JSON-LD Working Group Telco — Minutes

Date: 2019-10-26

See also the Agenda and the IRC Log

Attendees

Present: Rob Sanderson, Adam Soroka, Pierre-Antoine Champin, David I. Lehn, Dave Longley

Regrets: Gregg Kellogg, Benjamin Young, Jeff Mixter, Harold Solbrig

Guests:

Chair: Rob Sanderson

Scribe(s): Adam Soroka

Content:

1. Approve minutes of previous call

Proposed resolution: Approve minutes of previous call: https://www.w3.org/2018/json-ld-wg/Meetings/Minutes/2019/2019-10-18-json-ld (Rob Sanderson)

Rob Sanderson: +1

Adam Soroka: +1

Pierre-Antoine Champin: +1

Ivan Herman: +1

Resolution #1: Approve minutes of previous call: https://www.w3.org/2018/json-ld-wg/Meetings/Minutes/2019/2019-10-18-json-ld

David I. Lehn: +1

2. Announcements / Reminders

Rob Sanderson: We are rapidly apporaching CR

Pierre-Antoine Champin: next week’s call will be 1h earlier in Europe

Rob Sanderson: 5PM central europe, 4PM UK

3. Issues

3.1. Register profiles with IANA

Rob Sanderson: we have some new URIs for values of profile
… they must be registered with IANA
… bigbluehat offered to do this

Action #1: register profiles with IANA (Benjamin Young)

Ivan Herman: this should go via W3C channels, not directly
… not sure what the mechanism is for profiles
… I will ask

Action #2: check with plh if he is the right contact person (Ivan Herman)

Rob Sanderson: for the annotation WG it was easy
… no more topics to be discussed on calls

Rob Sanderson: https://github.com/orgs/w3c/projects/4?fullscreen=true

3.2. application/jsonld issue #287

Rob Sanderson: https://github.com/w3c/json-ld-syntax/issues/287

Ivan Herman: am representing the DID WG
… it is in that capacity that I raised this question

Dave Longley: this might be better if people could rad over the DID spec issue first,
… but it’s not clear how to handle mimetypes
… this is documented at the discussion there (at DID)

Pierre-Antoine Champin: the goal is to be able to declare application/did+jsonld?
… ?

Ivan Herman: Yes. It’s unclear whether application/did+ld+json is acceptable

Dave Longley: no one does this^ … but it doesn’t seem illegal … and questionable what tools would do.

Ivan Herman: there is no current registered mimetype that uses two + signs
… so we are not sure if this is really possible

Dave Longley: VC WG did not choose a mimetype

Dave Longley: VCWG didn’t pick a MIME type that i can see

Ivan Herman: that is a typical case where this might make sense

Rob Sanderson: https://www.w3.org/TR/annotation-protocol/#annotation-retrieval

Rob Sanderson: for Annotations, we went with a profile
… you can trigger a file extension from a profile, says IANA. If so, that would fulfill the requirements listed on the DID issue comment
… but I don’t have a reference for that
… we would have to go search the minutes

Rob Sanderson: https://www.iana.org/assignments/profile-uris/profile-uris.xhtml

Ivan Herman: we defined no extension for annotation

Action #3: find reference to file extensions for profiles re annotation (Rob Sanderson)

Rob Sanderson: if we did the application/jsonld, processors would have to know about it, which would be a normative change to the document

Ivan Herman: it would change the documents to allow this new media type in parallel to the older one
… it is simple for the document, but causes a lot of change elsewhere, because JSON-LD processors would have to change, too.

Rob Sanderson: we aren’t going to decide this today
… if there is no change here, we are on-track for CR
… a change here would push CR back by at least 2 weeks with Thanksgiving looming

Dave Longley: we want this group to enable other groups to be able to extend jsonld MIME type w/ file extension support – just a question of whether we need to register application/jsonld or not … either way should provide some advice to extenders on what to do

Dave Longley: we did discuss this in the 1.0 days, but it might not recorded in issues

Rob Sanderson: https://github.com/w3c/did-spec/issues/1#issuecomment-545103700

Rob Sanderson: if you can register a file extension to a profile, than that is what we could recommend
… how does application/jsonld solve the problem?

Ivan Herman: because then you can do application/did+jsonld
… no multiple plus signs

Rob Sanderson: this came up in XML days (application/xhtml+xml)

Ivan Herman: i couldn’t find any example of this (multiple plus signs) in IANA
… if I have a doc that is application/did+ld+json, and a JSON-LD proc, then that proc should handle it fine regardless of the DID aspect.
… if I do application/did+json, then a JSON-LD proc may or may not work correctly
… (handle it as JSON-LD)

Pierre-Antoine Champin: at the discussion at the DID WG, there is discussion of an alias
… is this an official construct of IANA?

Ivan Herman: I don’t know, but i think not; it just reflects that two media types may share suffix (but I may be wrong)
… I propose that azaroth digs up the remark from IANA about file extensions and we’d come back to the issue

Pierre-Antoine Champin: a profile approach could look pretty complicated to some people
… it might be a shame, in terms of marketing

Dave Longley: +1

Rob Sanderson: dlongley, the two WGs can meet at the next JSON-LD WG meeting
… to discuss this

Action #4: comment in issue inviting DID WG to JSON-LD WG regular call next week to discuss (Rob Sanderson)

3.3. closable issues

Rob Sanderson: https://github.com/w3c/json-ld-syntax/issues/269

Rob Sanderson: https://github.com/w3c/json-ld-framing/issues/69

4. Adjourn :)

5. Resolutions

6. Action Items