W3C home > Mailing lists > Public > semantic-web@w3.org > October 2016

JSON-LD Community group working on updated drafts

From: Gregg Kellogg <gregg@greggkellogg.net>
Date: Tue, 4 Oct 2016 12:21:28 -0700
Message-Id: <B07766CC-8322-42EC-A333-2C2DBBA0E44C@greggkellogg.net>
Cc: Linked JSON <public-linked-json@w3.org>
To: Linked Data community <public-lod@w3.org>, Semantic Web <semantic-web@w3.org>, W3C Web Schemas Task Force <public-vocabs@w3.org>
As many people on these lists likely know, the work on JSON-LD was originally fostered in the still active JSON for Linking Data Community Group [1]. Much of this work became input to the RDF 1.1 Working Group, which resulted in the publication of JSON-LD 1.0 [2] and JSON-LD API 1.0 [3] in January 2014.

Since this time, JSON-LD has seen great adoption, but there remain some issues that the group could not complete at the time, or which came to light subsequent to publication:

* JSON-LD Framing has been important for several groups, but it was left out of the 1.0 API document for lack of time. [4] (and several others)

* Named graphs are in greater use, but the JSON-LD support lacks a way to elegantly use this when considered as plain JSON. There is an issue [5] to create a @graph container type, which would make this more useful.

* Indexing allows greater use of key/value pairs without needing to sort through an array of different values, but it leaves out important use cases where properties tend to be gathered under a common key. [6]

* There is a common misunderstanding on the role of the @context, and the ability to define data in the context that is logically a part of each node value that uses it (for example, having a common @type). Similarly, people want the ability to change a context used for values of a given term, so that say the @base, or @vocab may be different, or a given term takes on a different meaning. This becomes important when composing data from multiple sources. Issue #426 [7] offers a comprehensive solution to many of these problems.

* GeoJSON and other specifications work awkwardly with JSON-LD because of the inability to assign meaning to different positions within an array. There are a couple of ways to look at this: one is to support JSON property values, there are also proposals to provide meaning to positional data. [8]

There are a number of other issues to work through as well, and more that may be brought up. Find a current list on GitHub [9].

Publication of updated specs by a Community Group contains no official weight, but does offer some patent protection. It also could potentially be input to some future Working Group which could generate a Recommendation.

The community maintains current drafts at http://json-ld.org:

* http://json-ld.org/spec/latest/json-ld/
* http://json-ld.org/spec/latest/json-ld-api/
* http://json-ld.org/spec/latest/json-ld-framing/

I encourage anyone interested to join the conversation at public-linked-json@w3.org, on GitHub at https://github.com/json-ld/, or on IRC at #json-ld on Freenode.

Gregg Kellogg
gregg@greggkellogg.net

[1] https://www.w3.org/community/json-ld/
[2] http://json-ld.org/spec/REC/json-ld/20140116/
[3] http://json-ld.org/spec/REC/json-ld-api/20140116/
[4] https://github.com/json-ld/json-ld.org/issues/110
[5] https://github.com/json-ld/json-ld.org/issues/195
[6] https://github.com/json-ld/json-ld.org/issues/246
[7] https://github.com/json-ld/json-ld.org/issues/426
[8] https://github.com/json-ld/json-ld.org/issues/397
[9] https://github.com/json-ld/json-ld.org/issues?q=is%3Aopen+is%3Aissue+label%3A1.1
Received on Tuesday, 4 October 2016 19:21:59 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 4 October 2016 19:22:02 UTC