W3C

– DRAFT –
WoT Plugfest/Testing

09 November 2022

Attendees

Present
Ege_Korkan, Kaz_Ashimura, Michael_Lagally, Tomoaki_Mizushima
Regrets
McCool, Sebastian
Chair
Lagally
Scribe
kaz

Meeting minutes

Agenda

Kaz: We need to clarify the targets for this call and also the Testfest in December
… I mean which specs to be discussed
… I think we should prioritize Architecture at least for today

Lagally: OK. Let's start with Architecture

Minutes

Nov-2

approved

WoT Architecture Implementation Report

WoT Architecture 1.1 spec

Implementation Report for WoT Architecture 1.1

Lagally: some duplication here
… e.g., Thing Description Playground

Ege: will look into the problem
… seems I forgot to fix the title for one of the two entries

Lagally: ok

Test results table

Lagally: (skims the table)

Kaz: before diving into the detailed check, we should revisit the issue I pointed out
… the question is that the order of the assertions within this result table are sorted by the ID names
… but they should be sorted by the order of their appearance
… i.e., the order of the text within the specification should be preserved for manual assertions

Lagally: agree

Kaz: let's create an Issue for wot-testing
… and ask McCool to modify the table for manual assertions

Lagally: ok
… (creates an Issue for wot-testing)

wot-testing Issue 477 - Order of assertions in Implementation Reports should follow order in specifications

Lagally: let's check the assertions then
… assertions on security to be made non-normative if we don't get sufficient implementations

Kaz: note that we need to update the SoTD section with the list of the features at-risk

Lagally: ok
… (visits the static HTML for CR)

<mlagally_> static HTML for CR

Kaz: it would make sense to compare the list within SoTD and the Implementation Report to make sure

Lagally: ok

Status section

Implementation Report

(16 assertions vs 16 assertions)

Lagally: confirmed the match

Ege: note that we need to update the Testimonial section for the Architecture Implementation Report

Lagally: would like to ask McCool to handle that

Kaz: we've already started some discussion on how to deal with testimonials for 1.1 version specs

Lagally: yes, let's continue the discussion during the main call

December Testfest

December Testfest

WoT Profile

Kaz: technically, we should clarify the requirements and the descriptions within the spec first
… and then think about how to test them
… when we think about how to test them, we need to consider which assertions to be tested how, automatically or manually
… the point is that not all the assertions need to be tested automatically using some smart tool

Lagally: agree

Ege: what about the UI for accessibility?
… e.g., speech synthesis
… note that from my viewpoint, human users are not Consumers
… that might be a philosophical question, though

Lagally: the question is how to process the data

Kaz: yeah
… editor tools and browser tools which include human controls are also WoT implementations
… the more important here is improving the text within the WoT Profile spec before extracting assertions
… so we should add that kind of preparation tasks as well for WoT Profile

Lagally: (then move ahead to the discussion on the test cases)
… standalone test of the Consumer assertions first?

Kaz: we should clarify the whole structure of the testing specifically for the WoT Profile spec
… some of the assertions can be tested based on the existing tests for the other specs like TD
… what still need to be tested for WoT Profile?

Ege: we need actual device implementations
… maybe Toumura-san, etc., can provide some

Lagally: not necessarily fully compliant to WoT Profile?

Kaz: sorry but getting kind of confused
… what we should discuss as the starting point for Testfests is rather how to get the result CSV

Lagally: right
… (adds note on people...)

Mizushima: we should clarify who would participate
… and who would tackle which assertions and provide results for them
… it's difficult for developers to fill in the results during the Testfest itself
… we should assign assertions to the participants in the Testfest
… I'm concerned about the output from the Testfest

Lagally: yeah, I share the same concern...

Kaz: we should clarify which assertions from the WoT Profile to be tested how
… WoT Profile is basically consolidated description of the whole WoT mechanism
… so we should be able to refer to the test results of the other WoT specs
… though newly added clarifications by the WoT Profile itself need to be tested by the WoT Profile spec itself

Lagally: ok

Lagally updates the "Goals and Priorities" section of the Testfest page

[adjourned]

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