W3C

– DRAFT –
WoT Architecture/Profile

16 November 2022

Attendees

Present
Ege_Korkan, Kaz_Ashimura, Michael_Lagally, Ryuichi_Matsukura, Sebastian_Kaebisch, Tomoaki_Mizushima
Regrets
McCool
Chair
Lagally
Scribe
sebastian

Meeting minutes

Agenda

Lagally: <shows the agenda>

https://www.w3.org/WoT/IG/wiki/WG_WoT_Architecture_WebConv#Architecture_.28Profile.29_Nov_16th.2C_2022

<ML adds the CR exit condition>

Minutes

Minutes 09 Nov 2022

Nov-9

Kaz: It seems we review the minutes twice

Lagally: any objections?

no

Minutes 10 Nov 2022

Nov-10

Lagally: some people can join only on Wed or Thu, so quickly skim both the minutes from Wed and Thu. objections?

no

Architecture

PR 878

<kaz> PR 878 - Provide CR exit statment in sotd section

Lagally: Sebastia provided an PR about the CR exit statement

Lagally: How long is the CR review phase?

<kaz> W3C Process Document says "CR period at least 28 days" but it can be longer

Sebastian: not 100% sure, I think min 1 month upto 2month

Kaz: the review phase is at least 28 days

Kaz: usually transition request should be done by the spec editor. Co-chairs and myself can help

Lagally: let's discuss this in the main call

<ML do live editing in the PR to swap the "at risk" sentance to the end>

Kaz: we should use this call to talk about the Arch topics

<ML shows the implementation report of Architecture>

Kaz: suggest to discuss the testing result in the testing call

<kaz> [ We've confirmed that Michael Lagally will work on the CR Transition Request for WoT Architecture and the Status Section (SoTD) for the exit criteria (the Chairs and the Team Contact will help) ]

Implementation status

Lagally: I always wondering why we have different numbers of test results.

<kaz> draft Implementation Results

Lagally: many security-based assertions missing

Ege: Does maybe Oracle support those features?

Lagally: Currently have poc for the Profile Spec.

Kaz: I suspect report submitters don't really understand what the assertions really mean given "The WoT Runtime SHOULD NOT directly expose native device interfaces to the script developers." is not implemented. We need to ask all the submitters about their intentions again.

Ege: maybe an explanation can be that there are different csv which have to be provided. Developer maybe are not aware of it.

Lagally: there are 10 implementations in the test folder, but the report has more

Ege: the report shows all WoT implementations that are involved in the testing

Lagally: we should invite people to provide contributions

Ege: agree
… it would be great to see your implementation also here

Kaz: we should check the current situation in the testing repo and should clearify what is missing first. after that, if some of the assertions which should have been covered already is not covered yet, we should explain what we really meant by each assertion to all the implementers and ask them to submit their results again..

Lagally: right. it's hard to understand what should be done for testing

Ege: actually, there are READMEs and they worked well in the past.

<ML goes back to the PR>

Kaz: The procedure is already defined by the process document. We should mention that we would like to have two or more implementaion.
… if we want to use the arch 1.0 version result we should check the status there.

Lagally: I have a problem to reuse from the past also from members that are not members anymore

Kaz: we should be clear which features are new in Arch 1.1 and should be explained in the IR

Lagally: we do not test that we are backwards compatible

Kaz: If we really want, we can define a policy to require WoT Architecture 1.1 implementations which cover all the features including both 1.0 compatible features and 1.1-specific features. However, we can expect 1.0-compatible features are implementable based on the results of the WoT Architecture 1.0 Implementation Report.

Ege: I just agree what Kaz said

Sebastian: in my understanding, in the CR phase the IR needs not to be finalized

Kaz: Anyway, given some of the simple/clear assertions like "abstraction of device interfaces" are not covered yet, we need to look into the test results themselves

Kaz: IR results are needed for the PR transition, so we can continue the detailed check in parallel.

Lagally: is the current version enough for the CR transition?

Kaz: I'm ok. the question is to check the results

Lagally: I'm going to merge this PR

Sebastian: we will use the main call to talk about Profile PRs

adjourn

Minutes manually created (not a transcript), formatted by scribe.perl version 196 (Thu Oct 27 17:06:44 2022 UTC).