The W3C JSON-LD Community Group

Go Back

W3C Logo

JSON-LD CG

Minutes for 2023-08-09

Present
Gregg Kellogg, Niklas Lindström, Ted Thibodeau Jr., alex , David I. Lehn, Benjamin Young
Regrets
Gregory Saumier-Finch
Chair(s)
Gregg Kellogg
Scribe(s)
Agenda
https://www.w3.org/events/meetings/1ab7df94-bb06-440e-a6b9-bc9022018c57/?recurrenceId=20230809T120000
Topics
  1. Announcements and Introductions
  2. YAML-LD
  3. JSON-LD Issue Discussion

Topic: Announcements and Introductions

Topic: YAML-LD

Topic: JSON-LD Issue Discussion

Subtopic: https://github.com/w3c/json-ld-syntax/pull/417

Subtopic: https://github.com/w3c/json-ld-framing/pull/154

David I. Lehn: Not sure how the framing profile parameter [scribe assist by Gregg Kellogg]
Gregg Kellogg: 'Framed' was discussed as signifying a document that had been framed [scribe assist by Niklas Lindström]
Niklas Lindström: ... These things are probably not used (don't break any tests)
Gregg Kellogg: The "framed" fragment was intended to indicate documents that had been framed rather that that is a frame. [scribe assist by Gregg Kellogg]
Ted Thibodeau Jr.: I don't think it's likely to break anything, and it's certainly broken now. [scribe assist by Gregg Kellogg]
Niklas Lindström: https://www.w3.org/ns/json-ld#frame: As an IRI, this link relation is used to associate a JSON-LD frame with a JSON-LD document. In an HTTP request header, specifies the location of a frame to use for framing.
Niklas Lindström: https://www.w3.org/ns/json-ld#framed: This profile URI is used to request or specify framed JSON-LD document form.
Benjamin Young: Not at Wiley anymore, now at Digital Bazaar. Still working on status updates. [scribe assist by Gregg Kellogg]

Subtopic: https://github.com/w3c/json-ld-api/pull/572

David I. Lehn: Haven't looked at it quite yet. [scribe assist by Gregg Kellogg]
Gregg Kellogg: Propose merging [scribe assist by Gregg Kellogg]

Subtopic: https://github.com/w3c/json-ld-api/pull/559

David I. Lehn: This is still in draft, as more is needed. [scribe assist by Gregg Kellogg]
Gregg Kellogg: Open since february. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... 58 Files change.
David I. Lehn: Several tests added. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... Also related is w3c/json-ld-api#560
https://github.com/w3c/json-ld-api/issues/560 -> Issue 560 Various `@json` processing issues. (by davidlehn) [test:missing-coverage] [ErratumRaised]
Gregg Kellogg: ... There are other links in there, including the issue from jsonld.js.
Gregg Kellogg: ... The problem is that when you're using a JSON literal, how to you deal with arrays?
Gregg Kellogg: ... When you're reading it, it can be interpreted as a set of literals within the array, or a single literal that is an array.
Gregg Kellogg: ... It involves spec changes, as we didn't have tests for this originally.
Gregg Kellogg: ... We could say it only works with objects, but that seems like an unreasonable restriction.
Gregg Kellogg: ... The tests just cover some use cases, but the overall issue hits different aspects of the spec.
Gregg Kellogg: ... It's about declaring things so you can handle the use case, but it means that `@container` might mean something different for JSON arrays.
Gregg Kellogg: There are different ways that the container attribute might be interpreted. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... Do you process it differently with and without the array description?
Gregg Kellogg: ... We need some basic rules, but you need to be more explicit about it.
David I. Lehn: I worked on my implementation until it seemed to do what you would expect. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... You'd need to know about this when dealing with JSON types that are arrays.
Niklas Lindström: Are there use cases in the wild that would be problematic? [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... I could imagine coordinates used as JSON lists that might be misintrepreted.
Niklas Lindström: "JsonLists": {"@set": [ [1, 2, 3], [4, 5, 6] ]}
Gregg Kellogg: ... I agree that it's odd that you can't use arrays as literals; if you need to use both, you might use "set" containers.
Gregg Kellogg: ... And a container version of that.
Gregg Kellogg: Does w3c/json-ld-api#560 comprehensively reference the problem and proposed solutions? [scribe assist by Gregg Kellogg]
https://github.com/w3c/json-ld-api/issues/560 -> Issue 560 Various `@json` processing issues. (by davidlehn) [test:missing-coverage] [ErratumRaised]
David I. Lehn: That is what I attempted to do. The jsonld.js issue referenced is what started this. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... Using a set container might not seem right either.
Gregg Kellogg: Seems like a fairly big change to tackle in this forum. [scribe assist by Gregg Kellogg]
David I. Lehn: I haven't seen a bit outcry to fix this issue. [scribe assist by Gregg Kellogg]
Gregg Kellogg: Maybe defer a bit longer? [scribe assist by Gregg Kellogg]
David I. Lehn: I don't think I've pushed all the changes that make this work. [scribe assist by Gregg Kellogg]
Niklas Lindström: Would it be a normative change. [scribe assist by Gregg Kellogg]
David I. Lehn: I didn't write spec text, but I believe it would be a normative change. [scribe assist by Gregg Kellogg]
Niklas Lindström: I'm more keen to solve RDF List issues. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... I'm more in favor of making a change sooner rather than later.
Gregg Kellogg: I think we're on the road to JSON-LD 1.2, but it's probably a couple of years out. [scribe assist by Gregg Kellogg]
Niklas Lindström: I'd like to work on issues like this. [scribe assist by Gregg Kellogg]
Gregg Kellogg: Propose investigating further and bringing it up again in the next meeting. [scribe assist by Gregg Kellogg]

Subtopic: https://github.com/w3c/json-ld-api/issues/558

Niklas Lindström: I hope that this isn't _too_ normative a change. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... It's an undesireable behavior.
Gregg Kellogg: I think that makes it an erratum. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... There is an erratum lifecycle we can stick with.
Gregg Kellogg: ... The changed algorithm would make current implementations misbehave?
Gregg Kellogg: ... The WG is chartered to do this work, subject to time an process.

Subtopic: json-ld-syntax#355

Gregg Kellogg: This was a proposal to replace the term "IRI" with "URL". [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... Propose doing a PR to change references for IRI into RDF 1.2 Concepts.
Benjamin Young: There's a note about it in the RDF 1.2 Concepts doc: https://www.w3.org/TR/rdf12-concepts/#h-note-3
Benjamin Young: Maybe we do the same?
Benjamin Young: I suggest we follow RDF 1.2. Many people use the terms interchangeably. [scribe assist by Gregg Kellogg]
Gregg Kellogg: ... The key is that there is a difference between IRI and URL.
Gregg Kellogg: ... Beyond simply being the right term, there are practical implications.
Niklas Lindström: https://xkcd.com/927/ ... :/