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

<kaz> scribenick: McCool

quick updates

kaz: status of fpwd

monday, found some minor issues, talked with editors and got clarifications

will be working with the webmaster, and will get them published tomorrow

matthias: dave offered to write a blog

we are still deciding where it will go

<inserted> scribenick: kaz

matthias: mccool, workshop status?

mccool: 2 proposals
... talked with the other proposer
... updated our proposal a bit
... to merge the 2 workshops together
... will circulate the updated proposal to the group
... we have to submit a paper to the NDSS workshop ourselves
... the other is IoT security
... security across various standards
... focus on Web of Things
... going to have to work on another proposal for IEEE workshop

matthias: what is the motivation for the 2nd one?

mccool: focus on standards
... NDSS one is broader
... would have more focused workshop for IEEE
... for NDSS workshop, paper deadline is in 2 months
... IEEE S&P is May
... much more time for review
... would generated an updated proposal for IEEE
... presenting at NDSS and get people involved
... and invite them to IEEE one

<inserted> scribenick: McCool

mccool: I will distribute the revised proposal; could use some more topics related to WoT if you have any suggestions

plugfest and tpac and next steps for wg

matthias: now we have released fpwd
... some changes; want to get plugfest aligned with the fpwd
... so one goal is that we want to update node-wot to match the fpwd
... would also like to target the scenarios as well
... does anyone see any blockers for updating their implementations
... to match the fpwd
... asking implementors and leads

kajimoto: what is viable target for plugfest implementors?
... consensus on what is viable
... thing description, changes at the plugfest every time
... so what is standard to define interoperability
... defined current practices as viable... but now we have fpwd

mccool: do you want to define a subset of the fpwd for the plugfest?

kajimoto: yes
... first step, update current practices to fpwd
... for thing description
... adding security support
... define purposes for next plugfest to prove concepts

kaz: proposing organizing and orchestrating plugfest as a starting point
... but not sure we need a profile or a minimum subset
... would like some clairification

matthias: about current practices document, have some responses

kajimoto: use cases are good... but we should have some guidelines or standards for how to describe our semantics
... the current practices document was good to have something to target
... but now we have the td only, less specific
... would like to update the current practices document to match the fpwd

matthias: overall, have plan that current practices needs to be adapted
... but main content is now in fpwd; normative content is there
... but there are some primary and secondary features; we should define which is which
... and it would also be useful to set some goals, for instance
... to describe their security features, test thing, use cases
... would this solve some of your issues?

kajimoto: yes

kaz: matthias has mostly covered my point
... we probably should update the current practices with whatever is not covered in other fpwd documents

plugfest update

matsukura: plugfest update, sharing screen

<kaz> Matsukura-san's slides

See issue #346

<kaz> Issue 346

<McCool_> scribenick: McCool_

matsukura: three kinds of servient
... exposed thing, consumed thing, servient (gateway)
... also integration model, 3-layer, 4-layer
... issues: interfaces between servients, authentication, discovery, TD exchange,
... NAT and firewall traversal
... event operation
... finally, TD management issues
... for TD management, how to create URI, who manages TD and how
... call for participation, want to accelerate preparation, share information
... so please share information about what you plan to do

dape: mentioned authentication... some content used quite a while ago
... but not sure if it is still the state of the art

<kaz> current practice doc

matthias: mechanism in current practices is just one way of how it could be done
... but really we want to describe what is existing
... rather than defining our own mechanisms
... main comment is also about management of TDs
... we have a Thing Directory... at the moment is standalone
... but could also be merged with a (cloud) servient

<Zakim> dape, you wanted to CP and authentication

matthias: but might also have a sparql endpoing

to support semantic search

matsukura: some variations are possible

matthias: could be included
... logically it could be a separate component

<inserted> scribenick: kaz

mccool: exposed OCF device as a Thing
... including CoAP as protocol
... question about security metada
... various decision to make about devices
... mechanisms for authentication
... we could for further
... we should pick examples

<McCool_> mccool: I think we should pick some other platforms and figure out how to support them

<inserted> scribenick: kaz

mccool: interested in Amazon Alexa interface
... that's doable but don't want to identify any devices
... need to find out how to handle discovery
... more than one Thing Directory

matthias: Thing Directory could be a good entry point

mccool: multiple directories
... starting with one entry point
... can be a working assumption

<inserted> scribenick: McCool_

mccool: I want to build an Alexa service that can discover things and talk to them in a generic way

matthias: good working assumption is one Thing Directory
... if there is more than one, they could still have a single front end and delegate, but we don't plan to do that for plugfest

kaz: do we need an additional document?

matsukura: yes, we need an additional document in addition to the PlugFest wiki
... need a specification
... we do want another document... but were? wiki? markdown? expand current practices document?

matthias: propose re-expanding the current practices document
... but ultimately should go into arch... but current practices can be used to draft things

<Zakim> dsr, you wanted to comment on work for TPAC on plain JSON and WebSocket based messaging protocol and ask about specific goals for interoperability testing

dsr: hope to collaborate with mozilla... websockets, etc
... mozilla hopes to bring some actual devices
... dsr to bring some simulated devices

<dsr> see: https://lists.w3.org/Archives/Public/public-wot-ig/2017Sep/0006.html

dsr: wants to see us as a group to clarify our goals for interop testing
... dsr also looking at "plain JSON" description in collab with mozilla
... unfortunately ben is unavailable for now... dsr not ready to share

matthias: would be better to share early

dsr: is plain JSON aspect, is also web socket aspect

matthias: can the ws aspect be aligned with panasonic?

kajimoto: can you explain the current situation with ws communication?
... (asking panasonic members)
... have a document that panasonic sent to mozilla
... need to understand the differences in the approaches, is useful

tokuyama: have to confirm these topics with rest of panasonic team and communications with mozilla
... but waiting for response from mozilla
... have already sent some material to them

matthias: best if moz and panasonic can collaborate directly

kaz: probably should send official invitation

dsr: we already talked to them about this
... they are just waiting for internal authorization
... so we already talked about this; can send another email, but already aware

kaz: point is they should enter their information on the wiki etc.

dsr: yes, but they are still waiting for internal approval

kaz: ok, will send a reminder

dsr: but they are away until the end of the month

kaz: please cc me when you contact him. I'll also cc you and the Chairs when I contact him.

dsr: ok

zkis: security and provisioning; even though we are not provisioning them, we still need to get client keys, etc. to talk to OCF devices
... so that is non-trivial
... we will have to provide some guidelines on how to do it

mccool: and I'm not sure where we find OCF implementations that actually do this...

matthias: anyone who brings device needs to figure this out
... the credentials can't go into TD, they have to be provisioned somehow
... ok, out of time...
... please do update info
... even siemens has to do this

<kaz> plugfest wiki

matthias: and next to-do is to update the current practices

<kaz> current practices

matthias: with focus on use cases

mccool: what is timing of node-wot updates?

matthias: will only be done just before the plugfest...
... will have to use current systems, and intercept
... need to work on a roadmap after cleanup

mccool: good topic for next meeting

matthias: may not be available next two weeks...

dape: hope we can get it done sooner than two weeks before the plugfest

matthias: would also be good to get test thing done early
... and that has a dependency on the scripting api

mccool: also, McCool to chair next two meetings, Matthias will not be available

<kaz> [adjourned]