email archive, and will work within the W3C Team.
Matthias call with Tom Burke and technical director of OPC
presented what WoT is doing, technical building blocks
main take-away: first time encountered another org with holistic framework for protocols, semantics, vocab
they started to integrate companion standards
OPC is pretty powerful... but WoT is perhaps even more
OPC also started with RPC and optimized for their use cases originally
have not quite figured out how to deal with collaboration with w3C
will have another meeting with someone at Siemens, discuss roadmap, come up with some actions for liaison
<inserted> Scripting minutes
discussed REST API in charter or not... determined to leave as it is
will however, put a RESTful definition somewhere; as an informative deliverable under something else, eg. Binding Templates
Could also do a Note
could also do an appendix
"REST API is generated from a TD like this"
McCool: an orientation session and a tutorial would be useful
-> consider when doing agenda for F2F
FAQ would be useful as well
had getting started slides
Q&A format is helpful to focus on small
input from newcomers would be useful
how do we optimize onboarding -> let's discuss at F2F
We also discussed the MMI group at SA meeting
Multi-Modal Interaction
<kaz> WoT and MMI discussion
how to write a web site with non-standard interfaces, eg speech, specialized I/O devices
<kaz> interaction between WoT's T2T and MMI's UI
architecture is similar... how can TD and modality components interact
will let us do things like connect to speech recognition etc.
would be a useful demo
proposal to invite MMI people to plenary calls and/or to F2F
<kaz> Resources on MMI for PlugFest:
- MMI lifecycle event transport
- MMI Interoperability Test
academic workshop on WoT... universal descriptions, generate right UI
ex, vision impairment; universal remote controller, person also in APA
<dsr> Dave notes that on the W3C staff, Shadi Abou-Zahra is responsible for accessibility for the web of things.
<inserted> Kaz mentions we also invited Michael Cooper from the APA WG to the WoT session during the Web Conf in Montreal.
Could turn into an umbrella topic: interactions between Users and Things
removed global definitions of endpoint
encode information at a local level
links is more used than endpoints
also use href rather than uri
also introduced a "base"
this is the uri for this TD, valid for all interactions
then hrefs are all relative from that
but... can still use absolute href
overriding base
this is the "release candidate", please comment as soon as possible
will freeze before Christmas so we have a stable version for the plugfest
events, still using structure we were using before
is an event example in the quick start
also uses communication meta-data on a local level using links
<mjkoster> I have comments but my audio is not coming through to the conference
is this enough information to set up stream-based events?
<mjkoster> have we defined stream based events?
for CoAP, can do Observe
<mjkoster> What other styles are we talking about?
<mjkoster> is it pub/sub?
not sure what other styles of event handling can be handled
<mjkoster> +1 jhund
<mjkoster> we need one model that expresses thru TD, scripting API, and REST binding
<mjkoster> my audio is not getting out so I can't unmute
call for implementations? which patterns? still some major work to get eventing right
dsr: has been investigating, don't need ALL the information, need "contract" information
should minimize information in TD
dsr: need to identify "driver"
mediaType - the right approach?
<mjkoster> We can use link relations and all web linking functionality
<mjkoster> +1 jhund
<mjkoster> yes!
is some protocols that can negotiate media, and some that can't
dsr: what is contract platform
should use to talk to another platform
... mediaType is wrong level of detail
<mjkoster> I have no audio
koster: agree with web linking direction
web links can point to pub/sub topics. rel can say it's a pubsub topic. link relations can be used for event mapping also. there is common vocabulary around relations, even a baseuri already in common use
<dsr> I wanted to state that the thing description needs to provide an identifier that the application platform needs to identify which contract to use to communicate with the remote platform. Details like IETF media types are part of that contract and should not be exposed to the application developer as they are not relevant to the object model that the application developer deals with.
needs more clarification
<dsr> Thanks Michael for the explanation, an example of a contract parameter is the OCF resource type and resource id
McCool: maybe contract defintion with arguments
what we need is a concrete definition
<mjkoster> hypermedia is a general contract
<mjkoster> I can hear, but I can'r speak
michael K, dsr, get together and define a concrete proposal for "contracts"?
outer contract, then "inner" format is encoding
dsr: another issue is that "mediaType" has been used for data formats, is already a standard
<mjkoster> the HTML media type defines quite an interaction model
it is PART of the contract...
<mjkoster> The OCF media type will also define it's interaction model application/vnd.ocf+cbor
dsr: agree there is a need for a registry
<mjkoster> I fully support the idea of making TD an extended web linking approach
<dsr> A registry of identifiers for the contract where the registration includes the metadata parameters involved
can we rely on things that are already "out there" via links?
<dsr> we could of course just use linked data rather than a centralised registry
and can even put parameters in uris...
(above comment from McCool)
Agenda for next meeting: REST discussion, contracts between application and platform
plans for next plugfest; logistics, agenda, etc
procedure for adding agenda items: edit wiki and put it in the "agenda stash"
[ adjourned ]
