See also: IRC log
Johannes: F2F@Montreal Updates. Draft agenda available
https://www.w3.org/WoT/IG/wiki/F2F_meeting_2016,_April,_11th_-_13th,_Montreal,_Canada
There's still room for contributions
(see Input/Comments section)
Sebastian: contribution to add from Bosch about the BEZIRK IoT platform
Johannes: there should be a
dedicated section for expert talks -> Sebastion will do
it.
... next topic: Plugfest.
room will be available on Sunday for demos, presentation on the same day as open contributions. Same setup as in Nice
See table on the F2F wiki page for contributions to the plugfest.
Put your name if you are willing to participate.
Reminder: deadline for registraton is April, 5th
Daniel: comment about the plugfest: would it make sense to move it later?
in the afternoon.
Johannes: new column in the participatin table for availability on Sunday -> Daniel will do it.
Sebastian: just updated it myself: added an Encoding column to know which encodings are used by the devices.
Louay: column for discovery. Still a bit early to tell what kind of discovery we want to setup during the plugfest.
Johannes: last time, was a central discovery with self-registration to a server (see webthing-repository on Github)
thingweb-repository*
Louay: my own demo will offer 3 discovery mechanisms. How a client will know which one(s) is (are) available
Johannes: How-to needed?
Sebastian: Feedback from last time: common vocabulary needed. Discovery should rely on that.
Louay: also important. Example of SSDP: SSDP requests (similar to UDP request) and response with some links? What should be the content of the requests? Should the response contain link to the Thing Description, something else?
Indeed, how-to/discussion needed.
Soumya: als wants to have a mDNS client, we should synchronize.
Johannes: this how-to should be included in the Current Practice document (http://w3c.github.io/wot/current-practices/wot-practices.html)
Sebastian: discussion about a common vocab should also start now.
Darko: we could use the Scenario table to foresee which concepts are needed.
Johannes: further topics about the F2F?
Soumya: in the agenda, joint breakout sessions. Should it be prepared by each TF?
Johannes: would make sense. Topics should be collected directly on the wiki page.
Next topic: Charter
Johannes:set up by Dave , kaz and Yingying.
contributions welcome through Github (issues
or pull requests)
https://github.com/w3c/charter-drafts/blob/gh-pages/wot-wg-2016.html
in our own Github repo, all raised issues in the WG/ folder should be added to the charter
Feedback is welcome about this charter, e.g. relevant agenda? IPR?
already submitted a PR: https://github.com/w3c/charter-drafts/pull/47
end of the IG topics. Moving to TD topics
<kaz> WoT Current Practices Document
Sebastian: see Current Practice document.
<scribe> new structure based on feedback from last time.
1. "href", new term to indicate how to access the Thing's properties, events, events.
2. removed "metadata". Instead metadata fields directly in the root object
<kaz> Example 1 at 3.1.1 Quick Start: TD Samples
Michael: clarification: intended to use standard URI rules (like concatenation of base URI and localname)?
(example was "href": "temp")
those rules are defined in RFC 3986
something that would be needed is to have composite Things
Victor: it has been added to the TD vocabulary: "associatedThing", to link different Things with each other
Johannes: other comment about mandatory and optional fields
<Soumya> I have to leave now, but I have experience the same Michael mentioned. I considered them as M2M device having several endpoints and described the device and endpoints seperately.
<Soumya> I can talk on this later if required
<Soumya> Bye for now
Sebastian: all this should be added in the Current Practice document
Soumya and Johannes agree with Michael
Johannes: TD is the blueprint of how to communicate with the Thing. Not known beforehand. Need for TD "template"?
s/"template§/"template"/
Darko: similar to the difference between class and instance. Using classes/templates could increase interoperability
Daniel: about Michael's question about "href". In case of URI arrays, cartesian product or one-to-one mapping?
Michael: definitely not in the RFC. Should have a kind of "map" of base URIs
where each base URI fits
Daniel: what if the resource structure is different depending on the scheme (HTTP, CoAP)?
Michael: chance of having such a situation?
Sebastian: this was about TD core. There's also the possibility to extend a TD with other vocabularies.
Darko will present this part
Darko: 1. machine-readable -> defined term in vocabs already available.
2. aditional properties to describe a Thing? Provided in an extended context, as shown in Example 2 of the Current Practice doc.
Ex: "@type": "sensor:Temperature"
or "sensor:unit": "sensor:Celsius"
Michael: it really makes sense to do such things.
have you tried to convert JSON-LD into RDF triples?
Darko: by the way, we have much better TDs if we start from RDF and turn them afterwards into JSON-LD than writing it directly in JSON-LD
Michael: things would be easier if we put as many things in the context, maybe?
Darko: want to standardize TD, we shouldn't re-standardize existing vocabs (such as units, quantity kinds)
way to go -> one TD context + one horizontal context for common concepts like units?
Sebastian: moving on to the next example of a TD (see Current Practices)
"uri", "decodings". And "security": requirements to access Thing's resources
self-contained. Example 3 shows auth with JSON Web Token (JWT).
A lot of new things, details to be found on the document. Comments welcome.
other questions about the TD?
End of the meeting.
This is scribe.perl Revision: 1.144 of Date: 2015/11/17 08:39:34 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/and Yingying./, kaz and Yingying./ Succeeded: s/see Current Practice document/Sebastian: see Current Practice document/ Succeeded: s/"template§/"template"/ FAILED: s/"template§/"template"/ No ScribeNick specified. Guessing ScribeNick: VictorCharpenay Inferring Scribes: VictorCharpenay Present: Kaz Johannes Achille Claes Daniel Darko Johannes Katsuyoshi Louay Michael N_Wang Ryuichi Soumya Takuki Toru Victor Yingying Got date from IRC log name: 16 Mar 2016 Guessing minutes URL: http://www.w3.org/2016/03/16-wot-minutes.html People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option.[End of scribe.perl diagnostic output]