08 Mar 2017


See also: IRC log


Achille, Daniel, Feng, Dave, Kaz, Michael_Koster, Michael_McCool, Nan_Wang, Nimura, Ohura, Sebastian, Taki, Tokuyama, Toumura, Yamada, Yingying, Yongjing, Zoltan, Matthias
Yongzing, McCool, Matthias


<scribe> scribenick: kaz

<scribe> Agenda: https://www.w3.org/WoT/IG/wiki/Main_WoT_WebConf#Agenda

OCF Liaison update

mm: OCF update
... MoU not yet
... OCF Liaison will be pretty busy for the next a few months
... btw, anyone who are members of OCF here?
... your help is welcome for the liaison

yz: my company is a member of OCF but I myself can't join the effort
... let's talk about that offline
... any other update?

Minutes from TF calls and this main call

kaz: confirmed minutes handling during the Chairs call today:
... - sending the minutes from the WG TF calls to the WG Members list
... - sending the minutes from the IG TF calls to the IG Members list
... - sending the minutes from the joint IG/WG main calls to the IG/WG Public lists

yz: anything else?


[Dave Raggett] Thing description requirements from a survey of OCF, oneM2M, BACnet and EchoNet

dsr: will put the slides somewhere reachable
... the project name is "CREATE-IoT"
... Survey of IoT Platforms

(Dave goes through the slides)

dsr: [ Survey of IoT Platforms ]
... work funded by the European Commission through the Create-IoT
... [ Coverage ]
... started to look at OCF/OIC 1.1, oneM2M, BACnet, ECHONET, Smart Device Template, ...

mm: what about MQTT, DDS, etc.?

dsr: on the roadmap
... [ Methodology ]
... read the IoT platform specs and look for common patterns
... [ OCF ]
... starting with OCF
... OIC 1.1
... RESTful interface over CoAP and HTTP
... spec uses Restful API Markup Language (RAML)
... modularization ongoing
... each resource is defined in RAML
... RAML is not used at run-time
... RAML is verbose with a high degree of predictability
... OCF spec online

mm: what about oneIoTa model?

dsr: not aware
... [ OCF OIC 1.1 release ]
... 63 resources/15 Devices
... [ oneM2M ]
... RESTful interfaces over CoAP, HTTP and MQTT
... defines devices in terms of composition of modules
... formalized in XML
... home appliances and information models

<McCool> Look at www.oneiota.org for a list of more up-to-date OCF/OIC data models

dsr: broad range of devices
... specification online
... [ oneM2M ] (contd.)
... 40 modules/13 devices

yz: first slide of oneM2M
... binding protocol now includes WebSocket as well
... DDS also
... there are tools as well
... service function models are represented as resources

mm: OCF is similar

dsr: what level of commonality is there?

yz: not 1-to-1 mapping yet

mm: 2 issues
... OCF define common resources
... roughly compared to oneM2M
... trying to work together
... the question is there are existing devices

yz: oneM2M was defining device models as an abstract ones
... there is another source, OMA
... oneM2M uses ontology but it doesn't govern the ontology
... there are some examples
... base ontology and domain specific ones

dsr: we can learn it
... [ BACnet ]
... building automation
... core specs define information model, metadata and protocols
... a bit different technology
... BACnet object corresponds to properties
... BACnet objects correspond to WoT properties
... BACnet properties correspond to WoT metadata
... Michael Newman's book available
... [ ECHONET ]
... home automation and personal health
... 22 properties
... 44 sensors
... [ ECHONET Device Groups ]
... air conditioners
... housing/facilities
... cooking/household
... health
... management
... ...
... [ HGI Smart Device Template ]
... defined by the Home Gateway Initiative
... XML Schema for smart home devices available online

yz: question on HGI
... home appliance model is available in oneM2M as well
... using smart device template

dsr: [ REST over CoAP and HTTP ]
... Roy Fielding's representational state transfer design pattern for stateless protocols
... different kind of abstraction
... [ Results ]
... [ Emergent Requirements ]
... strong support for objects with properties, actions and events
... object model for things
... common data types: boolean, string, integer, number
... less common data types: enumerations, unions, vectors, ordered/unordered collections
... some calls that "array"
... application types in terms of predefined types
... things as first class types
... compound/nested properties
... compound arguments for actions and events
... Metadata:
... min/max for numbers and integers
... read-only/read-write
... required/optional
... units of measure including scale factors
... static/changeable

mm: what about constraints of types?
... begin/end of some specific characters?

dsr: we should have mapping mechanism

mm: there is OIC bridging mechanism

dsr: characters and character sets as well

mm: regular expression too

dsr: sometimes metadata is static, but sometimes changeable
... mapping of enumerated values, property names, etc., to IDs
... Google protocol buffers show examples
... how versioning relates to the history
... [ Syntactic Modularity ]
... specs like to define things
... [ Semantic Modularity ]
... not used by any of the IoT platforms covered so far in the survey

mm: there is a modularity thing
... either supporting flattening or not

dsr: should look at use cases
... [ Next Steps ]
... plan to flesh out this survey as a CreateIoT report
... whilst also extending it to a broader IoT platforms

kaz: do you want to put the results of this survey into the tech landscape document
... would be a great input for that

dsr: yes

mm: good timing to collect data as well

yz: pattern of interface too
... sometimes need extended pattern

dsr: if there is already some driver for an IoT platform
... we don't need WoT
... but if users want to control the device, may need WoT

<McCool> (sorry, I have to go...)

yz: other questions?
... question on [ Semantic Modularity ]
... in oneM2M, there is something related
... semantic annotation
... semantic classes should be validated

dsr: LD TF are discussing that

yz: can provide ideas from the oneM2M side
... another question on [ Next Steps ]
... targets include OMA, etc.?

dsr: need to leave now
... will put the slides on the wiki

yz: tx

Osaka F2F

registration site

f2f wiki

kaz: mentions the registration site and wiki above
... sent papers to Yongjing

ohura: list of recommended hotels on the wiki

yz: the first hotel agora hotel was not available for my stay...
... asks about the papers for his visa procedure

kaz: clarifies the requirements

yz: any other questions?


WG reports

sk: give some report
... first call on Monday
... talked about logistics for spec generation
... GitHub, etc.
... already set up the first version
... update from the Santa Clara meeting
... will continue the discussion

dp: the content of the current practice document has been copied to the TD area
... what would happen to the old current practice document?

WG TD draft

IG current practice document

sk: we need to discuss how to handle it

zk: ok with keeping the old doc

dp: as an enter point for the upcoming WG draft?

kaz: we can keep the IG current practice document as is
... and also if we want, we can use the Current Practices area to record our progress/results from PlugFest demos

zk: yes, current practices are mainly records of PlugFest

yz: we need to check with Matthias as well

kaz: which would be better/easier for the Editors, (1) simply put the PlugFest results on the wiki or (2) once record the detail in the Current Practices document

yz: let's continue the discussion on the Editors ML

Communication TF

yc: chatted with Dave and the Mar/Comm Team
... about press release testimonials
... the first public wd should be published in April
... and would like testimonials from Members
... the second topic is recruiting for WG
... 218 people for IG while 69 for WG
... Alibaba has just joined
... contacting some more Chinese Members
... will reach out other Members as well
... especially for those who are participating in IG

yz: ok


yz: any other updates?
... security, scripting?

kaz: gives update from the Script TF
... talked about GitHub issues

GitHub issues on Script

kaz: Nimura-san pointed out we need more detailed use case descriptions
... that could be a separate informative document or part of the main use case document

yz: ok
... any other business?


[ adjourned ]

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2017/03/08 18:50:40 $