Stream: IG creation
Topic: IG for environment that does not use terminology.hl7.org
John Moehrke (Nov 02 2020 at 15:09):
I am writing IG for environments that are using the original URI for terminology (hl7.org/fhir/CodeSystem), but the IG build forces me to use terminology.hl7.org... is there a configuration I can use?
Jose Costa Teixeira (Nov 02 2020 at 15:45):
there is a command line parameter -tx xxxxx where xxxxxx is the other terminology server
Jose Costa Teixeira (Nov 02 2020 at 15:46):
I don't know if this is supposed to support your need
Jose Costa Teixeira (Nov 02 2020 at 15:46):
(i think the IG publisher uses tx.fhir.org ?)
John Moehrke (Nov 02 2020 at 16:47):
IG publisher is forcing me to use terminology.hl7.org... I would try the parameter, but don't understand how
Jose Costa Teixeira (Nov 02 2020 at 16:51):
when you run genOnce.bat, run _genOnce.bat -tx your.tx-server.com - but that is just for the server itself.
I don't know if this will solve the problem though because hl7.org/fhir/CodeSystem does not seem like a running terminology server
David Pyke (Nov 02 2020 at 16:51):
just modify the java call to th publisher with -tx XXX.XXXXXX.XXX
Oliver Egger (Nov 02 2020 at 20:04):
@John Moehrke do you mean codes with system hl7.org/fhir/CodeSystem? because tx Parameter is for setting the terminology server for validation and I don't think this has an effect (except to disable terminology validation and then it would probably not give an error).
Grahame Grieve (Nov 02 2020 at 20:26):
the terminology server is not relevant to the question
Grahame Grieve (Nov 02 2020 at 20:26):
I don't know why you say it "forces" you to use terminology.hl7.org. Perhaps you could discuss specific errors that you are getting
John Moehrke (Nov 03 2020 at 01:16):
When my profile uses hl7.org/fhir/CodeSystem I get the following
C:\Users\john.moehrke\Git\MHV-PGHD\input\profiles\StructureDefinition-VA.MHV.bodyWeight.json
Path Severity Message
StructureDefinition/VA.MHV.bodyWeight: StructureDefinition Process Info Validate resource against profile http://hl7.org/fhir/StructureDefinition/StructureDefinition
StructureDefinition/VA.MHV.bodyWeight: StructureDefinition.snapshot.element[21].pattern.ofType(CodeableConcept).coding[0] error Unknown Code System 'http://hl7.org/fhir/CodeSystem/observation-category'
StructureDefinition/VA.MHV.bodyWeight: StructureDefinition.snapshot.element[21].pattern.ofType(CodeableConcept).coding[0].system error URL value 'http://hl7.org/fhir/CodeSystem/observation-category' does not resolve
StructureDefinition/VA.MHV.bodyWeight: StructureDefinition.differential.element[11].pattern.ofType(CodeableConcept).coding[0] error Unknown Code System 'http://hl7.org/fhir/CodeSystem/observation-category'
StructureDefinition/VA.MHV.bodyWeight: StructureDefinition.differential.element[11].pattern.ofType(CodeableConcept).coding[0].system error URL value 'http://hl7.org/fhir/CodeSystem/observation-category' does not resolve
I do not get these errors when I use terminology.hl7.org/CodeSyste
John Moehrke (Nov 03 2020 at 01:23):
so, my Profile has (FSH)
Lloyd McKenzie (Nov 03 2020 at 01:40):
If you're building against R4, you need to use the terminology.hl7.org canonical
John Moehrke (Nov 03 2020 at 01:49):
but the system I am testing against are using hl7.org/fhir.... thus the data we would be validating will NEVER have terminology.hl7.org
John Moehrke (Nov 03 2020 at 01:49):
not everyone can switch to new stuff as HL7switches to new stuff.
Lloyd McKenzie (Nov 03 2020 at 02:23):
The systems you're testing are non-conformant, and the validator isn't going to pretend they are. You can ignore the errors, but the errors are real.
John Moehrke (Nov 03 2020 at 16:17):
so there is no support for IGs on DSTU2, STU3, or original release of R4????? Surely I am just missing something.
John Moehrke (Nov 03 2020 at 16:18):
meaning that an IG published back then, better not need to be re-compiled as that will fail?
John Moehrke (Nov 03 2020 at 16:19):
I am happy to use terminology for valuesets... but using it as the .system is problematic when the environment the system is being used was locked into configuration prior to HL7 deciding to move to terminology system URI
Lloyd McKenzie (Nov 03 2020 at 19:33):
If you define the IG as being based on DSTU2, then you can (and must) use the original http://hl7.org/fhir URL. If you're using R4, then you must use the terminology.hl7.org URL. All releases of R4 have used that code system URL.
Lloyd McKenzie (Nov 03 2020 at 19:33):
If you're converting DSTU2 data to R4, you have to translate the URLs as part of the conversion.
John Moehrke (Nov 03 2020 at 19:48):
I do need to have what I am working on in DSTU2 (yuck, I know.. forced by circumstances). I was doing that originally but had issues creating a DSTU2 IG using current build tools... I can switch back to fighting that fight.
Lloyd McKenzie (Nov 03 2020 at 19:56):
The tools should work for creating DSTU2, though few people are still doing that, so if we've broken things the test cases don't capture, it's possible no one's noticed and you'll be the first to bleed on any new rough edges. So report if you hit issues. But you can't build an R4 IG that will work for DSTU2 content.
John Moehrke (Nov 03 2020 at 20:13):
promises, promises... :-)
Lloyd McKenzie (Nov 03 2020 at 20:26):
Promising 'should' is easy :)
Grahame Grieve (Nov 03 2020 at 20:29):
your URLs are wrong, that's why there's a problem
Jean Duteau (Nov 03 2020 at 22:11):
John Moehrke said:
When my profile uses hl7.org/fhir/CodeSystem I get the following
I do not get these errors when I use terminology.hl7.org/CodeSyste
I just checked STU2 and the URLs don't have CodeSystem in them.
http://hl7.org/fhir/observation-category
Jean Duteau (Nov 03 2020 at 22:11):
Grahame Grieve said:
your URLs are wrong, that's why there's a problem
What he said. :)
Last updated: Apr 12 2022 at 19:14 UTC
