Canonical JSON-LD format analogous to n-triples?

I have a question about recommending a canonical JSON-LD profile to RDF 
users.  I would like to recommend a profile of JSON-LD that would be 
convenient for exchanging RDF data in JSON format, for use both by 
RDF-oriented processors and by JSON-oriented processors.  The ideal 
would probably be something analogous to n-triples: very simple (few or 
no options), very predictable, and very easy to generate and consume by 
machine.  Ease of human authoring is not a goal in this case.

I would like to be able to point both RDF and JSON developers at the 
appropriate sections of the JSON-LD spec, so that they don't have to 
learn all the ins and outs of the context and other features of JSON-LD 
that they won't need when they are simply using JSON-LD in this way as a 
JSON serialization of RDF data.

The goal would be something very simple, analogous to RDF/JSON
https://dvcs.w3.org/hg/rdf/raw-file/default/rdf-json/index.html#
though I don't care if the exact same conventions are used.

What JSON-LD profile should I recommend for this?  A combination of 
Expanded and Flattened document forms?

Thanks,
David

Received on Tuesday, 16 July 2013 17:15:25 UTC