Meeting minutes
Minutes Review
<kaz> Feb-19
McCool: Been a while, probably good to scan through
… meeting schedule should get organized
… PRs were merged
Kaz: The link is actually wrong in the wiki
McCool: Right, need to change that. Thank you for catching that
… (changes the link in the wiki)
… (goes over the correct minutes now)
… had a long discussion about versions
… one PR is still open, one got merged
… if we still have time, we should look at the calendar
… any objections to approving the minutes?
None, minutes are approved
Calendar
McCool: So, we have some upcoming holidays
… I wonder what we should do about holidays around easter
… Easter Monday is a holiday in Canada, in Germany as well I suppose
Jan: Exactly
McCool: So we should cancel April 1
McCool: Due to golden week, we should also cancel April 29
… (updates the Wiki)
Mizushima: On April 8 and 9, there are AC meetings
McCool: Is a direct conflict?
Kaz: Will be held in Hiroshima and there is going to be a after-party
McCool: So April 8 is going to be cancelled as well
… May 6 is also going to be a Monday, so it also going to be cancelled
… that's half our meetings cancelled for upcoming weeks, but can't do anything about it. Same is probably true for security
Resources
PR 537 and 536
<kaz> PR 537 - Expand rdfs:comments in Discovery ontology
<kaz> PR 536 - Render changelog in discovery ontology from RDF
McCool: So last time there were some problems which you tried to resolve
… and I guess we decided to change the version number to say a1
… but I think we agreed to use 1.0 to not get things out of order
… there was also the idea of creating a draft directory
… but then again, the discovery repo can be seen as a draft repo
… any objections to yet?
Jan: No, sounds good to me
McCool: I see you also updated the meta data
… but the first version is now saying 0.1.2
… in the changelog
Jan: That comes from a previous version
McCool: As we have a published version with 0.1.2 we could also consider using 0.1.3 for the new version
Jan: Or maybe use 0.2.0, as we've added some new "features" with the updated descriptions
McCool: Yeah, we made some larger changes
<McCool> suggest 0.2.0-a1
Jan: Updated and pushed now
McCool: Looks good, I will merge it now
… any objections?
None, merged
McCool: Please synchronize with the TD guys to get a consistent output
Discovery PRs
PR 538
<kaz> PR 538 - Introduce DataSchema for property 'things'
Jan: This reflects a discussion we had in node-wot regarding the discovery API
… using TDDs
… so Cristiano opened this PR
McCool: This mostly updates the JSON Schema and the TM of the TDD
… so I am fine with merging this
… did you have any chance to test this yet?
… we need to discuss testing in general, creating issues, regarding pagination and no pagination
… for example
Jan: I think there have not been any tests yet, but I don't know if Cristiano maybe updated his implementation already (probably not)
McCool: I'll generate an issue for the testing of the things property, if we already have a test then we can mark it as completed immediately
Kaz: Merging this PR itself is fine if everybody is ok with this change, but I am wondering about the relationship between the specification test and the schema
… is the specification text clear enough?
McCool: The TM of the TDD is in the spec
… not sure if it is normative or not at the moment
… it is also a TD and not a JSON Schema, right?
… it's in the API specification, so it is actually normative then
… but it is not the only source
… so we need to make source that it is consistent with the specification
… and we have a listing endpoint, which should actually be "things"
… so Cristiano is actually adding a schema definition that is describing the return values mentioned in the specification
Kaz: What to be tested is the text within the specification, so I asked if the spec text is clear enough. On the other hand, Schema can be used to make the test easier. So there are two levels of questions here.
McCool: Generally, we should organize testing better
<McCool> (McCool creates a new issue: https://
McCool: for the moment, we can merge this PR
… any objections?
None, merged
PR 542
<kaz> PR 542 - Replace custom JSONPath reference with RFC 9353
McCool: So this PR just replaces the two informative references of the JSONPath I-D with the RFC
… we need to check the RFC and investigate its features
… I will create an issue for that
… any objections to merging?
None, merged
Jan: There are already a couple of issues related to JSONPath
McCool: Yeah, so we can simply add a comment to an existing one
… (adds a comment to issue 387)
<kaz> Issue 387 - JSONPath queries filtering on semantic annotations|
<McCool> w3c/
McCool: The other issue that has been brought up is that JSONPath does not deal with JSON-LD very well
… Cristiano also mentions this in a comment in the issue
… this is especially related to expanding
… but we will deal with that next time. For now we are done with PRs
DID Registration
<kaz> did-spec-registries PR 486 - Add WotThing and WotDirectory service types
McCool: We now have a referencable URI
… I updated the PR against the did-spec-registries repository
… has been reviewed, should now be ready to be merged, but is still pending
… we need to get back to it next week, by then it will hopefully be merged
… so far, I have added some more explanatory text
… explaining the two terms and references the Discovery specification
… can you also check again if the PR references the correct terms/URIs from the discovery ontology?
Jan: Sure, will do
Planning
Main Calendar
McCool: Let's look at the schedule
… we have some dates here which are the worst case scenarios
… there is a bit of padding here
… we should probably aim for a earlier due date
… FPWD is in about a year
… we are mostly doing incremental changes
… we probably do not have enough time for a complete overhaul
… so probably it will rather be Discovery 1.1 instead of 2.0
<kaz> schedule.md
McCool: and hopefully we will not have breaking changes
… a good target will probably be end of this year, so we should have a first version of a finished document mid December
… (starts updating the schedule)
Kaz: Any date is fine if it's feasible
… as an initial target
… and we need to think how to get there
… and especially for the TD 2.0, we need to think how to get feedback from developers and the industry
… that is also true for discovery, right?
McCool: Ideally we should have our requirements by then
Kaz: my point is that the proposed date (end of 2024) is fine for FPWD as the initial target, and then we need to clarify the sub steps to get there as the next step :)
McCool: Let's look at the charter draft
… (opens the rendered document)
… looking at the timeline, we should have the use cases and requirements by Q3
… so September
… let me put use case and requirements and testing somewhere in our schedule
… (updates schedule.md)
… so I guess this would be a Note update (regarding use cases and requirments)
… so by September we probably want to have an updated Use Cases and Requirements document
… and then after the FPWD, we want to do testing in June 2025
… all of these deadlines are worst-case scenarios and include some padding
… I will not merge this but create a PR which we can discuss in the main call
… (commits the changes and opens a PR)
<McCool> w3c/
McCool: we are out of time for today but we made some progress
… for the next time, I think it would be nice to align our work items with the requirements
… I think we should pick a work item to work on in the meantime
… next time, we should look at the issue tracker and align with the things happening in the use cases task force
… we should probably go through the many issues, clean some up and change the "defer" label to "current"
Task Force Lead
McCool: Would you consider becoming TF lead, Jan?
Jan: I would consider it, but I would need some support
McCool: We need more people joining the calls
… let's discuss this topic next week
[adjourned]