JSON-LD CG

Minutes for 2023-07-12

Present

Gregg Kellogg, Anatoly Scherbakov, Ted Thibodeau Jr., Niklas Lindström, David I. Lehn

Chair(s)

Gregg Kellogg

Scribe(s)

Gregg Kellogg

Agenda

https://www.w3.org/events/meetings/1ab7df94-bb06-440e-a6b9-bc9022018c57/20230712T120000/

Topics

Announcements and Introductions
YAML-LD
JSON-LD Issue Discussion
JSON-LD-star
Open Discussion

Gregg Kellogg is scribing.

Topic: Announcements and Introductions ✪

Topic: YAML-LD ✪

Subtopic: PR json-ld/yaml-ld#100 – Unify examples ✪

Anatoly Scherbakov: I didn't do any additional commits after the discussion from the last meeting.

Subtopic: PR json-ld/yaml-ld#104 – schema.org @context does not require a trailing slash ✪

Anatoly Scherbakov: https://github.com/json-ld/yaml-ld/pull/110

Anatoly Scherbakov: I want to double-check.

Subtopic: PR json-ld/yaml-ld#106 – Reference two Convenience contexts from json-ld.org ✪

Anatoly Scherbakov: I used an existing context as a basis for context.jsonld.

Anatoly Scherbakov: The schema.org context does similar things with the `id` keyword.

Niklas Lindström: I agree with anatoly-scherbakov; the generic context feels generic enough for other things. The "$" version could also be used in other cases.

Niklas Lindström: Also, regarding filenames; I'd recommend not using "$" in a filename, although it's legal, but string expansion/interpolation is common, and could trip up people.

... perhaps "dollar_convenience.jsonld"

Subtopic: PR json-ld/yaml-ld#107 – List supported YAML features concisely ✪

Anatoly Scherbakov: From information POV, I thought it would be beneficial to have a section that concisely says what we support.

Subtopic: PR json-ld/yaml-ld#108 – Reference external example files ✪

Anatoly Scherbakov: I wrote several yaml-ld files and added links to them.

... not quite ready to merge.

Subtopic: PR json-ld/yaml-ld#109 – Implement a Structured Abstract ✪

Anatoly Scherbakov: From requirements at a conference, I encountered "structured abstracts".

Anatoly Scherbakov: I won't be at TPAC.

Topic: JSON-LD Issue Discussion ✪

Subtopic: PR w3c/json-ld-syntax#415 – Add structured suffix (changes requested) ✪

Ted Thibodeau Jr.: I'm not sure what's going on with this.

Gregg Kellogg: We may do a follow-on PR for style.

Subtopic: PR w3c/json-ld-api#586 – Fix context processing for reverse terms (needs reviews. IPR clear?) ✪

Niklas Lindström: I'm close to being able to join the WG after changing affiliation to DCMI.

Niklas Lindström: I merged TallTed's suggestions earlier.

Topic: JSON-LD-star ✪

Subtopic: json-ld/json-ld-star#45 – Keeping quoted triples sharing subject within one node object ✪

Niklas Lindström: I think pchampin probably would like to chime in. I need to extract parts of this as a use case for RDF-star.

... There are a lot of things going on there.

... It would be nice to use JSON-LD to succinctly describe this use case.

... It's related to the open questions on annotations in RDF-star. What appears to be transparent should be opaque.

... It might easier to go forward with if the annotation syntax questions are resolved.

Niklas Lindström: This uses but doesn't adequately discuss the relationship between quoted triples and named graphs.

... It would be interesting to see how graph terms (from N3) might relate to quoted triples.

... The relationship between named graphs and quoted triples. For the examples in the PR frame the ideas.

... The named graphs are used as two versions of the library catalog. The latest version is mirrored, but the core system stores different versions.

... The second example is like the union of both versions using a custom algorithm to make a "blame view", which uses quoted triples to show what has been said.

... Also, there's a quoted triple in the original, that is a "suggestion". So, it combines related use cases.

... One use case is making suggestions, the other showing version history.

Gregg Kellogg: There is also a `@quoted` keyword.

Niklas Lindström: That comes from a case where I have a lot of quotes, and serializing using existing formates makes a lot of assertions on the same subject.

... It would be nice to have an editing interface using a representation which groups quotes about the same subject syntactically in a tree would be useful.

... Just like is used for other forms within RDF serializations, but there is no different semantics.

Niklas Lindström: This is an issue, not a PR.

Gregg Kellogg: It would be good to disentangle into more primitive issues.

Topic: Open Discussion ✪

Subtopic: TPAC ✪

Gregg Kellogg: Proposal to give up the TPAC slot.

Anatoly Scherbakov: +1

Niklas Lindström: I agree

Ted Thibodeau Jr.: I don't usually attend in person.

Anatoly Scherbakov: I was looking forward to it, but my own financial limitations make it hard.

David I. Lehn: Were there any other topics.

... I'd also like to present something on "safe mode", which we've added to our implementation. But, that would involve me talking about it.

... There are some other issues about JSON literals.

... Some of the topics are good for people to think of over time, but in person time can be very constructive.

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

Niklas Lindström: Another topic of interest is on RDF lists in JSON-LD.

... Also, I don't know if there's interest in expression N3 in JSON-LD.

... This could crop up laster, as lists are used heavily in N3.

... I appreciate that RDF lists can be reduced to triples, but conceptually, they should perhaps be considered their own resources.

... Might relate to some notion of "abstract syntax trees"

... All RDF libraries have convenience methods for lists.

... I'm playing with hand-writing a SPARQL parser to see if I can represent SPARQL in JSON-LD.

... We have use cases which would be promising if there was a JSON representation of SPARQL.

David I. Lehn: We could fill up the TPAC time talking about the future.

The W3C JSON-LD Community Group

Go Back