Stream: committers
Topic: tasks in error
Grahame Grieve (Mar 18 2016 at 07:33):
task #9175 is in error. It appears to have been approved as a typo, but it's not a typo, and, in fact, it's not a mistake
Grahame Grieve (Mar 18 2016 at 07:54):
task #8990 - I do not believe this is a typo. I'm marking it as unapproved, and referring it to FHIR-I
Grahame Grieve (Mar 21 2016 at 04:12):
#task 8980 - I think this is just flat out wrong. References can be relative, not absolute. And for some reason it's already changed differently... any thoughts?
Brian Postlethwaite (Mar 21 2016 at 04:15):
I agree, that invariant prevents relative references being used at all too, but in that case should start with "ElementDefinition/"
Brian Postlethwaite (Mar 21 2016 at 04:17):
But Michael is right, only when using the ResourceReference type, not the *uri* which should be either the http or urn
Grahame Grieve (Mar 21 2016 at 04:19):
would usually start with, but not SHALL start with
Grahame Grieve (Mar 21 2016 at 04:19):
uri can start with urn, so could add that instead
Brian Postlethwaite (Mar 21 2016 at 04:21):
yup.
Grahame Grieve (Mar 21 2016 at 12:41):
http://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=9428&start=0: the resolution is confusing, because it's not possible to 'add a parameter' without signficant effect. can we discuss whether we really need to do that?
Grahame Grieve (Mar 21 2016 at 12:42):
http://gforge.hl7.org/gf/project/fhir/tracker/?action=TrackerItemEdit&tracker_item_id=9427&start=0 - what is the actual problem here? I can't see it
Lloyd McKenzie (Mar 21 2016 at 13:32):
9428 - significant effect how? And regardless of effect, yes it needs to be done. The populate operation must be able to pass back both a QuestionnaireResponse and an OperationOutcome.
Lloyd McKenzie (Mar 21 2016 at 13:34):
For 9427, the first paragraph should be multiple paragraphs, but line breaks aren't showing up.
Grahame Grieve (Mar 21 2016 at 20:04):
9427: make it multiple paragraphsin the source, then
Grahame Grieve (Mar 21 2016 at 20:05):
9428 - the operation is defined with a single parameter called 'return', so servers return just the populated questionnaire directly
Grahame Grieve (Mar 21 2016 at 20:05):
changing that will mean that a parameters is returned. what's the point of an operation outcome?
Lloyd McKenzie (Mar 21 2016 at 20:27):
On 9427, it did have them. They got stripped out when you removed the custom formatting from all of the spreadsheets. I've now put it back. (Though possible the same issue exists elsewhere). On 9428, it's going to be pretty common when populating a questionnaire that some pieces won't work cleanly - couldn't find certain pieces in the submitted CDA, couldn't map a particular code to the answer set allowed by a question, etc. You'll want to pass back some or all of this as warnings. The change will mean that rather than a simple "return", the response will now be a Parameters instance allowing both the QuestionnaireResponse and the list of warnings (if any)
Grahame Grieve (Mar 23 2016 at 00:26):
#9179 - it says to ducmentation a whole lot of stuff about conformance resources, but it says to do s o on the Reference data type. I don't think that this is correct - it should be somewhere in the conformance space
Grahame Grieve (Mar 23 2016 at 00:35):
is it just me who finds the resolution of 8885 vague? I'm going to say that it means that _summary amd _elements should be allowed for read, vread, search, history, and GET operations. But why should it not apply to POST operations? Or maybe it shouldn't apply to operations - that's a too much to ask for....
Grahame Grieve (Mar 23 2016 at 00:35):
I think it's too much
Brian Postlethwaite (Mar 23 2016 at 02:48):
I'm with you on that one Grahame, I've implemented it for the POST operations also in our server.
Grahame Grieve (Mar 23 2016 at 05:21):
@Eric Haas task 8527 - should be revisited?
Grahame Grieve (Mar 23 2016 at 05:48):
Task 8630 got approved, but no one actually specified an OID. Do I just register one in the Hl7 OID regsitry?
Grahame Grieve (Mar 23 2016 at 23:10):
another task in error: 8723 says: Add "binding" to ElementDefinition.base (for the usecase where extensible bindings need information about bindings in the parent to correctly decide whether a concept is already present in the ancestry of bindings).
Grahame Grieve (Mar 23 2016 at 23:10):
but this is wrong; the base is the reference to the base type, not type this is derived from. So it won't let you check the 'ancestry of the bindings', only against the ultimate base. And I think that if you're doing that kind of processing, you should just chase down the derivation yourself and figure it out
Brian Postlethwaite (Mar 24 2016 at 05:21):
is it just me or are some of the links on https://hl7-fhir.github.io/terminologies-v3.html broken?
-practice-codes
Brian Postlethwaite (Mar 24 2016 at 05:24):
(some work, maybe its a website deploy in progress thing)
Brian Postlethwaite (Mar 24 2016 at 05:29):
something seriously wierd with that page, looks WAY different to DSTU2 and partial names in there like "inistrative-gender"
Brian Postlethwaite (Mar 24 2016 at 05:36):
Logged it at 9751
Grahame Grieve (Mar 24 2016 at 19:49):
yes I broke it doing the code system change
Eric Haas (Mar 26 2016 at 02:06):
@Grahame Grieve re 8527 what part needs revisiting? I still think this would be an extension on components.
Grahame Grieve (Mar 26 2016 at 05:02):
Can't look at it now, but I thought it would be a component now? Something about the preconditions looked like it had changed to me
Grahame Grieve (Mar 28 2016 at 06:05):
actually, isn't 8527 about SampledData
Eric Haas (Mar 28 2016 at 08:48):
yes
Grahame Grieve (Mar 28 2016 at 16:53):
so an extension on SampledData?
Eric Haas (Mar 28 2016 at 17:01):
No I don't think so, but I need to go back to look to refresh my memory. I think so can do multi-channel using o.components- one channel per component.
Last updated: Apr 12 2022 at 19:14 UTC
