Web of Things IG, day 3 -- 31 Jul 2015

Report from yesterday's breakouts

<inserted> scribenick: kaz

Sebastian: makes report from TF-TD breakout

Johannes: makes report from TF-AP breakout
... architecture model
... technology landscape
... lifecycle states
... abstract resource model
... complete and transfer Tech Landscape

kaz: we did great discussion and got a basic consensus yesterday (and probably today as well)
... but given not all the IG participants could attend this meeting, we'll bring all the results to the telco, etc., and let the other participants as well know about the result. Right?

johannes: right

joerg: makes report from the TF-DI
... sketch of landscape
... relationship with TD
... discussion on requirements
... discovery should be independent of the communication technologies
... communication technology should be visible in the metadata
... things should be capable to automatically register themselves
... interaction patterns of discovery categories
... evaluation criteria for discovery tech landscape
... next steps: complete evaluation criteria, conduct evaluation of discovery tech

oliver: makes report from TF-S&P

johannes: will have discussions on interaction, security model, etc., during the upcoming calls

joerg: what's the best to discuss security&privacy?
... atomic use cases on security and privacy
... mapping and deep understanding
... how the different TFs started their contributions are different with each other
... wondering about the best way
... initial contributions are made on the wiki

dsr: don't have strong opinion at the moment
... GitHub is accessible as well
... issue tracking through pull request is available

johannes: just took a template
... looked into the wiki to see tech landscape
... got pull request from Jonathan
... GitHub is a possibility
... as a resource control system
... can continue to use wiki
... and make a decision when to transfer the content
... it's just copy&paste

joerg: working draft of deliveralbles
... format:
... continue the wiki until structure is confirmed
... for finalization benefit for tracking in GitHub
... structure of the documet has been proposed

johannes: have some contribution
... valuable to keep them

sebastian: overview landscape on the wiki
... can put it on GitHub

joerg: initial contributors provide support to the document structure

sebastian: yeah

johannes: three TF moderators should work with each other

joerg: review of TF of the document structure
... evaluations supported by contributors
... next step: ask for support in evaluation by controbutors
... sync in cases, where technologies are studied in several TFs

<dsr> scribenick: dsr

Jerry Wu, Delta, Point of View on IoT

Jerry introduces himself. Delta Electronics was formed in 1971. We have a broad business areas focusing on power electronics, etc.

We have sites all over the world. I come from Delta Research Center.

We are like an incubator for new ideas.

<JonathanJ2> Delta - http://www.deltaww.com/default.aspx?hl=en-US

IDC says the IoT market will reach $7.1 trillion by 2020. We all want a place in this market, but it isn’t easy.

There are many niche solutions from a long tail, and a fragmented market

Low popularity and difficulties in crossing domains.

We need to adapt dynamically to the client’s requirements. They may not know what they want, but do know what they don’t want

A common platform would increase economies of scale and reduce the costs.. We need a loosely coupled architecture.

We want to build commonalities on the IoT ecosystem.

Our current focus is on smart manufacturing, smart life and smart service

We plan to combine existing Delta business and needs to create IoT solutions.

We are working in an IoT common programming model. This currently uses Java but we want to also support JavaScript. This supports event driven, parameter driven, dynamic binding and rule based approaches.

We use an ontology editor that is web based and supports multiple views

We have a common component library and a universal kit (multisensory, flexible customisation and edge computing)

Thanks for listening.

Questions?

Joerg: you showed this long tail perspective, however, you also have first movers. Would you be interested in helping us to examine your use cases and the atomic use cases they involve?

Jerry: yes I will look into that

Ravi Subramaniam, Intel, Intro to OIC / Iotivity

Note that Ravi asked us not to minute this session.

See: http://openinterconnect.org

“The Open Interconnect Consortium is focused on delivering a specification, an open source implementation, and a certification program for wirelessly connecting devices.”

IoTivity is an open source software framework enabling seamless device-to-device connectivity to address the emerging needs of the Internet of Things, see https://www.iotivity.org

WG Charter

<kaz> scribenick: kaz

dsr: we're an IG
... we're developing use cases and requirements
... would talk about possible WG

[ From the Web of Pages to the Web of Things ]

[ Web of Things ]

dsr: clarifying the idea of proxy

[ The Web as the Global Data Bus ]

[ Horizontal & Vertical Metadata ]

[ Focus of W3C Contribution ]

[ Role of WoT Interest Group ]

dsr: collecting use cases
... studying requirements
... reachout other groups and share vision

<scribe> ... ongoing role to support work

[ Web of Things Working Group ]

dsr: proposed scope and work items
... define a core vocabulary for describing thing data models
... define a serialization to JSON-LD
... define bindings to common Internet protocols in collaboration with external groups
... define a vocabulary for servers...

[ Some Considerations ]

dsr: need to check WG's decisions are appropriate for a range of stakeholders
... also horizontal reviews

[ Data Models - Objectives ]

dsr: standard vocabulary for thig data models will enable servers to construct virtual objects
... enabling developers to interact

[ Data Models - Details ]

dsr: what kind of data type?
... building atomic use cases
... complex types including lists and maps from names to values
... analogous to "typedef" of C language

[ JSON-LD ]

dsr: RDF, JSON-based serialization of RDF, need for JSON-LD

[ JSON-LD and Data Types ]

dsr: RDF re-uses many of the XML Schema built-in datatypes
... JSON-LD context would allow "boolean"
... application/tdl MIME type
... additional terms: events, properties and actions
... stream data type
... "writable"

[ Bindings to Protocols ]

dsr: standard for how to signal events, property updates, action invocations, action results, metadata updates, etc.
... need to define this for common protocols: HTTP, Web Sockets, CoAP, MQTT, etc.

[ Compact Encoding ]

dsr: limited memory/packet sizes
... very high throughput
... data model of a thing
... compact encoding of JSON/JSONLD

johannes: why would this require us to form a new WG?

dsr: not sufficient this group to do standardization
... possible to joint work of the proposed WG and this IG

johannes: some thing special on WoT?
... there are already WGs on this work
... what is the real difference?

dsr: IG can't do standardization
... need to talk with WGs
... what is the expected relationship with other groups can be clarified in the Charter

johannes: horizontal technology like JSON-LD

dsr: a WG is needed for actual standardization

johannes: any specific WGs who are already working on our topics?

dsr: WoT datamodel could be a topic for the new WG

[ Out of Scope ]

dsr: vocabularies for specific application domains
... work related to security, assurance, prvacy and resilience
... could be added to future versions of the WG Charter, though

ravi: won't do fundamental work on security, but some security-related discussion possible?

dsr: narrow enough description is expected for the new WG's Charter

[ Opportunities ]

dsr: where there is a clear need based on agreed use cases, the WG could standardize vocabularies

[ What Next? ]

dsr: W3C staff contacts to prepare draft charter
... use GitHub for review and update it based on feedback
... identify candidates for the role of the Chair
... seek approval from the W3M
... AC review comments
... lanch the new WG
... first f2f

sebastian: WG vs IG?

dsr: IG does pre-standardization work

sebastian: not clear to me...

dsr: WG could do vocabularies, etc.

sebastian: doesn't make sense to simply stop the IG?

dsr: WG need to have a narrow scope
... IG can have broader scope

johannes: regarding the narrow scope...
... you mentioned some points in your presentation
... but didn't quite understand it

dsr: (proposed scope and work items)
... core vocabulary

johannes: negotiation for protocol mapping?
... make sense to have a bit more tight scope

dsr: two set of vocabularies

joerg: might want to identify our need
... IG started 3-4 months ago
... we're still on a transition state
... wonder if this is a trigger
... one of the points is that this IG is expected to identify requirements
... should we try to trigger new work for the WG?
... not sure if we already shared our understanding
... collecting tech landscape
... how to process compiling our deliverables?
... comments?

jonathan: how many deriverables?

dsr: 4 (plus optional one)
... after lunch we'll have discussion on implementation

Simon: would agree with Johannes that we (IG) could work with other WG

@@2: interested to see the relationship with OIC

scribe: duplication with their work?

ravi: see some complementary aspects
... semantic information independent of tech domains
... internet protocols
... resource modeling aspect
... potential conflict

dsr: possible collaboration with this group

ravi: just for answer to the previous question

Arnaud: WG is chartered for specification
... that is the biggest difference between WG and IG

dsr: W3C don't have to define concrete mapping for websocket, etc.
... can work with IETF

johannes: need clarification
... your opinion cover APIs?

dsr: bigger questions
... maybe standardized or maybe not

johhannes: protocol binding?

dsr: how to express it?
... there is a need for collaboration

joerg: based on the discussion so far, what would be the realistic way for us?
... would take this as a trigger to start discussion
... on scope, relationship with this IG/other groups, etc.

[ Lunch ]

<yo> quit

<dsr> we resume after lunch.

<dsr> scribenick: dsr

Joerg summarises his expectations in regards to collecting input for the working group charter.

(see his slide)

Work organisation and time plan, outreach

In respect to teleconferences, we’ve been rotating task forces through time slots. Some people are asking for fixed timeslots saying that this will be more convenient in practice.

The proposal is to have APIs and Protocols on Wednesday and Thing descriptions and Security and Privacy on Thursday.

Any other proposals?

The idea is to have calls at 4pm CEST on Wednesday.

Kaz: I agree that it is better to have calls at fixed slots, but I personally couldn’t make the 4pm CEST call.

Joerg asks for a show hands for keeping the rotation (none) or fixed times (around 6 people)

Joerg: we need to find the fixed timeslot according to the people interested in a particular task force.

Kaz: 3pm CEST would be 10pm in Japan and 6am in California

Joerg: we may make the call up to 90 mins.

In respect to face to face meetings, we have TPAC in Japan in October this year.

<inserted> TPAC 2015 in Sapporo: 26-30 October

He proposes Europe in January, and wonders if we could co-locate with the W3C AC meeting at MIT in March.

Dave: I can check to see if that is logistically feasible

Joerg: this would be followed by a meeting in July in Asia.

Joerg asks for people to check if they companies could host the January and July meetings.

Joerg: I would now like to discuss opportunities for joint meetings e.g. with the IETF

He asks for a show of hands in respect to the Yokohama IETF (1-6 Nov. 2015) the week following TPAC.

About 6 people put their hands up.

Question: would the joint meeting take place between TPAC (26-30 Oct. 2015) and IETF meeting (1-6 Nov. 2015) or during the IETF meeting?

Joerg: we propose the Saturday (31 Oct. 2015) and Sunday (1 Nov. 2015) in between.

Osamu offers to help with meeting rooms in Tokyo.

The focus would be on joint discussions with WoT TF AP / T2T Beyond REST, as well as privacy and security.

Joerg asks for volunteers to help with preparing slides for addressing the IoT vertical perspective.

Dave: the W3C staff contacts and Business Development staff could help.

Joerg: at TPAC we can have a demo session and should start to plan for that.

Implementation Plans

Joerg: can we aim to support demo sessions and a plugfest at TPAC?

Can we agree on some guidance on this and possible links to the IETF Thing to Thing activities.

<kaz> Announcement on demo area during TPAC 2015 (Member-only)

Joerg: at the IETF T2T session, the approach involved agreement on a restricted scenario, and to explore how REST can be applied.

Johannes: We have the idea that three people will create implementations of entities that can communicate with each other and then see how well they interoperate.

This involves setting out some minimal set of ground rules.

Johannes: who is interested in participating in such a plugfest?

Vlad: we have had some experience with this at EVRYTHNG

Simon: we’re doing an implementation with Andrea of T2T.

Johannes: Michael is also working on an implementation.

So we have the potential for several implementations, and need to agree on the ground rules.

Simon: agreeing on a TDL data model vocabulary should be easy enough.

Johannes: we would need to agree on the protocol bindings.

Simon: T2T is focusing purely on REST, which is perhaps oversimplifying the set up

(Dave worries about details of protocol bindings and encodings)

Johanes: I will set up a page on GitHub for us to work on the ground rules for the plugfest.

Dave: we would need to avoid developing servers without any protocols in common.

Johannes: we could create a register (table) listing implementations and the protocols they support.

Joerg asks Dave for his comments.

Dave: this sounds quite ambitious given the Summer break and other commitments. Another idea is to ask people to try to build IoT demos with other people’s Web of Things servers.

Kaz: what does plugfest mean? Maybe Japanese members are also interested in providing IoT demos

Dave notes that he is aiming for having servers ready for other people to use in 2016. October in 2015 is probably too ambitious.

Joerg: if people are just building on REST perhaps they don’t need to do so much coding from scratch.

Dave offers to show people the details of the code he has been working on the Arduino C++ and the NodeJS projects.

Johannes: the aim is to show case the building blocks we are working on.

He shows a slide describing the T2T framework scenario.

Kaz: very interesting idea. How about using some Japanese smart home devices for demo purposes.

kajimoto: Can we also have demos for the WoT principles?

Johannes: if we have some interesting use cases we could indeed feature them

Kaz: some of the Member companies might want to provide their own demos.
... The W3C TPAC management team are currently seeking ideas for the demo activities.

Joerg: how would we collect the proposals, using GitHub?

Kajimoto: we don’t have a concrete WoT specification as yet, so we can set up two kinds of demonstrations (plugtest and IoT demos)

Kaz: maybe Fujitsu, Toshiba, etc., also will have their own demo ideas, and anybody can put their ideas on the GitHub demo wiki

Dave: I am planning on some demos around the Arduino and some common sensors: DHT111, BMP180, HC-SR04 and the AD8232 ECG board.

Joerg: given the Summer vacation period, we need to have the plugfest ground rules in place within a week or two. Let’s say 8th August.

Kaz: so the demo proposals are needed by 15th September according to the W3C TPAC team.

<Kajimoto> <kaz>Demonstration of Application dead line is Sep. 15

Joerg: if there are no further agenda items, then I would very much thank Fujitsu for hosting is, it was really a pleasure and a nice environment.,

Joerg brings the meeting to a close.

- DRAFT -

Web of Things IG, day 3

31 Jul 2015

Attendees

Contents