- From: Kazuyuki Ashimura <ashimura@w3.org>
- Date: Thu, 28 May 2020 22:58:31 +0900
- To: public-wot-wg@w3.org
available at:
https://www.w3.org/2020/05/11-wot-discovery-minutes.html
also as text below.
Thanks a lot for taking the minutes, Andrea!
Kazuyuki
---
[1]W3C
[1] http://www.w3.org/
- DRAFT -
WoT Discovery
11 May 2020
[2]Agenda
[2] https://www.w3.org/WoT/IG/wiki/WG_WoT_Discovery_WebConf#11_May_2020
Attendees
Present
Andrea_Cimmino, Cristiano_Aguzzi, Farshid_Tavakolizadeh,
Kaz_Ashimura, Kevin_Olotu, Kunihiko_Toumura,
Michael_Koster, Michael_McCool, Takahisa_Suzuki,
Tomoaki_Mizushima, Zoltan_Kis
Regrets
Christian_Glomb
Chair
McCool
Scribe
acimmino, kaz
Contents
* [3]Topics
1. [4]Agenda and guests
2. [5]Previous minutes
3. [6]Use cases
4. [7]Requirements document
* [8]Summary of Action Items
* [9]Summary of Resolutions
__________________________________________________________
<inserted> scribenick: acimmino
Agenda and guests
Kaz set the agenda link in the chat
Christiano Aguzzi is invited as a guest for OAuth2 discussion
meeting starts, looking at the agenda
Review of PR is first issue
then, a propose resolution about pagination
then, a discussion about first draft will take place
also, added to the agenda Fujitsu directory implementation
discussion and UPM UC
Previous minutes
[10]May-4 minutes
[10] https://www.w3.org/2020/05/04-wot-discovery-minutes.html
McCool explains how the minutes of the meeting work
correction of attendees names for meeting of 04th of may,
"Cristiano_Cimmino" should be "Andrea_Cimmino"
identification of other typos in point 5 of the index, same
meeting minutes
another typo "addition use-case" should be "additional
use-case"
no comments on last meeting minutes
therefore, they are approved to be published
<kaz> [typos are just fixed]
Use cases
Meeting keeps with discussion of ongoing UC
first, the Fujitsu UC
a link to prior work has been added
a new subdirectory as been added with UC information
included a diagram for the UC of Fujistsu
the UC will be merged
<kaz> [11]PR 19
[11] https://github.com/w3c/wot-discovery/pull/19
no objections about merging the Fujitsu UC, therefore, approved
for merging
now UC of UPM
problem with pull due to the account of requester
<kaz> [12]PR 20
[12] https://github.com/w3c/wot-discovery/pull/20
markdown format problem due to template
adding note to discovery UC template so in future UC will rely
on the template updated n the wot-arquitecture
UC must follow the template specified in the wot-architecture
git repo
commiting UPM use case
Andrea explains the UC
<inserted> scribenick: kaz
Andrea: the idea is having distributed network
... offering data for different domains
... in this case, you want to solve discovery task
... relying on SPARQL query
... you'll have a lot of meta nodes
... there would be some trouble
McCool: delegate queries to other SPARQL end points?
Andrea: yes
... only to one-hop middle node
... just the one with direct connection
McCool: sounds reasonable
... capturing the requirements is important
... we can merge this PR 20 itself
<acimmino> trying to merge UC, despite technical problems
McCool: and then create another PR for architecture
<inserted> scribenick: acimmino
andrea needs to merge his account with the W3C
McCool suggest how to solve it in the comment of the pull
request
andrea must change a word and re-commit
Follow, meeting proceed checking the Fujitsu UC
Suzuki explains the UC
human discovers directories, which contain the TD of smart
appliances
some requirements are derived from the UC, they are specified
in a subsection of the UC file
some implementation clarifications are specified in a
subsection of the UC file
<kaz> [13]Fujitsu's discovery implementation
[13] https://github.com/w3c/wot-discovery/tree/master/prior-work/fujitsu
Diagram with discovery performed in this UC is explained
McCool: comments about the UC, the protocols used DHCP, and the
diagram for registration
McCool asks about Thing discovery diagram
McCool: seems unclear how the mDNS is performed
McCool suggest some changes
McCool: it would be really useful to have specification for
REST API (diagram consume phase)
... suggest to use standardised URL for describing entities in
REST API
... likely not all devices will be able to implement it
... question about query support, do this UC actually has it?
Suzuki: in this example query language was the one supported by
MongoDB
McCool: how keyword search can be done?
Suzuki: maybe query search is not supported
McCool: ...in general, here we have three things WoT name
space, Device, and Directory
... ...which is a reasonable structure
... ...next time deeper discussion about UC to catch
requirements
skip the propose resolution about pagination
jump into the requirement document due to time left is not much
Requirements document
need to capture requirements from design phase
<kaz> [14]Requirements (so far)
[14] https://github.com/w3c/wot-discovery/blob/master/requirements.md
McCool suggest to have a design document, from which capture
design decision and link them back to resolution
approved the creation of design document in the git repo
document has two subsections within Design section: Two-Phase
Architecture and Standardized Directory Service
McCool provides description for Two-Phase Architecture
authentication plays an important role during the discovery
McCool adds this to the Two-Phase Architecture description
no objections to include Resolution link in the notes
<McCool> proposal: The discovery process should have two main
phases: an introduction phase and an exploration phase. The
introduction phase can use existing discovery mechanisms and
its main output should be a pointer to an exploration service,
but it should not provide detailed metadata for actual devices.
Instead, the exploration service should provide detailed
per-device metadata after authentication.
no objections in making this resolution
RESOLUTION: The discovery process should have two main phases:
an introduction phase and an exploration phase. The
introduction phase can use existing discovery mechanisms and
its main output should be a pointer to an exploration service,
but it should not provide detailed metadata for actual devices.
Instead, the exploration service should provide detailed
per-device metadata after authentication.
in future meetings something similar will be done for
specifying the Standardised Directory Service
<kaz> [15]newly created (preliminary) design note
[15] https://github.com/w3c/wot-discovery/blob/master/design.md
no objections in merging the design file
no more issues for this meeting
meeting is end
<kaz> [adjourned]
Summary of Action Items
Summary of Resolutions
1. [16]The discovery process should have two main phases: an
introduction phase and an exploration phase. The
introduction phase can use existing discovery mechanisms
and its main output should be a pointer to an exploration
service, but it should not provide detailed metadata for
actual devices. Instead, the exploration service should
provide detailed per-device metadata after authentication.
[End of minutes]
__________________________________________________________
Minutes manually created (not a transcript), formatted by
David Booth's [17]scribe.perl version 1.154 ([18]CVS log)
$Date: 2020/05/25 14:09:58 $
[17] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm
[18] http://dev.w3.org/cvsweb/2002/scribe/
Received on Thursday, 28 May 2020 13:58:29 UTC