- From: Kazuyuki Ashimura <ashimura@w3.org>
- Date: Fri, 21 Feb 2020 13:04:03 +0900
- To: public-wot-wg@w3.org
available at:
https://www.w3.org/2020/02/13-wot-arch-minutes.html
also as text below.
Thanks,
Kazuyuki
---
[1]W3C
[1] http://www.w3.org/
- DRAFT -
WoT-Architecture
13 Feb 2020
Attendees
Present
Call 1: Kaz_Ashimura, Michael_Lagally, Kunihiko_Toumura,
Zoltan_Kis
Call 2: Kaz_Ashimura, Michael_Lagally, David_Ezell,
Michael_Koster
Regrets
Chair
Lagally
Scribe
kaz
Contents
* [2]Topics
1. [3]Call 1
1. [4]Previous minutes
2. [5]Issues
3. [6]PRs
2. [7]Call 2
1. [8]Previous minutes
2. [9]Issues
3. [10]PRs
4. [11]AOB
* [12]Summary of Action Items
* [13]Summary of Resolutions
__________________________________________________________
Call 1
<scribe> scribenick: kaz
Previous minutes
[14]Feb-6 minutes
[14] https://www.w3.org/2020/02/06-wot-arch-minutes.html
Lagally: (goes through the minutes)
... Zoltan, have uploaded your slides yet?
Zoltan: yes, created a Pullrequest
Lagally: (continues to look into the minutes)
Kaz: URL for Zoltan's slides?
Lagally: let's merge the PR to fix the URL
[15]PR 434
[15] https://github.com/w3c/wot-architecture/pull/434
Lagally: merged
... and the minutes are approved
[16]Zoltan's slides
[16] https://github.com/w3c/wot-architecture/blob/master/proposals/Device-lifecycle-comparisons.pdf
<scribe> ACTION: kaz to add the URL for Zoltan's slides to the
minutes
Issues
[17]Issue 435
[17] https://github.com/w3c/wot-architecture/issues/435
Lagally: Fleet management use cases
[18]Issue 436
[18] https://github.com/w3c/wot-architecture/issues/436
Lagally: requirements for home network scenarios
... based on the old use case descriptions by the Web&TV IG's
Home Network TF
[19]Home network use cases
[19] https://www.w3.org/TR/hnreq/
Lagally: we should discuss the detail when Chris Needham is
available
... need to use more time for the discussion expected next week
[20]Issue 433
[20] https://github.com/w3c/wot-architecture/issues/433
Lagally: Michael McCool will describe this maybe the second
call
[21]Issue 432
[21] https://github.com/w3c/wot-architecture/issues/432
Lagally: I'm working on this
[22]other issues
[22] https://github.com/w3c/wot-architecture/issues
Lagally: we still have other issues and need volunteers
... e.g., data streaming (issue 387)
... possibly could get comments/ideas from NHK and the MEIG
guys
[23]https://github.com/w3c/wot-architecture/issues/25
[23] https://github.com/w3c/wot-architecture/issues/25
Lagally: WoT in the Web browser
... how to involve browser vendors?
Zoltan: many discussions about that so far
... possible JavaScript library which uses Fetch, etc.
Lagally: (adds comments to Issue 25)
Zoltan: got discussion with browser vendors but it seems
generic API-based approach would not be promising for WoT
... there is no browser vendors who want to implement our ideas
like extensions based on the fetch API or the WebAssembly
solution
... so at the moment, we might want to generate polyfill
implementations
Toumura: browser is a consumer of devices?
Zoltan: primarily a consumer but could be a producer
Toumura: ok
Lagally: so nobody is implementing WoT natively within the
browsers
... but do we have any concrete use cases for browser
integration?
Zoltan: browsers included in sensor fusion
... myself don't have a use case which include a Web browser as
a producer
Lagally: no concrete use case for exposing things yet
... can consider sensor fusion/mashup at the moment
... this use case is already covered, however,
WebAssembly/WebWorkers would be a different way to implement
Kaz: when we say "web browsers", we should think about use
cases for both their hardware side (when empbedded within a
device) and the software side (when installed on a
PC/smartphone)
... and then think about what kind of features to be used for
which use cases
... like digital TV set as a device and speech interface on a
smartphone
<zkis> Example for using node-wot from browser:
[24]https://github.com/eclipse/thingweb.node-wot#as-a-browser-l
ibrary
[24] https://github.com/eclipse/thingweb.node-wot#as-a-browser-library
Zoltan: the above is the link for using node-wot on a browser
PRs
[25]PR 428
[25] https://github.com/w3c/wot-architecture/pull/428
Lagally: Digital twin use cases
... seems it's mixed up and need some cleaning
... let's concentrate on the digital-twin-use-case.md for now
[26]digital-twin-use-case.md
[26] https://github.com/w3c/wot-architecture/pull/428/files#diff-28a786469d3456b28e0e345541c9d866
Lagally: (goes through the "Motivation" section)
... terminology for "digital twin"
... also called as "twin" or "shadow"
... (goes through the "Expected Devices" section)
... (Expected Data)
... (Dependencies)
... WoT Architecture, TD, Profile, and possibly Scripting?
... need to have a way to extract the data model
... though digital twin itself uses abstract-level data
... (Description)
... better visibility, accurate prediction, what-if analysis,
documentation and communication, integration of disparate
systems
... (Variants)
... * virtual twin
... ** intermittent connectivity, protocol abstraction,
business rules
... ** and examples
... * predictive twin
... * twin projections
... that was quick run-through of the digital twin use case
... regarding the "Gaps"
... we have ranges for values and units already
... but we don't currently have a way to describe simulated
behavior
... also templating mechanism
... and data model separation from the binding
... we still have 10 more minutes, so let's see another use
case
[27]PR 431
[27] https://github.com/w3c/wot-architecture/pull/431
Lagally: there is only one use case description here
[28]X-Protocol Interworking.md
[28] https://github.com/w3c/wot-architecture/pull/431/files#diff-912132dcd97d0282ba66431f9bdeea92
Lagally: Cross protocol interworking
... (Motivation)
... bridge more than one protocol
... for interoperability accross protocols
... possible limitation based on the capability of each
protocol
... if we want to think about interoperability, we need to
consider that point
... (Description)
... a couple of examples here
... smart home environment
... industrial environment
... smart city environment
Zoltan: wondering who would really use multiple protocols at
once
... we have a specific/concrete use case already
... and do we want to have this kind of generic use case as
well?
Lagally: would think about a device which exposes its
capability in multiple ways
Zoltan: we could use native bridges between WoT and native
connection
... if this approach is successful, it would be great
... but a good homework
Lagally: I see a big value
... we could have wide variety of devices
... and a lot of implementation complexity
Kaz: think we should clarify the concrete descriptions about
each example use case a bit more
... and then we could think about (1) a category of use cases
which includes "cross-protocol interworking" or (2)
requirements for "cross-protocol interworking" based on the use
cases later
Lagally: ok
... (adds comments to PR 431 about that point)
[29]Lagally's comments
[29] https://github.com/w3c/wot-architecture/pull/431#issuecomment-585601403
Toumura: in that case, probably we should think about some
concrete use case like smart home for Discovery purposes as
well?
Lagally: right
[Call1 adjourned]
__________________________________________________________
Call 2
<scribe> scribenick: kaz
Previous minutes
[30]Feb-6 minutes
[30] https://www.w3.org/2020/02/06-wot-arch-minutes.html
Lagally: (goes through the minutes)
... discussion on the state diagram
... onboarding/discovery and additional things
... and device lifecycle
... Zoltan did great job for the survey
... end-to-end security by McCool
... Zoltan's slides now available
... PRs for use cases
... can approve the minutes?
Koster+David: fine
Lagally: minutes approved
[31]Zoltan's slides
[31] https://github.com/w3c/wot-architecture/blob/master/proposals/Device-lifecycle-comparisons.pdf
Issues
[32]Issue 436
[32] https://github.com/w3c/wot-architecture/issues/436
Lagally: use cases/Requirements from the Web&TV IG's Home
Network TF
... which of those could be still relevant?
[33]Requirements for Home Networking Scenarios
[33] https://www.w3.org/TR/hnreq/
Lagally: 9 years old
... Opera, CableLabs, Ericsson, Telecom ParisTech, BBC, Samsung
and Sony were involved
Koster: might be able to contact the Editors
Lagally: next week, Chris Needham, one of the MEIG Chairs, is
expected to join the Architecture call
... let's revisit this issue then
... if you have time, please look into the document
[34]Issue 25
[34] https://github.com/w3c/wot-architecture/issues/25
Lagally: WoT in the Web browser
... 2.5 years back
... had discussion with Zoltan during the first call today
... it seems browser-native implementation is currently out of
scope
... there is a possibility of using a Web browser as a Thing
exposer
David: should keep this issue open
... someone should talk with TAG and browser vendors
Lagally: what specific question should we ask them?
David: would see McCool's advocacy on edge computing based on
service worker
... wish his comments
Kaz: we already have several pieces but I think we should think
about both the device side and the software stack side of the
Web browser (as I mentioned during the first call today)
Lagally: (adds comments to Issue 25)
... this may be related to service workers. we should think
about separating physical device (hardware) and Web browser
(software stack)
Koster: maybe embedded browsers use web socket protocol to
interact with outside servients
... that is a possible use case
... e.g., TV could be a smart Things hub
Lagally: would be good to generate a concrete use case
description
Koster: will do that
Lagally: (assigns Issue 25 to Koster)
... the other issues on discovery (425), digital twin (426),
edge device (427), etc.
... but let's see PRs
PRs
[35]Digital Twin use case
[35] https://github.com/w3c/wot-architecture/pull/428
Lagally: (goes through the use case description)
... (Motivation)
... representative of a real Thing
... you can use it to collect data including time series of
data
... scaling up to multiple devices
... can monitor everything
... can be called as "twin" or "shadow"
Koster: we have an end-point app
... would be a same thing
Lagally: cloud API as interface to the virtual thing?
Koster: yes
... API for cloud input
Lagally: ok
... there are several connection patterns
... can use the model for prediction
... and create a simulator
... can create a what-if analysis system
... also a combination of simulators
... also integration of simulators and physical devices
... (Virtual Twin)
... intermittent connectivity
... (Predictive Twin)
... you can use machine learning for precise prediction
... (Twin Projections)
... questions?
(none)
Lagally: can quickly talk about the "Gaps"
... managing multiple devices using the template mechanism
... various instances could be involved but we don't have a
template mechanism for that situation
... (adds comments mentioning we reviewed this PR)
... (and merges PR 428)
[36]PR 431
[36] https://github.com/w3c/wot-architecture/pull/431
Lagally: we need some more stakeholders like McCool for this
discussion
... so would defer this to next week
AOB
David: DID WG meeting in Schiphol
... suggested they coordinate with WoT
<dezell>
[37]https://www.w3.org/2019/did-wg/Meetings/Minutes/2020-01-29-
did#section4
[37] https://www.w3.org/2019/did-wg/Meetings/Minutes/2020-01-29-did#section4
David: includes the link to the slides
... retail industry should be open
... that is very important
... many of the WoT use cases are applicable to retail industry
... encouraged Conexxus guys to generate TDs
... will also go to the Helsinki f2f
Lagally: tx for the quick wrap-up
... should we create a specific retail use case?
... would work together to generate one
David: (mentions some possible use case)
... this kind of control and safety
Lagally: ok
... pretty concrete use case
... let's write it down based on the use case template
David: happy to help
<scribe> ACTION: Lagally and David to work on a use case for
retail services
Kaz: completely agree with you both that we should work with
the DID WG, and should create concrete use cases starting with
retail services
Lagally: aob for today?
(none)
[Call 2 adjourned]
Summary of Action Items
[NEW] ACTION: kaz to add the URL for Zoltan's slides to the
minutes
[NEW] ACTION: Lagally and David to work on a use case for
retail services
Summary of Resolutions
[End of minutes]
__________________________________________________________
Minutes manually created (not a transcript), formatted by
David Booth's [38]scribe.perl version 1.154 ([39]CVS log)
$Date: 2020/02/17 08:58:42 $
[38] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
[39] http://dev.w3.org/cvsweb/2002/scribe/
Received on Friday, 21 February 2020 04:04:10 UTC