FHIR Chat · Extensions · terminology / utg

Stream: terminology / utg

Topic: Extensions

Ted Klein (Apr 04 2019 at 11:46):

@Grahame Grieve I think the extension is left over from the first pass, or even from your very first original work a year ago, and has not been touched as this is the first time it is flagged. Best to note it in the Resource Review document (Google Doc) which lists all of the issues we run into as we look at the resources. It absolutely should be fixed. The other item - mismatched codes for workgroups between V3 and FHIR - was supposed to have been addressed, I need to follow up on why it is still there. Probably a bug that was only partially fixed.

Ted Klein (Apr 04 2019 at 11:47):

wtf, zulip did not process the reply properly, I must have some setting wrong here...sigh...

Grahame Grieve (Apr 04 2019 at 12:35):

well, can you follow up on it here?

Grahame Grieve (Apr 04 2019 at 12:36):

as in, invite the apelon guys doing there work here so we all see the discussion?

Ted Klein (Apr 04 2019 at 13:04):

yes will do that. Better to have discussions here, but lists of found bugs to be fixed and tracking their completion is in a separate document which the PM is supposed to be staying on top of.

Carol Macumber (Apr 04 2019 at 14:41):

Adding the erroneous URL (http://healthintersections.com.au/fhir/StructureDefinition/valueset-generate) into the Google Doc being used to track all extension work. Google Doc located here: https://docs.google.com/document/d/1Ym1MElmb5tpQXugP7UMM_R9ahY9NsbYKZLZ-7itGCYc/edit#heading=h.qinruk9vblj4

Carol Macumber (Apr 04 2019 at 14:49):

Mismatched codes for workgroups between V3 and FHIR is listed as work item #14 in the overall task list (located here: https://docs.google.com/document/d/16ApgcoESRt86FXhP6dgVz8iTQNB9aXLL_v6fJtfqIxY/edit) @Grahame Grieve we can prioritize this one if it's impeding your immediate progress?

Grahame Grieve (Apr 04 2019 at 20:26):

it's not the mismatched codes causing me problems, it's the use of a different extension- can we have just one?

Grahame Grieve (Apr 04 2019 at 21:50):

looking at this:

    <property>
      <extension url="http://hl7.org/fhir/StructureDefinition/resource-concept-binding-strength">
        <valueCode value="CNE"/>
      </extension>
      <code value="contextBindingUV-valueSet"/>
      <valueString value="2.16.840.1.113883.1.11.155"/>
    </property>

Grahame Grieve (Apr 04 2019 at 21:50):

we need a way to tell the UI to show this extension when it doesn't exist. I propose that we change the definition:

Grahame Grieve (Apr 04 2019 at 21:51):

  <property>
    <extension url="http://hl7.org/fhir/StructureDefinition/codesystem-property-extension">
      <valueUrl value="http://hl7.org/fhir/StructureDefinition/resource-concept-binding-strength"/>
    </extension>
    <code value="contextBindingUV-valueSet"/>
    <uri value="http://terminology.hl7.org/CodeSystem/ConceptDomain/"/>
    <type value="string"/>
  </property>

Grahame Grieve (Apr 04 2019 at 23:23):

I implemented that - the binding strength option will only appear in the UI if the property is so marked

Grahame Grieve (Apr 04 2019 at 23:24):

what does valueset-generate mean?

Grahame Grieve (Apr 05 2019 at 00:46):

ok I did a new release. It works... mostly...

Grahame Grieve (Apr 05 2019 at 00:46):

at least as most as I can sort given the questions above. but testing is good

Ted Klein (Apr 05 2019 at 14:07):

Hmmm...that looks like some kind of error, we should have only one extension, I will look into what is going on with the two different ones. Agree we should have only one. I will also research valueset-generate, I suspect it is a flag for RoseTree and the balloting tool to generate an expansion, not sure. Will post here once I understand what it actually is. "...did a new release". ?? The Editor? IGPublisher? Not sure what exactly you are working on.

Ted Klein (Apr 05 2019 at 14:22):

ok I need to look more carefully = the issue with this new proposed extension (it is in the list of extensions that must be added) is that concept domains are modeled as concepts in the conceptDomains.xml code system. A binding on a concept domain is 1 realm code to 1..n value sets, whah relationship having a coding strength. So it is basically a repeating tripartite property on a concept in that code system. There are a couple different ways of doing it, we were not sure (and still not) on the Brest way to do this wrt you tools and the FHIR infrastructure.

Ted Klein (Apr 05 2019 at 14:23):

sorry for the typos

Carol Macumber (Apr 05 2019 at 18:43):

Adding the erroneous URL (http://healthintersections.com.au/fhir/StructureDefinition/valueset-generate) into the Google Doc being used to track all extension work. Google Doc located here: https://docs.google.com/document/d/1Ym1MElmb5tpQXugP7UMM_R9ahY9NsbYKZLZ-7itGCYc/edit#heading=h.qinruk9vblj4

@Grahame Grieve We've (re)confirmed that valueset-generate is not needed as it's internal to Frank's V2 generation operations.

Grahame Grieve (Apr 05 2019 at 20:02):

ok I've removed the generate support

Ted Klein (Jan 06 2020 at 19:43):

Need to deal with the many extensions that do not have valid URI values in the UTG code system resources - because we do not yet have the formal FHIR definitions for the UTG extensions documented in the Google doc so do not yet have URIs for them to put into the resources. How can we move forward with this item before the Sydney WGM?

Grahame Grieve (Jan 09 2020 at 21:41):

where are they going to be defined? in UTG itself?

Grahame Grieve (Jan 14 2020 at 20:26):

I just got sent this in an email:

Listed below are the open high priority tasks we have for Grahame:

• Display of extended UTG properties on codes on rendered pages
• Fix Urls on the extensions
• Ability to create/edit Naming Systems in UTG Editor
• FHIR R4 toolkit bugs fixed (works with R3 now)
• Apple Mac version of the UTG Editor and Vocabulary Server to 64 bit for new OSX (Catalina)
• Fix About -> Help section bug on Vocabulary Server
• Add ability to view diffs to IG Publisher
• Possibly support implementation of publishing

Grahame Grieve (Jan 14 2020 at 20:27):

extended UTG properties... we're still discussing the extension URLs here, but I have no outstanding work

Grahame Grieve (Jan 14 2020 at 20:30):

naming systems in UTG editor: done and ready for testing

Grahame Grieve (Jan 14 2020 at 20:30):

R4 toolkit bugs... I'm not aware of any open bugs

Grahame Grieve (Jan 14 2020 at 20:31):

64bit... I have them compiling, will release ASAP

Grahame Grieve (Jan 14 2020 at 20:32):

I don't know what the about/help thing is

Grahame Grieve (Jan 14 2020 at 20:33):

I thought we decided not to do the whole diffs thing?

Grahame Grieve (Jan 14 2020 at 20:33):

I don't know what

Grahame Grieve (Jan 14 2020 at 20:33):

support implementation of publishing

Grahame Grieve (Jan 14 2020 at 20:33):

means

Grahame Grieve (Jan 14 2020 at 20:33):

@Ted Klein

Ted Klein (Jan 14 2020 at 22:49):

OK, lemme see if I can clarify sone of this.

Ted Klein (Jan 14 2020 at 22:51):

1. The current resources have the URLs that you and I agreed upon, and you said that rather than us creating the code system of the extended properties you would do so. If you are not going to do it let me know. The URLs look like: "http://terminology.hl7.org/CodeSystem/utg-concept-properties#v2-version-introduced". If you wish it to be something different please let me know and we'll fix it toute de suite.

Ted Klein (Jan 14 2020 at 22:53):

2. The URLs of the extensions are still dangling - they are not valid in many of the resources. You had said last year that you would create the extensions as part of putting together the IG. I think this fell off the table as work to do. Let me know if we need to further discuss, but we are waiting on you to create them and tell us what the URIs need to be in the resources that reference them.

Ted Klein (Jan 14 2020 at 22:53):

3. Super, we'll test editing Naming Systems this week.

Ted Klein (Jan 14 2020 at 22:54):

4. I'll ask Jess to send you screen shots of the bug when you select R4 instead of R3 when launching the toolkit.

Ted Klein (Jan 14 2020 at 22:54):

5. Fantastic, I think Sheila has Catalina and can check this.

Ted Klein (Jan 14 2020 at 22:55):

6. I'll ask Jess to send you screenshots again of the error that comes up when you do a Help->About

Ted Klein (Jan 14 2020 at 22:56):

7. Yes we did decide not to do this, but folks are asking again. We will explore this further in the Sydney WGM, but to do anything on it before we figure out what we want to do.

Ted Klein (Jan 14 2020 at 22:58):

8. You told me at the last WGM that you need to put some mods on the publishing scripts to do things like launch Lloyd's coremif generator, something about specific uRLs to suit the release at which are supposed to be where the redirects go to, etc. We may need to further discuss to identify what exactly needs doing. I don't know the details as I am not completely familiar with the publishing operations beyond what is documented in the Confluence pages.

Grahame Grieve (Jan 15 2020 at 00:19):

ok so, Ill go ahead and defein the code system in UTG based on the discussion we had earlier

Grahame Grieve (Jan 15 2020 at 00:20):

help about does nothing...

Jessica Bota (Jan 15 2020 at 14:13):

Very cool about the Naming Systems being editable! I sent over the info for the two items I was tagged in via email.

Jessica Bota (Jan 16 2020 at 14:05):

@Grahame Grieve , I tried testing the editing on Naming Systems and got this error: Unsupported resource type: NamingSystem. I also do not see a way to create a new Naming System. I downloaded the latest toolkit from http://www.healthintersections.com.au/FhirServer/.

Grahame Grieve (Jan 19 2020 at 20:27):

what version are you running> (help...about)

Ted Klein (Jan 19 2020 at 22:04):

I am not sure which version Jess is running, but the latest one on your tools download page is labeled 0.0.59; I just now downloaded it from that page (http://www.healthintersections.com.au/FhirServer/) as I had the FHIRToolkit 0.0.56 installed (which shows the Help->About just fine, but I am on OSX), and did an install of the latest Mac version from the download link on that page. Entertaininly, the Help->About works...but shows version 0.0.58; So perhaps the OSX version of the latest 0.0.59 is not linked for download there, and/or perhaps only the Windows version of 0.0.59 is the one that does not show it. It is Sunday evening here, so Jess will respond when she gets in tomorrow morning I would imagine.

Ted Klein (Jan 20 2020 at 09:46):

@Grahame Grieve please let me know when we can go over these and other items. I think there are several small things we have verbally discussed but are not nailed down, and I really want to get on top of them prior to the meetings in Sydney in 2 weeks.

Jessica Bota (Jan 21 2020 at 16:34):

@Grahame Grieve To clarify, the Help -> About issue is in the Vocabulary server, not the Editor. The Vocab server hasn't been updated since 2017, at least not here: http://www.healthintersections.com.au/FhirServer/. That says it is v0.0.6.

Jessica Bota (Oct 22 2020 at 18:37):

@Grahame Grieve Per the HTA, I am creating a number of new Naming Systems. We are currently representing title, url, and version in extensions like:
<extension url="http://hl7.org/fhir/tools/StructureDefinition/extension-title">
<valueString value="American Type Culture Collection"/>
</extension>
<extension
url="http://hl7.org/fhir/5.0/StructureDefinition/extension-NamingSystem.url">
<valueUri value="http://terminology.hl7.org/NamingSystem/atc"/>
</extension>
<extension
url="http://terminology.hl7.org/StructureDefinition/ext-namingsystem-version">
<valueString value="2.0.1"/>
</extension>

Ted mentioned that you may have added these elements in the base. Are we able to use <title>, <url>, and <version> now as we do for code systems instead of the extensions above? If so, will they render on the build pages?

Grahame Grieve (Oct 27 2020 at 00:21):

I believe they should all render correctly

Ted Klein (Jun 15 2021 at 02:28):

However, 600 errors were recently introduced, all the same. They are in the form: NamingSystem/ethernet-address-identifier: NamingSystem.extension[1][url='http://hl7.org/fhir/5.0/StructureDefinition/extension-NamingSystem.url'] error Extension url 'http://hl7.org/fhir/5.0/StructureDefinition/extension-NamingSystem.url' is not valid (invalid Version '5.0')
NamingSystem/ethernet-address-identifier: NamingSystem.extension[1] error The extension http://hl7.org/fhir/5.0/StructureDefinition/extension-NamingSystem.url is unknown, and not allowed here

Lloyd McKenzie (Jun 24 2021 at 01:40):

I suspect that to make everything completely happy, what the validator wants to see is http://hl7.org/fhir/4.6/.... @Grahame Grieve Is this something we should change in the validator or in the UTG source? (Keeping it at 5.0 would be nice because there'll be transforms for 5.0 stuff, but little chance of transforms for 4.6...)

Grahame Grieve (Jun 24 2021 at 01:42):

no it's supposed to 5.0. It was before, and I haven't had time to look

Last updated: Apr 12 2022 at 19:14 UTC