Stream: committers
Topic: Technical Corrections
Grahame Grieve (Jul 29 2019 at 20:40):
I am starting to work on forthcoming technical corrections for R2, R3, and R4. Here's my current task list for them:
Grahame Grieve (Jul 29 2019 at 20:40):
R2:
- New Package Id
- Fix licensing note on UB04 codes
Grahame Grieve (Jul 29 2019 at 20:40):
R3
- New Package Id
- Fix licensing note on UB04 codes
- GF#20335: fix invariant sdf-15 for logical models
- GF#20334: fix invariant sdf-8 for logical models
- GF#20336: fix invariant sdf-12 for logical models
- GF#20339: Fix Invariant sdf-20 on structureDefinition
- GF#20331: fix erroneous FHIRPaths
- GF#22942: Fix Condition con-4 invariant
- GF#22712: eld-10 should only apply to snapshot
- GF#20399: Correction to v3 schema dateTime regex
- GF#22808: "instant" should have a regex specified in types-profiles
Also:
- Non-technical corrections: GF#13865 : Apply these R4 changes as technical corrections to STU3 ?
Grahame Grieve (Jul 29 2019 at 20:42):
R4
- New Package Id
- Fix licensing note on UB04 codes
Definition Changes
- GF#20487: Parameters.parameter.valueMeta prescribed but not existing
- GF#18441: Canonicalization & normalization to single-space
Search Parameter Corrections:
- GF#20463 Mother's Maiden name search parameter
- GF#20470: SearchParameter declarations in R4 are problematic
- GF#22495: add modifier 'ge' to date searchparameter
Formal Definition fix-ups
- GF#22807: code systems look up
- GF#20493: Fix incorrect version appended by validator to terminology.fhir.org value sets
- GF#22681: Incorrect element ids in profiles-others
- GF#22670: ElementDefinition.type.code is invalid for Element.id (and thus all datatypes)
- GF#22493: Type profiles derived from Element have incorrect <type> for Element.id
- GF#20465: Element.id is inconsistent between snapshot and differential in FHIR 4.0.0's JSON definitions
- GF#19968: Schematron implementation of eld-19 technically incorrect
- GF#23008: Invariants in Condition and AllergyIntolerance broken
Fixed Invariants
- GF#20335: fix invariant sdf-15 for logical models
- GF#20334: fix invariant sdf-8 for logical models
- GF#22986: con-3 is wrong
- GF#20331: fix erroneous FHIRPaths
- GF#20320: Fix eld-4
- GF#22985: ait-1 and ait-2 are wrong
- GF#20319: Fix Observation constraint
- GF#22986: con-3 is wrong
- GF#22709: Bundle bd-10 wrong fhirpath
Broken Links:
- GF#20561: R4-candidate fhir.schema.shex.zip is missing
- GF#20393: FamilyMemberHistory.sex binding
- GF#20592: 404 Error
- GF#20409: non-resolving urls for stu3
Documentation fixes
- GF#20472: What version of JSON Schema does FHIR target?
- GF#20584: R4 FMM levels are incorrect
- GF#19969: Normative example value set with 0 members
- GF#20855: Missing 3 Substance Resources from Alpha and Categorized
- GF#20618: STU to R4 update
- GF#20199: STU3 or R4
Grahame Grieve (Jul 29 2019 at 20:42):
Please review
Grahame Grieve (Jul 29 2019 at 21:07):
Note that changing the package ids will force a a re-publication for every IG there has been (including those not published by HL7), and new versions of all tools. This is effectively a reset of the entire eco-system. I think I'm going to make the IG changes without formally republishing them (e.g. changing their stated version), but this will be the last time that happens
Lloyd McKenzie (Jul 29 2019 at 21:44):
Will the change to package ids matter for those who are still using versions of tools that use igpacks rather than the NMP packages?
Grahame Grieve (Jul 29 2019 at 21:45):
no idea
Eric Haas (Jul 30 2019 at 15:23):
What is the projected date for the Technical corrected version of R4?
Eric Haas (Jul 30 2019 at 15:24):
I assume all IGs will need to cut over to it
Grahame Grieve (Jul 30 2019 at 19:19):
while ballot is opened. All IGs, published or not, will need to cut over to it
Richard Ettema (Jul 31 2019 at 19:40):
For R3 the obs-7 invariant also needs to be corrected. This done for R4 under GF#14073
Michelle (Moseman) Miller (Aug 05 2019 at 19:49):
Reviewed the list. I'm checking with @Lloyd McKenzie to confirm, but I think GF#23008 might be a duplicate of GF#22986 and GF#22985.
Grahame Grieve (Aug 05 2019 at 20:08):
ok thanks.
Lloyd McKenzie (Aug 05 2019 at 22:35):
It is
Lloyd McKenzie (Aug 05 2019 at 22:35):
(a duplicate)
Peter Jordan (Aug 06 2019 at 00:56):
GF#23062
On selecting the R3 Conversions Tab of any resource, the Breadcrumb Trail below the Menu Bar ends with 'R2/R3 Conversions', rather than 'R3/R4 Conversions'.
Grahame Grieve (Aug 06 2019 at 01:07):
thanks. got it
Grahame Grieve (Aug 06 2019 at 06:48):
ok, well, I have R4 building successfully. Now I can start working on the R4 technical corrections..
Grahame Grieve (Aug 06 2019 at 06:49):
@Chris Grenz it would be great if you could review the structure definitions for extensions and profiles in the current build - I believe that they are all now correct by our new definition of correct
Richard Townley-O'Neill (Nov 04 2019 at 03:57):
@Grahame Grieve
Is there a list of the changes from 3.0.1 to 3.0.2? Where?
http://hl7.org/fhir/history.html has lots on the 4.0.1 change, but nothing on the STU3 changes.
Dusica Bojicic (Nov 06 2019 at 01:47):
@Richard Townley-O'Neill, see http://hl7.org/fhir/STU3/history.html#v3.0.2.
Chris Moesel (Nov 06 2019 at 15:34):
Apologies if I missed this elsewhere, but I'm surprised to find that the Publication History (http://hl7.org/fhir/directory.html) has no entries for 3.0.0, 3.0.1, or 4.0.0. Why list every ballot version but not the prior official versions (for which there are possibly running production systems)?
I think it would be important to have access to these prior versions because: (1) there are still systems out there built against 3.0.1, so I want to see what 3.0.1 is; (2) the change log says things like "Remove ElementDefinition Invariant eld-10" -- but there is no apparent way for me to actually look up what eld-10 was. I can't truly understand the changes without being able to compare to previous versions.
Chris Moesel (Nov 06 2019 at 15:34):
So... is this just an oversight or was it intentional?
Chris Moesel (Nov 06 2019 at 16:01):
(deleted)
Chris Moesel (Nov 06 2019 at 16:03):
Also, in the detailed version history for 4.0.1, the entry for
Fix up the definition of ElementDefinition.type.code for special elements (Element.id, Extension.url, [Primitive].value)
has a task link that points to 23967, which doesn't seem to be related. Is it intended to point to 22670?
Grahame Grieve (Nov 06 2019 at 21:35):
@Chris Moesel thanks for the task correction. I've made that. With regard to the past versions, our practice is that we don't want people to find the uncorrected content on the web. So the older versions are there for download as a zip should you wish to do comparison with regard to the technical correction
Chris Moesel (Nov 06 2019 at 21:37):
@Grahame Grieve -- I totally missed the zips. I see them now. I think that's a fine (and probably wise) compromise. Thank you!
Last updated: Apr 12 2022 at 19:14 UTC
