Stream: IG creation
Topic: CapabilityStatement and core extension search
John Moehrke (Dec 17 2021 at 22:37):
When an IG is calling on a FHIR core extension, and that core extension has a defined search parameter... (e.g. mothersMaidenName).
John Moehrke (Dec 17 2021 at 22:38):
It is clear how to call for the extension in a StructureDefinition.
John Moehrke (Dec 17 2021 at 22:38):
It is not clear what all is needed in a CapabilityStatement (requirements) to call for the search parameter for that extension.
John Moehrke (Dec 17 2021 at 22:40):
is it sufficient to just address it the same way as core defined parameters?
John Moehrke (Dec 17 2021 at 22:41):
or one must find the canonical URI of the SearchParameter? For example for mothersMaidenName
http://hl7.org/fhir/SearchParameter/patient-extensions-Patient-mothersMaidenName
John Moehrke (Dec 17 2021 at 22:42):
if the canonical URI is needed, then we really MUST make these more clear. I only found this by searching deeply the FHIR core build output. I can't find mention of this anywhere in the narrative.
Vassil Peytchev (Dec 17 2021 at 22:45):
It is needed, and I learned to find the canonical in the core package (since it is a tar.gz package, you can just look inside with you favorite compression app).
John Moehrke (Dec 17 2021 at 22:46):
yup, geeks can find... but not the kind of author we want to attract to IG publication.
John Moehrke (Dec 17 2021 at 22:49):
Lloyd McKenzie (Dec 17 2021 at 23:38):
Having the canonical urls visible in the core spec is an agreed but unapplied change
Eric Haas (Dec 18 2021 at 02:59):
John Moehrke said:
yup, geeks can find... but not the kind of author we want to attract to IG publication.
what kind of author do you want to attract vs what we have today?
Lloyd McKenzie (Dec 18 2021 at 03:05):
We certainly want IG creation to not require digging into download packages to search for search parameter URLs...
Eric Haas (Dec 18 2021 at 03:29):
Do you mean it is too hard to know where to look or what you are looking for?
Sarah Gaunt (Dec 18 2021 at 03:53):
John Moehrke said:
yup, geeks can find... but not the kind of author we want to attract to IG publication.
Choosing not to be offended! :joy:
Lloyd McKenzie (Dec 18 2021 at 04:51):
IG developers shouldn't have to dig for information they need to do their job.
Eric Haas (Dec 19 2021 at 19:17):
Yes well very little value there IMO.
John Moehrke (Dec 20 2021 at 13:49):
finding the canonical URI for a search parameter that is defined in the FHIR core seems rather important. The alternative that many will choose is to just re-invent the exact same SearchParameter so they can use their own canonical URI.
Lloyd McKenzie (Dec 20 2021 at 15:16):
Right. Which is why it's an agreed change for R5.
Eric Haas (Dec 20 2021 at 16:49):
So lemme get this straight it is easier to make a new SP than dig up a URI ...
John Moehrke (Dec 20 2021 at 17:13):
if the solution is to dig thru the FHIR core source... yes. Many people have no access to the FHIR core source. -- in fact the human readable part that is in FHIR core gives you the string you would add to your custom SearchParameter, so it seems to be guiding you to create your own.
Eric Haas (Dec 20 2021 at 17:43):
For the rest of us is easier to to navigate a package or read the FHIR spec here: http://build.fhir.org/searchparameter-registry.html and figure it out :-)
John Moehrke (Dec 20 2021 at 18:23):
@Eric Haas mothersMaidenName search parameter is NOT IN THAT REGISTRY. That is my point. That is my WHOLE point. That is what @Lloyd McKenzie says will get fixed.
Lloyd McKenzie (Dec 20 2021 at 19:25):
No one should have to go to a registry to see what the URL is for something in the core spec. The URLs should be clearly visible in the human-readable part of the spec.
Last updated: Apr 12 2022 at 19:14 UTC
