<kaz> scribenick: McCool
Koster: let's review checklists
... participants and security
... let's also discuss semantic interop
... reserve last 20m for sematics
<mjkoster> General WoT Plugfest Participant Questionnaire
Koster: PR has been accepted for
participant questionnaire
... want to collect this information so we can see what people
are thinking
... would also like to have a lot of scenarios
... we also want to document both what we plan to do
... AND what we would do for a real product deployment
... for example, we may do something manually in an ad-hoc
way
... that in a real deployment we would automate
... even if you are just bringing connected things, but not a
client
... still good to document
... (discussion of application scenario section)
McCool: what is the timeline on the scenario questionnaire?
Koster: probably next week
... good to have a document that is more focused on use cases
... I also plan to work up a more concrete example...
... is also a survey of what components people are bringing
... important if you want to build scenarios that cross
components brought by different people
... as an additional topic: how do we add extrinsic semantics
... are looking at that for the IETF Hackathon the weekend before IETF
... March 17-18
... what we want to look at there is how to do discovery
... how can we do some complementary things
... some additional information needed to make sense of
data
... thing directory would be the obvious place (Koster: IMO) to
add this information
... next topic is semantic integration
... what ARE the types and capabilties, in English?
... what information do you want to express?
... what kind of information would you want to find out?
... then there's the issue of adapting to engineering units.
etc.
... now protocol bindings
... what protocols you plan to use, and how
... are you building a driver, or using the information in the
protocol binding?
... how are observables/events handled in particular? (MQTT, SSE, WS, LP, etc)
... then how do you generate bindings for exposed things
... next topic: system level
... proxies, nat traversal, caching, translation, etc.
... how does it interact with other things and services
... topic security: a few high-level stuff, but look at security questionnaire
McCool: under accessiblility, we
probably need to thing about semantic tagging
... for example, to indicate physical events and ui options
... to know when we have to worry about translation into alternative sensory modalities
<Zakim> kaz, you wanted to ask about possible template for use cases
<kaz> use case template
Kaz: there is another template
for use cases
... the multimodal interaction working group looked at this
... and developed a template
Koster: nice, we can review that at perhaps reuse
Kaz: work itself was regarding
accessibility and user interface
... but we can extend it to the more general case, using the
template
Darko: when do you expect to
obtain some results?
... it looks very extensive, but what is the timeline?
... how can we get the results as soon as possible?
Koster: optional, easy to update,
etc.
... it's all optional
... should make it easy for people to update information
... we still have six weeks, but...
Darko: what format? Is the wiki the right place?
Koster: if there was any easy way to do a survey...
Darko: could use google
... using doodle is a little awkward
Kaz: there is a questionairre functionality we can use in W3C
<kaz> [ we can use hyperlinks using W3C WBS (Web-based straw-poll system like https://www.w3.org/2002/09/wbs/1/WoTF2F201803/?login)
Koster: would be nice if it can also do file attachments
McCool: ok if people can just hyperlink to their own github
... I think two weeks out do get some initial drafts would be good
Darko: but it is important to be
able to see answers from others
... it is useful so that one can look at others for
examples
Koster: so... we need to go look at tools, can do that offline
<kaz> Security questionnaire
<inserted> scribenick: kaz
McCool: (goes through the
questionnaire)
... would like to add figures
... there is picture in the Security note
... would like to check if we can use SVG
... concrete example would be helpful
Koster: concrete example and implementation guidance would be useful
McCool: right now the security note
is quite high level
... need concrete example on provisioning key, etc.
... let us know if anything to be added
... or need clarification
<scribe> scribenick: McCool
Soumya: project with another
company
... set up tools that can insure interoperability
... will create a couple of slides that can share in the next
meeting
Koster: can you discuss the idea
of testing semantic interop test?
... what is the system being tested, what are the inputs and
outputs?
Soumya: better to let me put it
down on some slides first
... better than saying things now
Koster: are you available for the Friday meeting?
Soumya: yes, I am
<kaz> (4pm in Europe)
Koster: that would be good time
to discuss the conceptual aspects
... also participation in the Eurocom testing
Darko's slides (esp. p5 on "Next Steps")
Darko: let me show what we are
thinking about for the next steps for semantic interop
... want a set of scenarios
... extend scenarios with "WoT challenges"
... which in order to be implemented need semantic
integration
... provide semantic artifacts (queries, capabilities, recipes,
etc) needed for these scenarios
... in addition, Eurocom may want to participate with some
other semantic models
... rather than using iotschema
... may want to use other semantics models...
... iotschema in particular has three levels
... can talk about interoperability at these levels
... but other models may have other levels
Soumya: think that iotschema is a good place to start
Darko: that makes things
easy...
... but we can look at additional capabilities
Koster: I did a presentation that
added a "thing type"
... that would be useful
... I think it's far more useful for capabilities
... but some people still want to think in terms of
... types
Darko: it would be great to get
feedback from telcos on templates
... so we can direct our work in the direction of results from
participants
Koster: thing we discussed in
Wishi group earlier this week
... extrinsic information
... is this something we want to add?
... seems to be in scope, but we haven't really discussed
McCool: concrete examples?
Koster: location is one for sure
Darko: SOSA
... location, also "feature of interest"
... eg: what are you measuring?
... is a pattern, not everyone will want it
... but will be useful for some people
Koster: location is kind of an
aspect of FOI
... want to know which capabilities are applied to which
FOIs
Darko: sometimes the FOI is more
important that the capability
... for instance, I may want to find out everything I can about
some FOI
... but don't necessarily know in advance what things are being
sensed about that FOI
Benjamin: is this integrated into Iotschema.org
Darko: not integrated/public
yet
... SOSA would be an additional pattern, as opposed to a
horizontal layer
... in order to keep it simple
... but we need to go back and look at SOSA pattern
Koster: ok, will take this up in the Friday meeting
<inserted> scribenick: kaz
McCool: when is the iot.schema meeting?
Koster: once charter, etc., is clarified, would have one
McCool: github, etc.?
Koster: there is a iot.schema repo
McCool: place for pullrequests?
Koster: yes, there is a place for
that purpose
... ontology pattern is very simple
... iot.schema is based on ontology
McCool: legal framework?
Koster: charter would clarify
license, etc.
... you need to part of the W3C CG
McCool: we're already part of W3C
Koster: the venue will go under the W3C CG framework
<inserted> scribenick: McCool
Koster: there have been some
important decisions made recently
... interactions seem to be central
darko: properties may not be
top-level
... also mention shape constraints
... some companies some things may be a property
... for others may be an action
... be more generally, have interaction patterns
Koster: let's put that down as a
topic for the next iotschema meeting...
... our hour is up
kaz: regarding CG for
iotschema...
... not created yet, correct?
Koster: right. first thing is to
create the charter
... thought we could use the existing WoT CG
... timeline is... we write up the charter
... and then have to decide whether we want to create a new CG
or reuse the existing WoT one
... any more business? No? then adjourn...
<kaz> [adjourned]