The W3C JSON-LD Community Group

Go Back

W3C Logo

JSON-LD WG

Minutes for 2024-09-04

Present
Anatoly Scherbakov, Ted Thibodeau Jr., Benjamin Young, Pierre-Antoine Champin, Gregg Kellogg, Niklas Lindström, dlehn2
Chair(s)
Benjamin Young
Scribe(s)
Benjamin Young, Gregg Kellogg
Agenda
https://www.w3.org/events/meetings/08563546-30fd-497a-ac16-77f96e96e8c0/20240904T120000/
Topics
  1. Announcements and Introductions
  2. TPAC
  3. Other topics
Benjamin Young is scribing.

Topic: Announcements and Introductions

Topic: TPAC

Gregg Kellogg: TPAC discussion
... we're meeting with Spatial, RDF-star, and Web of Things
... where are we at on our JSON-LD support for RDF-star
Niklas Lindström: Yes. happy to look more into that
Gregg Kellogg: We should also be prepared to discuss things around `@json` and the `RDF:json` data type
... I think our dedicated time is Thursday
Group scowls at calendar
Gregg Kellogg: Our times are missing from the TPAC week
... and we're also not listed on the RDF-star joint meeting
Benjamin Young: Pchampin could you find out where those went?
Pierre-Antoine Champin: Yes. I'll take a look
Gregg Kellogg: So, what thinks should we talk about at TPAC?
... let's do a quick roll call
Pierre-Antoine Champin: I'll be there
Benjamin Young: I'll be there
Gregg Kellogg: So...three of us...and Zoom
... we do have PRs still pending and several things that still need classification
Gregg Kellogg: One big topic is our charter
... and relatedly our breakout session
... what are the next steps charter
Benjamin Young: Not sure what next steps for the charter are; a couple of PRs have been merged. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... We need to decide when it's ready and then send in for feedback.
Gregg Kellogg: ... The question is, have we covered the topics of interest in the proposed charter?
Gregg Kellogg: ... YAML-LD, CBOR-LD, and multiple-language values.
Gregg Kellogg: ... We haven't said we're going to do a new version of JSON-LD.
Gregg Kellogg: Timing for a 2.0 is probably dependent on RDF-star
... but we should probably include updating the specs based on erratum
... and the implications of RDF-star
Gregg Kellogg: I think the RDF-star related things need to be in the charter. [scribe assist by Gregg Kellogg]
Benjamin Young: There are classes of changes we can't change. [scribe assist by Gregg Kellogg]
Gregg Kellogg is scribing.
Pierre-Antoine Champin: The JSON-LD specs don't allow new feature in maintanence mode. Class 3 changes at most.
... This is fixing errors in normative text but not adding new features.
... Triple terms would be new features.
... In maintenance mode, we can't make the discussed changes.
... We could update the charter in that spirit.
... Of course, there are uncertainties in the RDF-star work, itself.
https://github.com/json-ld/json-ld-wg-charter/pull/5 -> Pull Request 5 update end-date and timeline (by pchampin)
... We do have a pending PR with updated deadlines.
... The timeline in the current proposal isn't realistic.
... I think we could send it to horizontal review, but if we want to make changes to allow us to update the spec.
Gregg Kellogg: It's a question of scale of the changes
... and whether it invalidates any previous JSON-LD
... I don't think the community would support that
... so we do need to focus on backwards compatiblity
Gregg Kellogg: I think we might be able to do in a 1.2 vs a 2.0 version of the specs.
Anatoly Scherbakov: What do you think of making the charter open ended
Anatoly Scherbakov: What do you think of making the charter open-ended.
Benjamin Young: I wanted to discuss what we can change about JSON-LD.
Anatoly Scherbakov: I thought the charter could allow inclusion of more *-LD languages, but I agree with the point about the defined scope.
... Within the rechartering, we are changing from a Maintance Group to a Working Group that can allow us to make changes.
... Do we need something in the charter to talk about changing the core specs.
... We've talked about doing normative changes.
... We've at least addressed errata and RDF-star changes, but that is outside the scope of a Maintenance group.
... A version 2.0 is probably beyond what we can do.
Pierre-Antoine Champin: I realize some piece of text has been accidentally removed in the description of our liaison with RDF-star.
Pierre-Antoine Champin: The process says nothing about "Maintanence Group", we could say that it is a WG without new deliverables. It's chartered to maintain and update documents according to what the process allows.
...That's why this group is a hybrid, as it's currently considered, with maintenance for the existing specs and work on new specs.
Benjamin Young: It might be too early to know how many new features we would have.
... I suggest we look to use TPAC to determine what we want.
Pierre-Antoine Champin: +1 To use TPAC to gauge the needs of other communities
Benjamin Young: K. then let's wait to change the charter in that way until after TPAC
Gregg Kellogg: So I think it does come down to what things may come from RDF 1.2, etc.
... but we do also have open "future-work" issues we could consider
... so maybe we triage that list and come up with a guess at what we may do
... we don't have to know that list nor put it in the charter
... but that could help us know what's hoped for
Gregg Kellogg: We've been talking about new changes for years
... so to not look at these open issues would be strange
Benjamin Young: For the last four months, the charter has said we're just maintaining the specs.
... We probably need to change the charter to allow for revising the specs.
Pierre-Antoine Champin: I think the new charter has been around for a while; there was a strategy for YAML-LD and CBOR-LD, but to keep things the same otherwise.
... The goal was to be quick, which we haven't been.
... The urgency for changes has receded, so we can probably work on larger changes to the charter.
Gregg Kellogg: I'll work on a PR to the chater.
Benjamin Young: Let's get the current PR in first.
Pierre-Antoine Champin: I think we can merge that now. Timeline will change if we add more work.
Agenda?

Subtopic: TPAC breakout sessions.

Benjamin Young: The general topic at TPAC is rechartering, and presenting YAML-LD and CBOR-LD.
... pchampin, bigbluehat, and gkellogg can co-present on other discussions.
Anatoly Scherbakov: Is the time of the breakout session already defined or I have missed it?
Anatoly Scherbakov: I assume I need to prepare a presentation about YAML-LD. Is that right?
Gregg Kellogg: I don't think we've seen a breakout schedule yet.
Anatoly Scherbakov: How much time do I have for that presentation?
Benjamin Young: No, and I don't know that it's up to us
Pierre-Antoine Champin: Their starting to work on the schedule now.
Anatoly Scherbakov: Acknowledged. Is it right that anyone can join uor session so this should be for general audience? Not just a recap for us
https://github.com/w3c/tpac2024-breakouts/issues/35 -> Issue 35 CBOR-LD, YAML-LD & the JSON-LD Recharter (by mandyvenables) [session]
Gregg Kellogg: If the session is an hour, we probably budget 15 minutes each for YAML-LD and CBOR-LD.
Anatoly Scherbakov: Acknowledged
Benjamin Young: The goal of the breakout is to stir interest in the work.
Anatoly Scherbakov: So, this will be unbounded boasting of how cool our group is
... (Mostly marketing).
Anatoly Scherbakov: You want to be a cool kid on your block? Come join us and use YAML-LD to list your groceries!
... I'll get something on the calendar to go over the slides early next week.
Anatoly Scherbakov: Great, will try to get slides draft done before that time
Benjamin Young: We'll make it a CG call just to discuss meeting materials and presentations.

Topic: Other topics

https://github.com/w3c/json-ld-api/issues/558 -> Issue 558 Compaction cannot round-trip terms using `@container: @list` and `@type: @vocab` (by niklasl) [spec:enhancement] [spec:substantive] [ErratumRaised]
Niklas Lindström: I was planning on making a PR for this; I have a fix in my impleentation.
Anatoly Scherbakov: Thanks for your recent comment there gkellogg, I need to address it. I haven't done that yet. I think if I manage to do it this PR will be ready for re-review & maybe merge
David I. Lehn: More of a general question on versioning. Is TPAC a good place to ask what people want of JSON-LD?
Anatoly Scherbakov: …Or of YAML-LD
Gregg Kellogg: We can put an agenda item on our dedicated meeting time.
David I. Lehn: I've noticed that people have problems writing contexts right; there may be the possibility of compiling things into the context.
... I don't see much deep use of contexts.
Gregg Kellogg: Perhaps its a tooling issue.
Anatoly Scherbakov: I've been thinking of compiling a context from an ontology. But didn't do about that
... Maybe to visualize contexts and make changes by description.
Anatoly Scherbakov: Do we have any contacts with FAIR practitioners? They are expected to write JSON-LD. Maybe to offer them YAML-LD?
Benjamin Young: Did you want another breakout, or to include it in the session.
... I'd love to spend a bunch of time on this kind of thing.
Anatoly Scherbakov: I did get ChatGPT write JSON-LD & RDF for me
Gregg Kellogg: Getting folks to come to the breakout session and the group meeting would be a good start
Benjamin Young: If there are examples out there, the AI tools can do pretty well.
... There's also a proposal is to say "make this JSON JSON-LD using known vocabularies".
Anatoly Scherbakov: You can also explain it what to do and it will do that. I gave it YAML-LD spec and it did write YAML-LD documents with dollar-context
David I. Lehn: Do robots understand JSON-LD yet? How it works?
Benjamin Young: I think for the most part, it's just going to add a context and not do things more nuanced.
David I. Lehn: I asked it to describe something once, and it got it totally wrong.
Benjamin Young: Writing vocabularies is hard, and people don't understand the difference between a context and a vocabulary.
Anatoly Scherbakov: ChatGPT makes mistakes all the time. Iterative approach can help a lot.
... Is anyone maintaining LOV (Linked Open Vocabularies).
Pierre-Antoine Champin: I knew something working on the project and has done some ongoing maintanence.
... It's hosted at university of Madrid, although it seems to be down right now.
Anatoly Scherbakov: https://lov.linkeddata.es/dataset/lov/about → Service Unavailable
Anatoly Scherbakov: Prefix.cc is cool, I often use it
Benjamin Young: Maybe we should just invite people to demo their JSON-LD.
Anatoly Scherbakov: I'm making visualization for ontologies and I'd like to make a demo of my stuff at some point
Niklas Lindström: +1 I'll think of demoing
Anatoly Scherbakov: Please throw such an invite at me! thanks!
... I just sent out an invite for a CG call to discuss TPAC and we'll probably have another WG meeting (or not) before TPAC
Anatoly Scherbakov: Thanks all, sorry for my audio