The W3C JSON-LD Community Group

Go Back

W3C Logo

JSON-LD CG

Minutes for 2024-04-17

Present
Pierre-Antoine Champin, Gregg Kellogg, Benjamin Young, Niklas Lindström, Ted Thibodeau Jr., David I. Lehn
Chair(s)
Gregg Kellogg
Scribe(s)
Gregg Kellogg
Agenda
https://www.w3.org/events/meetings/398465c7-3787-40df-856a-434a438d9f05/20240417T120000/
Topics
  1. Announcements and Introductions
  2. YAML-LD
  3. CBOR-LD
  4. JSON-LD Issue Discussion
  5. Open Discussion
  6. website
  7. next meeting
Gregg Kellogg is scribing.

Topic: Announcements and Introductions

Kamal_singh: I've joined to discuss CBOR-LD; we're implementing CBOR-LD in C.
... Mostly work by grad students; good enough for prototyping now.

Topic: YAML-LD

Topic: CBOR-LD

Kamal_singh: I was trying to use CBOR-LD, but I was wondering if there were plans to compress the output further?
... The problem we had is that the size available is so small.
Benjamin Young: The spec is available in the json-ld org.
Kamal_singh: I've wondered why not to use bits for encoding, vs bytes. I think the spec is a bit old compared with the implementation.
Benjamin Young: Please file issues on the repo. The spec is old, and DB has a plan to update it.
... Is there more that needs to be updated besides the specs?
Kamal_singh: I'll create some issues based on my observations.
Gregg Kellogg: Issues would be useful.
Kamal_singh: what's the timeframe for finalizing the spec?

Topic: JSON-LD Issue Discussion

Gregg Kellogg: It would be nice if we could update the GitHub project to be more automated.
Pierre-Antoine Champin: I'll see about adding the actions we use for RDF-star to work on JSON-LD repos.

Subtopic: w3c/json-ld-syntax#425

https://github.com/w3c/json-ld-syntax/issues/425 -> Issue 425 how to "retype" rdf:JSON to geo:geoJSONLiteral? (by VladimirAlexiev)
Gregg Kellogg: Perhaps support using regular strings with datatypes.
Pierre-Antoine Champin: Looking at the, issues I had one on w3c/json-ld-syntax#420
https://github.com/w3c/json-ld-syntax/issues/420 -> Issue 420 feature request: support `$schema` as an alias for `@context` (by pchampin)
Benjamin Young: +1
Pierre-Antoine Champin: I think we can close this issue marking it as "don't fix".
Niklas Lindström: Agreed
Gregg Kellogg: We've tried to unify with JSON schema for a while.
Niklas Lindström: Other open issues are defined adequately to create PRs.
Benjamin Young: It would be great to have JSON-LD and JSON Schema as a WG topic.
Gregg Kellogg: Might be difficult to work with normatively if it's not a standard.
Pierre-Antoine Champin: WoT WG found a way to reference it normatively, but they have a more structured way of publishing snapshots, which makes it appropriate to cite normatively.
Gregg Kellogg: It's going to take someone to champion the effort.
Pierre-Antoine Champin: I'll do some research.

Topic: Open Discussion

Pierre-Antoine Champin: I found some time to start working on changes to the charter PR. The schedule is out of date.
... It would have started in 2022 :)
... The charter currently says Q12025 for both specs, but I'll need to update that.
... I need to create an issue on the strategy group to formally start the process of re-chartering.
Gregg Kellogg: Dependencies may be RDF-star + 1 year.
Pierre-Antoine Champin: For the moment, we're not mentioning JSON-LD 1.2. We just mention maintanence.
... I think it's best to keep in maintenance for JSON-LD specs until the situation is clarified.
... It's going to be dodgy anyway, and there is an RDF-star recharter being worked on.
... JSON-LD charter should focus on YAML-LD and CBOR-LD until the situation is clarified.
Niklas Lindström: I would wish that wasn't the case, but can see why it is.
... One thing is the "living standard" situation, which might allow a way to do such work.
... We're going to have to jerry-rig things. I think we can live without having a standard rec, but that's a risk.
Benjamin Young: Wondering about timing overlap.
... I'm in favor of pchampin's view. I'd hate to get to the end of our chartering process.
Pierre-Antoine Champin: The initial RDF-star charter ends in August, but we won't be done by then. We're planning on asking for another year for the RDF/SPARQL specs.
... I'm due to produce a new charter for RDF-star. It's more than an extension, as it's more than 6 months.
... A possible way forward is to charter the group to produce JSON-LD 1.2 and open it to new features.
... We could then integrate the RDF 1.2 features when they are published.
Benjamin Young: +1 To staying flexible--however we do that ;)
... We could start working on the new rec-track documents.
... Does that make sense? Or could we do it as maintenance of JSON-LD 1.1? Not sure that buys us anything.
... The living standard seemed good, but I don't think we need to do anything beyond that until RDF-star is clarified.
... Maybe trying to move faster than RDF isn't a great idea.
Gregg Kellogg: Maybe point out in our charter about tracking RDF-star changes.
https://github.com/json-ld/json-ld-star/issues/49 -> Issue 49 Adapting to Triple Terms and Reifiers (by gkellogg)
Gregg Kellogg: Think if that issue as a place to brainstorm.
Niklas Lindström: The triple keyword came up before as well, I'll find a link.
David I. Lehn: Are these keywords useful enough?
David I. Lehn: Given the problems of convincing people to use regular RDF, it might put people off.
Niklas Lindström: "Concretize" could be a possibility.
... Annotation is the primary thing our data needs. Depending on what happens, we might need to put suggestions in the data.
... The many-to-many use cases are interesting, although may not continue.
... Because we overload at-type, it might be overloaded further.
Gregg Kellogg: Muses about YARS-PG.

Topic: website

David I. Lehn: I think we can link to the GitHub pages site for some of what we're doing.
... Minutes could just redirect to the GitHub repo.
... I'm worried that we're going to hit GH limits.
Benjamin Young: +1 To separating minutes out...somehow
... It's too hard to check out the site and edit it.
Benjamin Young: I've wondered if we can get large file support in GitHub, which is now supported. But, at an additional charge.
David I. Lehn: It didn't seem like quite the same thing.
... They're offering similar capabilities to other systems, for a price.
Gregg Kellogg: Maybe we can just move the old audio files and redirect?
David I. Lehn: Do we want to move the minutes and scripts into the main repo?
... The other problem is the playground proxy thing. We've been using another product which is working well.
... Once that's written, we could publish the site as is to point to CloudFlare.

Topic: next meeting

Gregg Kellogg: Next meeting on May 1