W3C

WoT-WG - TD-TF

25 Nov 2020

Agenda

Attendees

Present
Kaz_Ashimura, Michael_McCool, Michael_Lagally, Sebastian_Kaebisch, Taki_Kamiya, Ege_Korkan, Michael_Koster, Tomoaki_Mizushima
Regrets
Chair
Sebastian
Scribe
sebastian, kaz

Contents


new WoT web page

<inserted> scribenick: sebastian

need some short abstract about the task forces

including the task forces leaders

<kaz> Welcome page

McCool: we should use a template for each section

<kaz> TD-TF

<kaz> Ege's issue 85

McCool: there is an issue from Ege that gives an overview about the current content of all the task force wiki pages

McCool: we should agree on a template
... who can help here to create a template?

Ege can provide a template.

Ege: Which points should be covered?

McCool: Main page has the most contents and covers many points

Ege: Discovery can be also used to create template

Kaz: Daniel developed an CSS file for the webpage. Shall we use this for the template as well?

McCool: Prefer the wiki style
... we should keep the wiki structure

Kaz: we should have the detail structure of the MD files for task forces
... should discuss this in the marketing call

Ege: https://w3c.github.io/wot-marketing/activities/tf-td/ is html file which will be not changed
... only the content of the wiki page will changed like the agenda

Kaz: I'm OK with writing the details of each TF's activity using, but the discussion on the structure of the whole pages should be done during the marketing call

Lagally: +1 that the structure should be agreed during marketing call

Ege will prepare a proposal for the next marketing call

next week marketing call is canceled

so, will discuss this one week after

Please other TF should also think about the content

Lagally: Yes, should be not that complicated to provide some sentences about the scope of the TF

<Ege> https://github.com/w3c/wot-marketing/pull/99/files

Status TD 1.1

Kaz: we should also provide the last minute changes in the index.html in github master

<kaz> Kaz's editorial updates for publication (Member-only)

TD 2.0 roadmap

There were some discussion about TD 2.0 roadmap

McCool: we should provide a CR 1.1 version in March to be on time

<kaz> Milestones calculator, fyi

McCool: it's quite tight in schedule
... different option such as work in parallel on TD 2.0 and TD 1.1

I think this is not a good idea

Kaz: if we can not finish or start the work on TD 2.0 we need to get rechartered for ver 2.0 and some more possible additional deliverables

<kaz> scribenick: kaz

Kaz: wondering about the actual "roadmap" for 2.0 version

McCool: let's get a poll about deferring the 2.0 version to the next Charter
... need to discuss during the main call as well, though

<McCool> summary it seems we are leaning towards deferring 2.0 to the next charter and focusing on 1.1 in this charter

Issue 999

Issue 999

Sebastian: provided an example
... (put the example TD to edi{TD}or)

Sebastian's comment including the example

McCool: need a list of placeholders?

(like "URL_PLACEHOLDER, ID_PLACEHOLDER", ...)

Lagally's comment

Lagally: also versioning should be mandatory

McCool: yes, it should be mandatory
... we should put version identifier in the id and "version" metadata

Ege: how to handle major/minor/patch versions?

McCool: several different views possibly
... TD is instance of model
... if we have a new version of TD, what would happen?

Ege: depending on the use cases
... some people might think semantic versioning wouldn't work for API versioning

<Ege> https://w3c.github.io/wot-thing-description/#example-31

Ege: version for instance and version for modl

Sebastian: (loos into Issue 1000)
... example of Eclipse Vorto model

<Ege> https://github.com/w3c/wot-thing-description/issues/947

Issue 1000

Sebastian: would it make sense to add "model" term to "version"?

McCool: another possibility is that Thing Model's version number simply refers to the version of the Thing Model specification

Lagally: would be confusing

McCool: that's true

Sebastian: what should we do then?

McCool: for Thing Model, should only have "model"
... without link

Sebastian: ok

"version" : {"model" : "1.0.0"},

Sebastian: let's quickly do that
... (reusing the "version" container, skip `instance` nd introduce a new term `model`)
... (adds an updated example to Issue 1000)

McCool: we could think about DID here

Sebastian's latest comment

version definition for the model:

"version" : {"model" : "1.0.0" },

version definition for the instance:

"version" : {"instance": "1.0.0",

"model" : "1.0.0" },

"links" : [{

"rel" : "instanceOf",

"href" : "<address of the TM model>"

}],

Issue 1003

Issue 1003

table within section "5.4 Default Value Definitions"

Kaz: note there is "5.4" twice within the section title

Sebastian: right
... (creates an issue on that)
... regarding the issue 1003 itself, would it make sense to have more detailed description on op?

(no objections)

Sebastian: Ege, can you create a PR for this?

Ege: ok

Sebastian's updated comment

Issue 983

Issue 983

Sebastian: there is actually a section for Multilanguage (https://www.w3.org/TR/wot-thing-description11/#multilanguage)
... (adds a comment)

Sebastian's comment

Issue 977

Issue 977

Sebastian: the IP address where the TD comes from

McCool: model is meant to be more abstract
... directory entry for discovery needed

Ege: IP address possibly changes based on the environment for mobile usage, etc.

McCool: maybe you could describe the device itself
... would suggest we write up a use case for that

Kaz: agree we should generate a use case description about this
... related to how to deal with session lifecycle

McCool: two possible issues
... no IP address
... and change of IP address

Kaz: no IP address specification for Thing Model may imply dynamic address assignment
... there are several possible settings

updated comment

Issue 1005

Issue 1005

Sebastian: some of the combinations should be prohibited
... e.g., both readOnly and writeOnly as "true"

McCool: maybe JSONSchema guys know how to avoid it

Koster: should be compatible with oneDM

McCool: readable/writable would be clearer

Sebastian: (adds Henry Andrews for JSON Schema and Mchael Koster for SDF)

updated comment

[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: 2020/12/16 08:15:23 $