The W3C JSON-LD Community Group

Go Back

W3C Logo

JSON-LD CG

Minutes for 2024-01-10

Present
Gregg Kellogg, Pierre-Antoine Champin, Ted Thibodeau Jr., Kevin Herndon, Benjamin Young, Niklas Lindström, Anatoly Scherbakov, David I. Lehn
Chair(s)
Gregg Kellogg
Scribe(s)
Gregg Kellogg
Agenda
https://www.w3.org/events/meetings/398465c7-3787-40df-856a-434a438d9f05/20240110T120000/
Topics
  1. Announcements and Introductions
  2. YAML-LD
  3. CBOR-LD
  4. JSON-LD Issue Discussion
  5. Open Discussion
Gregg Kellogg is scribing.

Topic: Announcements and Introductions

Kevin Herndon: I had an email conversation with gkellogg a couple of years ago about how to get more involved.
... May be too technical for me, but I wanted to find out.
... I come from accounting/finance, but as a hobby, I'm a musician. I put together a website for my music, but started incorporating JSON-LD/schema.org into my website.
... I've left my job, and am looking for something new to get into.
Pierre-Antoine Champin: On Mastadon I anounced that I leased version 0.8 of Sophia with full support for JSON-LD.
... Sophia is a Rust library for RDF.
Pierre-Antoine Champin: https://crates.io/crates/sophia
Anatoly Scherbakov: й+
Pierre-Antoine Champin: There should be an IE invitation in progress for adding anatoly-scherbakov
Anatoly Scherbakov: I am already an invited expert.
... I hope to bring the YAML-LD spec to Rec.
Niklas Lindström: To be clear, I'm not a member of the WG, because of the DCMI exit, and the National Library of Sweden hasn't done the paperwork for me. This should happen ASAP.
Pierre-Antoine Champin: I wanted to point out about niklasl and kherndon that it's still possible for chairs to invite non-members on a case-by-case basis.
... Particularly if there is some temporary issue.
... Regarding our charter, it talks about YAML and CBOR as non-normative work.
Ted Thibodeau Jr.: The chair can invite anyone, but substantive contributions aren't covered by this. It's helpful to keep up on discussion, but not to make contributions.
Niklas Lindström: That's good to know.

Topic: YAML-LD

Gregg Kellogg: Generally, avoid contributing new PRs to work in progress.
Anatoly Scherbakov: I found an issue that the spec doesn't mention, about how we handle YAML-LD embedded in HTML.
... the extractAllScripts tag would likely extract not just the scripts but also all the documents from a stream.
... If I for the repo and do some PRs, will it be merged into the current version of YAML-LD.

Topic: CBOR-LD

David I. Lehn: I was supposed to move the spec from the DB repo to the json-ld org. It's hard to keep things working.
Gregg Kellogg: There is some dance to do on this.
David I. Lehn: I'll work on this, but we're working on other related things too.
Gregg Kellogg: We should continue to develop the CBOR-LD spec in the CG.
... I think the CG patent policy allows us to work in the CG.
David I. Lehn: Writing down the issues? We have a CBOR-LD spec and if we're going to try to mix work.
Gregg Kellogg: There are parts from YAML-LD, pchampin's JSON-LD in CBOR and DB's CBOR-LD spec to me mixed together.

Topic: JSON-LD Issue Discussion

https://github.com/w3c/json-ld-api/issues/581 -> Issue 581 If processing mode is json-ld-1.0, IRI-like terms should still be allowed to differ from their expanded forms (by niklasl)
Niklas Lindström: A couple of sessions ago we had a discussion on some detail, that I'll need to find.
... I was trying to find out why my test suite failed on something and thought I had missed something in the algorithm, but I found that the algorithm should behave differently if the algorithm is 1.0.
... It works fine in 1.1, but we don't describe the 1.0 behavior.
... To be correct, the algorithm should have a check that makes it behave differently in 1.0 processing mode.
... It should not through an IRI processing error unless processing mode is 1.0.
Gregg Kellogg: Could be considered an editorial change, as it doesn't affect 1.1 behavior.
Pierre-Antoine Champin: There may be some privilege loss that happens automatically.
Niklas Lindström: It's a "formal" issue, but doesn't affect any behavior in the wild I'm aware of.
David I. Lehn: In the API there are a couple of things that have been added.
W3c/json-ld-api#582
https://github.com/w3c/json-ld-api/issues/582 -> Pull Request 582 Add protected graph container null context test. (by davidlehn) [test:missing-coverage]
W3c/json-ld-api#585
https://github.com/w3c/json-ld-api/issues/585 -> Pull Request 585 Add graph container array tests. (by davidlehn) [test:missing-coverage]
David I. Lehn: 582 Is a case of missing test coverage.
... 585 deals with missing coverge.
David I. Lehn: In some cases, there is some spec work needed.
David I. Lehn: In some cases, we need to determine what output should be.

Topic: Open Discussion

David I. Lehn: Regarding the website, I've talked with bigbluehat about using a different infrastructure to manage the site.
... Previous change requests were for the bootstrap version, but it broke stuff. I think the right way to go is to update the static-site generation softward.
... The PHP proxy doesn't work right, and should be hosted elsewhere. It's hard to mix both static and dynamiting things together.
... Minutes take too much space, and we need to figure out how to manage.
... Also, the playground needs work and needs to be done differently to be able to maintain it.
Benjamin Young: When I worked on this over 7 years ago, we got partly converted to Jekyl, but got hung up.
... I wanted to see if we can go through 11P on a case-by-case basis.
... The biggest piece to take care of is the playground.
... DB needs to spend time focusing on this.
... It would be helpful for people to be able to review PRs.
Niklas Lindström: I can also give encouragement; I haves a bit more time to work on JavaScript related things and may be able to help with some of the work.
... We have a lot of experience with related technologies.
Benjamin Young: Rewriting the playground would be hard, and we should try to iterate it, maybe with a package manager.
... The playground could also go on a sub-domain, which could make it easier to host.
... The only bit that's not static is a proxy for CORS.
... The minutes could be moved somewhere as well.
... We can't screw up things too much either.
David I. Lehn: If people want to start on new Playground things we need to coordinate.
Gkellogg next meeting on February 7th.