- From: David Booth <david@dbooth.org>
- Date: Thu, 26 Feb 2015 19:15:39 -0500
- To: Josh Mandel <Joshua.Mandel@childrens.harvard.edu>
- CC: Jim McCusker <mccusj@rpi.edu>, "public-linked-json@w3.org" <public-linked-json@w3.org>, Manu Sporny <msporny@digitalbazaar.com>, Markus Lanthaler <markus.lanthaler@gmx.net>, Pat Hayes <phayes@ihmc.us>
On 02/26/2015 06:18 PM, Josh Mandel wrote:
> One quick note: order matters within FHIR JSON arrays, so if you need to
> support round-trips, you'd also want to add {"@container":"@list"} to
> your context for every property that allows multiple cardinality.
Thanks Josh, I was going to ask about that separately, but you beat me
to it!
That makes sense. That is unfortunate for RDF inference purposes,
because it means that inference rules will have to deal with objects in
RDF lists, but it sounds like there is no way around that.
But what about FHIR extensions? Am I correct in understanding that
FHIR has a closed content model, i.e., FHIR extensions cannot create any
new JSON properties?
http://www.hl7.org/implement/standards/fhir/extensibility.html
This seems crucial, because if extensions could create new properties
then we would also have to somehow record whether those new properties
could hold lists.
Thanks,
David Booth
Received on Friday, 27 February 2015 00:16:09 UTC