W3C

– DRAFT –
WoT-WG - TD-TF - Slot 1

22 May 2024

Attendees

Present
Cristiano_Aguzzi, Daniel_Peintner, Ege_Korkan, Jan_Romann, Kaz_Ashimura, Kunihiko_Toumura, Luca_Barbato, Mahda_Noura
Regrets
Koster
Chair
Ege
Scribe
JKRhb

Meeting minutes

Agenda Review

Ege: Today's agenda is mostly about errata fixes
… a report from the sub-taskforce regarding LinkML
… will report on Binding Note publication and discuss project management

Minutes Review

<kaz> May-15

Ege: Already went through them, looked good to me, does anyone have remarks?
… everyone should have received the minutes
… not hearing anything, minutes are approved

Regrets

Ege: Next week on Thursday, I cannot join, but my Wednesdays have been converted back to normal workdays

Errata

<kaz> PR 2023 - Fixes to Errata Documents

Ege: There are some errata fixes which we can merge today
… Discovery and Architecture are probably going to be merged in the main call, as the calls are not happening at the moment
… (shows the diff of PR 2023)
… the main thing I did was to remove javascript
… removing the script did not do anything bad
… also removed some attributes from HTML tags
… could be incorporated into the workflow
… this was for Errata 1.1, I applied the fixes to the 1.0 version
… there was also some obsolete content that I removed that was not applicable to us
… (shows the rendered version)
… the script I removed does not do anything unexpected, the rendered version looks the same as before
… (shows the previous version and the new one side-by-side to illustrate that there no changes to result)
… just noticed that there is one thing that has changed (a missing link) and the table of contents has disappeared
… not sure if it is that important, though
… that is something that could also be added manually

Daniel: I don't think the table of contents is needed
… the whole page should be rather short anyway

Kaz: I agree
… given that the content is short enough, we don't need to put the table of contents or the sections numbers
… by the way, you plan to add links to all the errata contents, right?

Ege: Yeah
… just noticed that we can also update the date
… and that we can update the errata count
… and copy-paste the submission text
… (edits the PR description)

Daniel: Maybe it would make sense to add short how-to on how to add a new entry to the errata

Ege: You mean for the editor of the document, right?

Daniel: Yes, if it is just the date it is not that important, but if there is more than instructions would be useful

Ege: I will add that information to the document itself
… (adds that to the PR description)

Ege: I will work on that and then the next time, the documents will be in a better shape
… then the next time, we can actually discuss the policies in the main call

Tooling Sub-TF

Ege: Today we met with five people before the TD call
… we have a dedicated repository we are working in
… where there are draft PRs coming improving the automation
… so far the README and the project structure have been updated by Mahda
… her idea is to have a central file that can be used for generating the tables in the TD spec
… then in an index.template.html file, we can have the additional text that accompanies the tables

<kaz> resources area within the wot-thing-description-toolchain-tmp repo

Ege: there were some JSON Schema issues that we still need to investigate furhter
… we meet 30 minutes before the first TD call under the same Zoom link, if anyone wants to join

Binding Templates

<kaz> Main call resolution for Binding Templates publication

Ege: There was a resolution on publication in the main call
… we merged one PR by Daniel improving the contrast in one image
… with that the remaining issue has been fixed
… we still need to update the changelog, right Kaz?
… after the main call, however, I mentioned that this point is not valid anymore, but I can still send an email

<kaz> Issue 364 - Figure Improvement and minor bug

Kaz: Technically, we should have sent the reminder before the resolution, but since we have already made it, we can simply go ahead

Ege: We can then start preparing the document
… (creates an issue for that task)
… (adds it to the "in progress" column in the TD project)
… creating the static files can be done until next week

TD Project/Backlog

<kaz> WoT TD Project

Ege: Here, we can have a look into what has been happening during week
… there are two topics
… one is looking into the issues that have been worked on
… the other one is looking into one of the big issues, as discussed with Michael Koster
… the big topic we could discuss is "Reusable Connections"
… but first we should look into the project

PR 2017

<kaz> PR 2017 - Add README for json-ld files

Ege: This one can be merged, after the unwanted changes have been resolved

Cristiano: Not sure why these changes were appearing, fixed them manually on my machine

Ege: Will explain shortly why they were happening

PR is merged

PR 2018

<kaz> PR 2018 - Add text regarding choosing between events and observable properties

Ege: There has been an interesting discussion here that would be interesting for everyone
… regarding best practices
… I had a short discussion with Jan here, so he is adding a sort of best practice, which we have in multiple places
… would be nice to have a single place to put this information
… the text itself in the PR is regarding the use of events vs observable properties
… he asked where the best place to put this
… I think we could have some way of collecting this information from across the document and then give this information to developers as a guideline
… seeing no one in the queue at the moment, I will add a separate issue for that as a reminder to revisit this
… (creates issue 2024)
… the PR itself can be merged then

Jan: Will rebase/clean up the diff

Ege: Then we can merge it asynchronously afterward

PR 2021

<kaz> PR 2021 - Generating index.html

Ege: This PR addresses the issues Cristiano was experiencing in his PR
… so the issue was here that the template was not formatted before rendering but then the index.html did not get formatted
… causing the "ghost changes"

Cristiano: That explains the issue, but maybe we can find a cleaner solution

<cris> +1

Ege: I will do the changes to fix this, changing the prettier config and GitHub Action, as also noted by Daniel
… that will provide a clean fix. Afterward, we can merge the PR asynchronously
… (adds a comment to the PR)

PR 365

<kaz> wot-binding-templates PR 365 - Align Readme

Ege: This PR updates the README.md with the template created by the marketing TF
… now contains links and information regarding tooling
… if there are no objections then I will merge this

Issue 1972

<kaz> w3c/wot-thing-description#1972 Issue 1972 - unwanted any type for title

Ege: Mahda and I have been discussing this in the office

Mahda: The thing is that some while ago that there was an issue from someone regarding the SHACL shapes
… the issue that the id in the shapes is defined as a language strings
… which causes the validation to fail if the string is empty
… in the SHACL shape I added a fix that accepts both regular and language strings, but there is a bug in the tooling that causes the validation to fail
… so the SHACL is correct but the tool cannot deal with it
… I also noticed that Cristiano did a lot of work on this a year ago

Cristiano: There was an issue with the differentiation between single values and arrays, if I remember correctly
… maybe it is better to have correct SHACL, even if we have to sacrifice a bit on the tooling. At the moment, it is also a good opportunity to refactor

Mahda: The title is also an "any type" at the moment

Cristiano: Maybe we can a workaround for the moment and then solve it for good with the new tooling
… I will check what I did in my PR I did a while ago
… and then report back to you

Mahda: In one discussion recently, you actually provided a link, will try to find it
… I will ping you, Ege, once I find it, so you can update your comment in the PR

Ege: I will assign both Cristiano and Mahda for this issue

Mahda: I just pasted the link in the chat for you, Ege

Ege: Then let's see how we can deal with this issue, if it is going to be a "dirty fix", then that should be fine

Reusable Connections

Ege: We probably don't have enough time for this issue
… but we can have a brief look
… the idea here is that you can avoid repetition
… so we have the base, but it is a bit annoying that you need to repeat other stuff
… for example the default content-type
… so, in Modbus you always need to repeat the content-type
… or in WebSockets you also need to provide the details of how the connection should be set up

<Ege> https://github.com/w3c/wot/blob/main/planning/ThingDescription/td-next-work-items/usability-and-design.md#reusable-connections

Ege: all of that is mentioned in the document linked above
… as everybody can see, there are issues linked in the document that started this overarching issue
… still, it is probably at the easier side of things here
… any opinions so far?

Cristiano: Definitely good to start working on this, not opposing
… however, I would say there is a subtle difference between reusable connections and avoiding repetitions in a TD
… so connection reuse has a greater impact than templating
… just to keep that in mind when we are discussion this

Ege: I agree, the different bullet points mentioned in the document have different weight/semantics

Cristiano: Yeah, but that is probably also just an implementation detail
… did we already discuss how we should tackle this issue?

Ege: Brain storming should happen here, then we can start with concrete proposals
… maybe we should create a big issue to collect all of the ideas
… and then close the other issues that are currently linked

Cristiano: As you suggested, maybe we can reserve 30 minutes the next time to clear waters here

Daniel: Maybe a minor point, but maybe we can also mention "initial connections" in the title, since I had to read the text itself to realize that this is also included in this topic

Ege: Sounds good. We should then reserve some time to talk about this issue

AOB

None

[adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 221 (Fri Jul 21 14:01:30 2023 UTC).