15:00:16 <RRSAgent> RRSAgent has joined #hcls
15:00:21 <RRSAgent> logging to https://www.w3.org/2024/03/21-hcls-irc
15:00:24 <dbooth> rrsagent, make logs public
15:00:31 <dbooth> Meeting: FHIR RDF
15:00:34 <dbooth> Chair: David Booth
15:02:53 <dbooth> Topic: IRI stems
15:04:05 <dbooth> gaurav: Stuck in content check because THO is still on R4, but our change is in R5.  When they try to update, it causes an error.  Mark Dutto (sp?) is looking into it.
15:04:36 <dbooth> ... May be nothing to do until thye go to R5.  Could put the iri-stem into the comment field until then.
15:05:44 <dbooth> .. Second update: Emailed radlex, forum.  It's obvious what their IRI stem is, and they refer to it as their PURL, and use it in their OWL file, but unable to find official documentaiton about it.
15:06:23 <dbooth> ... In email I suggested a way to move forward.
15:07:37 <dbooth> dbooth: it should be at the follow-yuour-nose location
15:08:07 <dbooth> gaurav: Re OBO ont, there are only two on THO: Mondo and HPO.
15:08:27 <dbooth> ... For mondo, have to look at two docs to figure out IRI stem.
15:09:04 <dbooth> ... We can cite them.   Emailed them about having it in one place.
15:09:19 <dbooth> ... Haven't found it for HBO yet.
15:09:36 <dbooth> jim: OBO foundry ID policy should do it.
15:09:48 <dbooth> gaurav: Yes, but I need it for HPO.
15:10:44 <dbooth> Topic: DICOM
15:11:19 <dbooth> erich: Raised an issue on json-ld group, to speack w Gregg Kellogg.  In DICOM a lot of fields are multi valued, ordering matters, and a lot of them involve nulls.
15:12:05 <dbooth> ... No good sparql for it.  Asked Gregg how when you put nulls in, then get filtered out or used for defaults in contexts.  Except for lists.
15:12:50 <dbooth> ... In RDF null means not expressing the triple.  Remove first list element to indicate null?  Gregg said thats not well formed RDF list.
15:13:10 <dbooth> ... HE said sparql group is discussing what is well formed list.
15:13:39 <dbooth> ... Jena gives an unbound var in that situation.
15:13:57 <dbooth> ... Gregg said json-ld will follow sparql group.
15:14:18 <dbooth> ... RDF seq collections have the same issue.  Could leave out an element to indicate null.
15:15:06 <dbooth> ... Would be clean if lists w nulls are supported.  It's in the hands of the sparql group.
15:15:51 <dbooth> dbooth: May have stepped into a hornet's nest.
15:16:39 <dbooth> dbooth: What about using an explicit null URI?
15:16:51 <dbooth> erich: No discussion on that.
15:17:28 <dbooth> ... In jena you can map a native type to an RDF type.
15:17:46 <dbooth> ... e.g., to deal w numeric or date types.
15:26:16 <dbooth> dbooth: value spaces like int don't often allow null.
15:28:13 <dbooth> erich: Will try to speak w sparql WG folks.   But they're talking about reification.
15:28:59 <dbooth> Topic: FHIR conversion testing
15:29:13 <dbooth> jim: Spoke w Tim Prudhomme, but haven't had time to pursue.
15:30:58 <dbooth> Present: David Booth, Gaurav Vaidya, Jim Balhoff, Rob Hausam, Erich Bremer Detlef Grittner.
15:31:04 <dbooth> ADJOURNED
15:31:11 <dbooth> rrsagent, draft minutes
15:31:13 <RRSAgent> I have made the request to generate https://www.w3.org/2024/03/21-hcls-minutes.html dbooth
16:29:14 <TallTed> TallTed has joined #hcls