- From: Kazuyuki Ashimura <ashimura@w3.org>
- Date: Thu, 26 Jan 2017 01:14:14 +0900
- To: Public Web of Things IG <public-wot-ig@w3.org>
- Message-ID: <CAJ8iq9ViYG0rrPu05PMjxs+O3LGf3e0U-w9QnbeOMOFj3gr6ww@mail.gmail.com>
available at:
https://www.w3.org/2017/01/25-wot-td-minutes.html
also as text below.
Thanks a lot for taking these minutes, Taki!
Kazuyuki
---
[1]W3C
[1] http://www.w3.org/
- DRAFT -
WoT IG - TD Restructuring
25 Jan 2017
See also: [2]IRC log
[2] http://www.w3.org/2017/01/25-wot-td-irc
Attendees
Present
Daniel, Dave, Katsu, Kaz, Sebastian, Taki, Victor
Regrets
Chair
Sebastian
Scribe
Taki
Contents
* [3]Topics
1. [4]GitHub Issues
2. [5]Dave's proposal
3. [6]PlugFest
* [7]Summary of Action Items
* [8]Summary of Resolutions
__________________________________________________________
<scribe> scribe: TK
<scribe> scribeNick: taki
GitHub Issues
DP: Can we go through github open issues?
... JSON-LD uses @base and we use base
... How can we align them?
Sebastian: @base is only be able to use in context definition.
DP: We should align with JSON-LD.
Sebastian: We have to check it. If everybody is ok.
... To always use singular name.
DP: JSON-LD chose it. When we introduce new concepts. we
should follow the same way.
Sebastian: "interactions", for example
... It is a good topic to discuss during face-to-face.
... TD model should be independent of specific serialization
format.
... We need to define TD model.
... And then, should be able to provide various serialization.
... Core definition. Then comes the serialization discussion.
... Simplify TD. The difinition using JSON schema is a bit ugly
now.
... "Type" postfix may not be necessary. For example,
"valueType".
... Add a class field to the TD. This is looking for grouping.
... This is an old issue.
... I will double-check with Alex.
... to see whether we can close it or not.
... Allow the definition of resource parameters in TD.
... Rest-based servient. Query parameters, currently we have no
way to support it.
... Please make comments on issues.
... Vocabularies for UI, for example. Vocabularies for images,
for example.
... proposed by Johannes.
... Avoiding arrays. (by Dave)
... We can discuss this topic in Santa Clara.
... Separating discussion of requirements from serialization
formats. (by Dave)
... Template may not be possible in the new structure. (by
Darko)
... But it turned out that was not the case.
... I will close it.
... Actions vs properties.
... We should make a decision.
... There are a lot of definitions coming from other
consortiums.
... Do we want to define ours? Do we want to depend on existing
ones?
... Please comment on issues. When we have agreements, I will
close them and integrate into current practices.
Dave's proposal
[9]Dave's writeup
[9]
https://lists.w3.org/Archives/Public/public-web-of-things/2017Jan/0002.html
Sebastian: Overview of what Dave did.
DR: Many platforms. OPC, OneM2M, etc
... We need abstraction
... We need to be through in looking at use cases.
... Application contracts decouples from the rest.
... Web of things demo page...
... What kind of requirements emerge from those use cases?
... Cyber-Physical system for a tank
... This is available on the web.
... Water tank with action and properties.
... RDF, turtle provides independent model.
... Showing graph...
... Properties... each property has name.
... Context is part of model.
... Shopping basket use cases.
... It has basket, consisting of product, quantity...
... product is a union of meat, fruit, vegetables. and so on...
... Sometimes turtle is easier to read than diagrams.
... enum and union are used.
... We need to distinguish integers from numbers.
... Color Choices
... rank and color are defined as enumeration.
... diagram is generated by JavaScript.
... oneM2M refrigerator
... it defines device ontologies.
... importing a number of different modules.
... That should be enough now.
... boolean, string, number, integer, collections, enums,
union, typedef, vectors.
... 3D printing
... streaming property...
... Application contracts. "platform" uri. and "sink".
streaming properties provides blocks of data.
... Showing nested properties.
DP: You collected all the use cases.
... RDF, turtle, JSON mapping. Is it similar to what we are
doing?
DR: I use RDF as the model.
DP: OK, it seems similar to what we do.
... Some people does not seem to like RDF.
DR: It depends on the audience.
DP: You seem to have different type system.
... How should we converge them?
DR: I looked at more use cases. We need some more such as
union, enumerations, collections and vectors.
... Numbers vs integers.
DP: Schema also supports some of them.
DR: OCF and OneM2M use cases requires some more.
... It is not really good at descrining linked data.
Sebastian: 80% of the use cases, I think we can use
JSON-schema.
DR: That is not true.
... meta-data is critical.
Sebastian: We would like to rely on existing technologies.
... Matthias mentioned JSON-schema standardization members will
join face-to-face meeting. We can collaborate with them.
DR: We have to address broad range of communities, we have to
bear in mind.
... By the time of face-to-face, I will do more analysis.
... OPC uses their own modeling, similar to RDF.
... Healthcare is another use case.
<Zakim> kaz, you wanted to ask about (1) unit and type and (2)
if we want to include the figure/idea of "app contracts" to the
architecture doc
KA: How to handle datatype. We can consider JSON-schema.
... New diagrams. IoT platform contracts.
... IoT platform contracts. Is it different from protocol
mapping?
DR: Yes
KA: Application contracts. Is it TD and Scripting?
... You mean various application languages?
DR: JavaScript, Java, whatever.
KA: Our scope was abstraction layer, I believe.
DR: We also need security, provacy, etc. Non-functionals.
... Contracts between applications.
<kaz> [10]automotive architecture
[10]
https://www.w3.org/TR/2016/WD-vehicle-information-service-20161020/images/W3C_Vehicle_APIs_Component_Diagram_v1_3.png
KA: I am not saying your diagram does not make sense.
... Application contracts layer possibly includes WebSocket
interactions.
... Then this approach is a bit different from the current
architecture.
DR: I believe it is same.
KA: Mechanism seems different.
DR: Current architecture is less about layering.
... I have coordinated with OneM2M, OCF. But the work was
mostly done by myself.
<inserted> [ Maybe a good topic for the OCF liaison call :) ]
Sebastian: Thank you Dave.
PlugFest
Sebastian: Plugfest.
... Showing WoT features list...
... We have testcases.
... We can go through the list during plugfest.
... For example, discover things, read properties, etc.
... It would be nice to use table to find what is working and
what is not working.
... Let's discuss later during main telecon.
<Zakim> kaz, you wanted to ask if this check list is available
online
<sebastian>
[11]https://github.com/w3c/wot/blob/master/plugfest/2016-beijin
g/TestCaseCoverage.xlsx
[11]
https://github.com/w3c/wot/blob/master/plugfest/2016-beijing/TestCaseCoverage.xlsx
TK: Sorry I did not ask Nimura-san what he will bring to
face-to-face.
<victor> [12]http://json-ld.org/spec/latest/json-ld/
[12] http://json-ld.org/spec/latest/json-ld/
victor: I wanted to mention that JSON-LD spec was updated
recently.
... I want to discuss this during face-to-face.
Sebastian: Mr. Kellogg will come to face-to-face.
KA: Is this ver. 1.1 a draft CG report, right?
victor: yes.
... it will stay as a community group report.
... Dave is trying to specify. The similar is also formalized
in the document.
... therefore, it is worth taking a look at it.
<kaz> [13]7. Relationship to RDF
[13] http://json-ld.org/spec/latest/json-ld/#relationship-to-rdf
DR: Thank you. We need to have dialog across group. I will do
that.
Sebastian: I should close. Talk to you later in plenary
meeting.
[ adjourned ]
Summary of Action Items
Summary of Resolutions
[End of minutes]
__________________________________________________________
Minutes formatted by David Booth's [14]scribe.perl version
1.147 ([15]CVS log)
$Date: 2017/01/25 16:09:23 $
[14] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
[15] http://dev.w3.org/cvsweb/2002/scribe/
Received on Wednesday, 25 January 2017 16:15:32 UTC