Meeting minutes
Minutes Review
<kaz> Apr-3
Ege: Let's start the minutes from last week
… the minutes looked good at a first glance
Kaz: Since we had the AC meeting yesterday, we are reviewing the minutes from the last meetings, right?
Ege: Correct
… there is a leftover abbreviated name
… otherwise, they look good to me
… does anyone have issues?
Kaz: Issue has been fixed
Ege: Good, then minutes are approved
<kaz> Apr-4
Ege: (shows the minutes of the second meeting)
Ege: there have been some IRC formatting issues
… two resolutions in total
… we can approve the minutes, however
Minutes are approved
Agenda
Ege: There are some weird things in the agenda
… as we are closer to switching to Github projects for project management, we are marking some points as italic, therefore, the formatting gets messed up
… you can ignore it, though
TD Next Editor's Draft
<Ege> PR 1997 - Preparation for TD Next Work
<kaz> Preview
Ege: This is the current state of the PR
… currently, it looks as if we are changing the old version
… therefore, the PR proposes to change to title
… currently, it just adds a "Next", but we can discuss that
… also adds some additional information regarding breaking changes
… only textual changes, but still important to discuss some points
… also did some changes to the README
… there are some unrelated changes at the moment, only the ones in the index.html are relevant
… would like to hear some comments, whether you like the changes and whether we should use "Next" or "2.0", for example
… but this should send the message that we are working on a new version
Kaz: Basically, you copied all the content from the Binding Template side to the TD document, right?
Ege: No, that is not included in this PR
Kaz: Okay. We should also update the list of editors by the way
Ege: Will update that when we will move the content of the Binding Templates over
… for editors, we should have a separate discussion as well, e.g. regarding authors vs. editors
Cristiano: Nothing more to add, like the direction of the PR
Ege: Any opinions on 2.0 vs Next?
Luca: I guess we can already go for 2.0
… but that could even be done as a second step
… you already did a lot of work here
Ege: Would be a small change, though
Cristiano: Not having a strong opinion, "Next" would give us more flexibility
… but then again, 2.0 is still in the charter
Koster: We can start with our semantic versioning at any time
Kaz: Given that our charter only says "TD update", we can go with "TD Next" and then think about the title as a second step as Luca mentioned. I personally think everyone seems to be okay with 2.0, though.
<kaz> WoT WG Charter
Ege: Regarding the context URL, we should use a temporary one, I will create an issue for that
… will also try to fix the large amount of unrelated changes.
… (adds a comment to the PR)
<kaz> Ege's comment
Ege: will fix these issues and then merge it asynchronously
Toolchain Discussion
Ege: There have been some discussions but nothing new here
… Mahda is looking into using LinkML and will present some initial results in one of the next meetings
… there is a related PR in the Binding Template respository related to the toolchain which we can use as the basis for the discussion
Modbus PR
<kaz> Modbus Binding Template - 4.4 Function
Ege: I have noticed some bugs in the toolchain that should need to be fixed
… so in the Modbus binding, we have a context, ontology and a JSON Schema
… there are some inconsistencies between ontology and the resulting Binding Template
… caused by a bug where the tool was lowercasing everything
… causing an specification to be wrong if it looks at the implementation or the JSON Schema
… another example was a typo in the JSON Schema, where an "s" was missing, causing bugs when read by a machine
… these are examples where our current tooling causes issues
Mahda: This is related to a recent issue
… in the TD repository
… issue number 1988 to be precise
… caused by a different naming in the context and the ontology file
<cris2> +1
Ege: In the future, there should simply be no possibility to make this mistake
… another change in the PR is that I added Cristiano as an editor
… also fixed some capitalization issues
Cristiano: One thing: While you explaning the changes, I noticed something
<kaz> wot-bindig-templates PR 359 - Modbus fixes
Cristiano: you mentioned that an "s" was missing, so maybe the Schema was wrong but the ontology was right
… maybe we need to double check the terms and use plural if required
Ege: So you are saying that we need to check with the Modbus specification
Cristiano: The fix itself is correct, so maybe we can add a follow-up issue and check again later
Ege: (adds a comment to the issue)
… thank you, good point
Kaz: Thank you very much for your hard work, but at least for the discussion today I am a bit confused, maybe we need to make it a bit clearer what the actual issue was
… so there is an issue in the TD repository, but a PR in the Binding Template repository
… so we need to clarify what is wrong on which side and if there is an issue in the TD specification, we might need to consider issuing an errata
Ege: So this is only on the side of the Binding Templates repository
… so we don't need an errata, I think
… otherwise, it is a human error, one letter makes a big difference
… no one has noticed it so far
Kaz: This implies that we might need an even nicer review mechanism as well
Ege: True, but it would also be nice if there was no way to make this mistake
Kaz: We could maybe document this and add it to the review policy at some point as well
Ege: (updates his comment in the PR)
… I will not merge this PR so far, please also have look at the changes, Cristiano
… then we can go to the next topic
Project Management Discussion
<Ege> PR 1990 - Simplified lifecycle diagram
Ege: Did not have time to look into versioning again, therefore I would propose going forward with the project management
… in the PR above we propose a simplification of the process I discussed with the chairs
… the diagram was a bit too complicated at the top
… should be a bit easier to understand now
… also making sure that we have branching based on the decision
… also, the whole process is now issue-driven
… issues might be delegated to the use case TF
… if an issue is accepted, it will be prioritized and assigned, work will be done via PRs
Cristiano: Looks good to me
Kaz: Thank you very much for this part as well!
Daniel: Just noticed that steps 8. and 9. are flipped, but it is not a big issue
Ege: Sure, can adjust this quickly
… (adds a comment to the PR)
… I mean, even in this diagram, steps 1 and 2 are from left to right
… any other points or rejections to the PR?
… then I will fix these points and merge asynchronously
Ege: The other actual thing I wanted to discuss is starting to work this way
… we have the table and did some categorization, e.g., regarding use case relevance
… once we have the UC template, we can do that as well
<kaz> Rendered MD
Ege: but in the other aspects, we can already start working this way
… do we need a main call resolution for that?
… I think working this way will make the things we work on more visible
Kaz: I think giving a quick presentation during the main call would make sense
… also we can ask Michael Koster for his opinion
Koster: I just agree
Ege: (updates his comment)
PR 1998
<cris2> +1
PR 1998 - Adding ignored paths to GH Action to match prettierignore for prettier changes
(no objections and merged)
PR 1999
PR 1999 - Term inconsistency among the ontology and context file
Ege: is this PR ready?
Mahda: PR itself is ready, I think
… but still need to look into Prettier
Kaz: would confirm the changes
Ege: kind of bug fixes
… moving the description on hasConfigurationInstance and then renamed it
Cristiano: right
… note the changes include the resources (td.ttl)
Daniel: might need an Errata, maybe...
… we can fix the bug but there was a bug within the spec
Ege: right, but the spec HTML itself was correct
Kaz: so no issue or problem with the index.html of TD spec
… but related resources like ontology/td.html and ontology/td.ttl files are to be fixed
… so it might make sense to describe that on the Errata
… also we should think about what to be handled by the versioning mechanism and what to be handled by the Errata mechanism as part of our policy
Ege: good point
… for this problem itself, we should have a follow-up PR for the resources
[adjourned]