14:59:29 <RRSAgent> RRSAgent has joined #hcls
14:59:29 <RRSAgent> logging to http://www.w3.org/2016/03/22-hcls-irc
14:59:31 <trackbot> RRSAgent, make logs world
14:59:31 <Zakim> Zakim has joined #hcls
14:59:33 <trackbot> Zakim, this will be HCLS
14:59:33 <Zakim> I do not see a conference matching that name scheduled within the next hour, trackbot
14:59:34 <trackbot> Meeting: Semantic Web Health Care and Life Sciences Interest Group Teleconference
14:59:34 <trackbot> Date: 22 March 2016
15:03:09 <Tony> Tony has joined #HCLS
15:07:38 <rhausam> rhausam has joined #HCLS
15:08:31 <rhausam> I have some conflicts today - will try to join audio soon
15:19:13 <dbooth> Topic: Loading FHIR instance data into Protege
15:19:43 <dbooth> Eric: You can use a separate loader file, that imports both your ontology and then your instance data, and load that loader file.
15:20:39 <dbooth> Tony: I'll try it
15:23:25 <dbooth> tony: It looks like it assigns local genid names to bnodes on loading.
15:24:12 <dbooth> eric: Yes, because bnodes are temporary names.  They can be renamed at any time by the engine.
15:25:20 <dbooth> ... Maybe I can talk to the maintainers of the code this week at AMIA.
15:25:33 <dbooth> ... Changes to OWL API are reflected in Protege within a day.
15:27:08 <dbooth> Topic: Editing the FHIR RDF document
15:28:09 <dbooth> dbooth: Grahame wants the edits to be done directly on http://w3c.github.io/hcls-fhir-rdf/spec/rdf.html for version control reasons
15:29:36 <dbooth> dbooth: Should at least align with http://w3c.github.io/hcls-fhir-rdf/spec/rdf.html as soon as possible.
15:30:13 <dbooth> dbooth: Rob, did you get section 4 put in?
15:30:20 <dbooth> rob: Not yet.
15:30:35 <dbooth> dbooth: It becomes section 3 in the github version.
15:33:16 <dbooth> eric: deepak edited section 3, without knowing that rob was doing it.  :(
15:34:29 <dbooth> eric: The purple sections in http://w3c.github.io/hcls-fhir-rdf/spec/rdf.html are duplicated with the Ontology page, and will need to be de-duplicated at some point.
15:35:18 <dbooth> s|directly on http://w3c.github.io/hcls-fhir-rdf/spec/rdf.html|directly on https://hl7-fhir.github.io/allergyintolerance-example.ttl.html|
15:35:23 <dbooth> rrsagent, draft minutes
15:35:23 <RRSAgent> I have made the request to generate http://www.w3.org/2016/03/22-hcls-minutes.html dbooth
15:35:38 <dbooth> eric: I will edit the doc for a week
15:37:25 <dbooth> ... Maybe we can integrate with Grahame's version soon after that, and then switch to the subversion version?
15:40:26 <dbooth> dbooth: Would like to capitalize on Rob's work too.
15:41:03 <dbooth> ACTION: Rob to edit section 3 (Terminology)
15:41:03 <trackbot> Created ACTION-50 - Edit section 3 (terminology) [on Rob Frost - due 2016-03-29].
15:41:23 <dbooth> ACTION: Eric to edit the FHIR RDF doc except section 3
15:41:23 <trackbot> Created ACTION-51 - Edit the fhir rdf doc except section 3 [on Eric Prud'hommeaux - due 2016-03-29].
15:42:27 <dbooth> Topic: Revisiting fhir:resourceType to identify the root element
15:44:37 <dbooth> dbooth: Grahame is pushing back on using fhir:resourceType to identify the root element: "Any time a resource is included - whether it is contained or not, it will have a fhir:resourceType"
15:46:57 <dbooth> dbooth: We said that the fhir:resourceType should have a *string* value that is the name of the root element type.
15:47:31 <dbooth> ... But rdf:type has a different purpose.
15:50:14 <dbooth> dbooth: "fhir:resourceType is not intended to replace rdf:type declarations.  Our intent is to use fhir:resourceType *only* to indicate the root element of a FHIR resource instance.  It is not intended for use on other contained resources."
15:50:38 <rhausam_> rhausam_ has joined #HCLS
15:51:51 <dbooth> ... It is indicating two things: 1. which node is the root node; and 2. the type of that node (as a string).
15:52:32 <dbooth> s/type/FHIR type/
15:55:36 <dbooth> The same two facts could be indicated in two separate statements instead:
15:55:37 <dbooth>  :myAllergyIntolerance fhir:isRootNode true .
15:55:37 <dbooth>  :myAllergyIntolerance fhir:hasFhirTypeAsString "AllergyIntolerance .
15:58:36 <ericP> i wonder if this would help avoid confusion: :hasFhirElementName "AllergyIntolerance"
15:59:47 <ericP> http://www.w3.org/2016/FHIR-tutorial/Constellations?logical
16:00:20 <ericP> look for a fhir:Specimen
16:00:24 <ericP> look for "a fhir:Specimen"
16:03:22 <dbooth> eric: If we look for anything that is one of these resource types, and use that as the root, we would not be able to roundtrip, because we have both fhir:DiagnosticOrder and a fhir:Specimen and we do not know which one should be the root.
16:05:05 <dbooth> ... So we need a way to signal that <DiagnosticOrder/12345-68> is the root node.
16:06:25 <dbooth> Present: Tony Mallia, David Booth, EricP, Rob Hausam, Thomas Lukasik
16:11:02 <dbooth> eric: All FHIR resources in JSON will have resourceType property on them, even nested ones.
16:12:37 <dbooth> eric: Suggest using rdf:type instead of fhir:resourceType to indicate the type of every FHIR resource.
16:12:56 <dbooth> ... And explicitly indicate the root node using fhir:isRootNode
16:13:04 <dbooth> dbooth: Like this:  :myAllergyIntolerance fhir:isRootNode true .
16:15:41 <dbooth> or:   :myAllergyIntolerance a fhir:RootNode .
16:18:05 <dbooth> Topic: 5pm call?
16:18:30 <dbooth> dbooth: Let's see who shows up.  If we don't have sufficient people by 5:10pm, I'll cancel.
16:18:52 <dbooth> Chair: David Booth
16:18:57 <dbooth> rrsagent, draft minutes
16:18:57 <RRSAgent> I have made the request to generate http://www.w3.org/2016/03/22-hcls-minutes.html dbooth
18:13:21 <Zakim> Zakim has left #hcls