FHIR Chat · Must Support · IG creation

Stream: IG creation

Topic: Must Support

Grahame Grieve (Jan 29 2019 at 08:00):

http://www.healthintersections.com.au/?p=2908

John Moehrke (Jan 29 2019 at 15:01):

Excellent... Note that in IHE we have chosen to use Must-Support at least to indicate when the Interop Profile (IG) requests the sender populate the element if it has a value. This has historically been indicated with R2 or R3 (Required if Known). Further refinement is expected. IHE has found that R2 is needed because often a sending system will only populate mandatory elements, thus they need the R2 indication to push them to populate critical elements to best-case Interop.

Brian Postlethwaite (Jul 01 2019 at 13:34):

Those interested in a MustSupport only version of the StructureDefinition, I have a version of the IG Publisher that now produces this.
Comments on the visual, and labelling?
pasted image
This is the snapshot for the profile:
pasted image
I stripped out the bindings and examples from the must support view, as the examples are in the detailed display, and the must support bindings are available as a table to add after the structure tree.
If anyone wants this branch, I'll push it to github on my fork for folks to check it over.

Brian Postlethwaite (Jul 01 2019 at 13:37):

This is the commit on my fork with the changes...
https://github.com/brianpos/fhir-ig-publisher/commit/107fd966b27b0a3006c197a0a8998e2c33ba2203

Eric Haas (Jul 01 2019 at 17:27):

@Brian Postlethwaite can you explain what the benefit of this is over the existing diff view?

Brian Postlethwaite (Jul 01 2019 at 22:09):

That is the differential view at the bottom, no change. The top display is the must support tree.
It's a more consise display, only showing nodes marked with must support that are not constrained out.
So it focuses the display on what the author wants to display.
The removed and extra properties aren't shown.

Brian Postlethwaite (Jul 01 2019 at 22:12):

I'll add a screenshot of the full snapshot so you can see the difference.

Richard Townley-O'Neill (Jul 02 2019 at 01:35):

One reason: If a profile (B) adds must support to another profile which has must support (A), the difference view on profile B will not show the elements that are made must support in profile A.

Richard Townley-O'Neill (Jul 02 2019 at 01:35):

This new view will show them.

Brian Postlethwaite (Jul 02 2019 at 01:40):

Here's the complete set showing the differences between the 3 views.
pasted image
pasted image
pasted image

Brian Postlethwaite (Jul 02 2019 at 01:42):

one thing to note is that I removed all the red must support flags, as they look like errors in this view (as every property has them).

Richard Townley-O'Neill (Jul 02 2019 at 01:44):

Is it normal for 0..0 elements like statusHistory to not appear in the snapshot?

Richard Townley-O'Neill (Jul 02 2019 at 01:45):

I expected the must-support view to be a filtered view of the snapshot.
But it includes more.

Brian Postlethwaite (Jul 02 2019 at 01:46):

Yes, 0..0 elements are not in the snapshot, and haven't been before either.
It is a filtered view of the snapshot, that's how I generated it.
(cloning the snapshot list, then filtering/modifying the elements in the list before passing it to the render tree function)

Brian Postlethwaite (Jul 02 2019 at 01:49):

@Eric Haas , does that clarify things for you?

Richard Townley-O'Neill (Jul 02 2019 at 01:50):

Why do you want the 0..0 elements in the must support list?

Brian Postlethwaite (Jul 02 2019 at 01:56):

They aren't in there.

Richard Townley-O'Neill (Jul 02 2019 at 01:58):

Your second image contains implicitRules 0..0

Brian Postlethwaite (Jul 02 2019 at 01:58):

That second view is the current differntial view

Brian Postlethwaite (Jul 02 2019 at 01:59):

(and last is the current snapshot)

Richard Townley-O'Neill (Jul 02 2019 at 01:59):

Doh!

Richard Townley-O'Neill (Jul 02 2019 at 02:00):

:embarrassed:

Eric Haas (Jul 02 2019 at 02:03):

I guess I personally don't find it that useful (probably because I MS everything anyway). That being said I do reserve the right to change my mind. :-)

Lloyd McKenzie (Jul 02 2019 at 02:19):

I'm nervous about removing the red indicator. Consistency is a good thing (and it sort of re-enforces what the view is doing).

Brian Postlethwaite (Jul 02 2019 at 02:38):

On that display it's just offensive noise.
The idea of the display is to try and simplify the view and remove as much noise as possible. (with all the layers collapsed)

Brian Postlethwaite (Jul 02 2019 at 02:45):

I'd like to see profiles of the EOB with this view on, they would be MUCH nicer I'd suspect.

Richard Townley-O'Neill (Jul 02 2019 at 03:54):

Maybe some muting of the must support icon, analogous to unchanged items in the differential view, would be enough.

Richard Townley-O'Neill (Dec 02 2019 at 06:22):

What happened to expanded snapshot?

Grahame Grieve (Dec 02 2019 at 07:01):

?

Richard Townley-O'Neill (Dec 02 2019 at 23:33):

@Brian Postlethwaite worked on a view of structure definition hierarchies that showed must-support elements.
Is it usable? Does anything use it? How can it be used?

Oliver Egger (Jun 19 2020 at 08:28):

@Richard Townley-O'Neill looking for ways to show a MustSupport view I came across this thread.

IG Publisher creates two artifacts per profile: ./temp/pages/_includes/StructureDefinition-...-grid.xhtml and /temp/pages/_includes/StructureDefinition-...snapshot-by-mustsupport.xhml. The second is the one from @Brian Postlethwaite described above I think.

To use it, you need do define an IG template which includes the artifact. See a possible usage here where an additional tab Snapshot Table (Must Support) has been added. The ig refers to a special ig-template where the layout-profile.hml hast been overwritten to include the additional tab, as far as I see it cannot be directly parameterized in the current templates.

Lloyd McKenzie (Jun 19 2020 at 15:59):

@Jose Costa Teixeira Can we merge this into the base template?

Jose Costa Teixeira (Jun 19 2020 at 16:17):

we can.

Jose Costa Teixeira (Jun 19 2020 at 16:18):

do you want to merge the previous PR (adding history tab) so that I can test properly? This seems a small addition but best to be safe than to debug.

Jose Costa Teixeira (Jun 19 2020 at 16:23):

@Lloyd McKenzie this exists in the layout:

{% unless excludexml == 'y' %}
<!--      <li>
        <a href="#tabs-xml">Pseudo-XML</a>
      </li>-->
{% endunless %}
{% unless excludejson == 'y' %}
<!--      <li>
        <a href="#tabs-json">Pseudo-JSON</a>
      </li>-->
{% endunless %}
{% unless excludettl == 'y' %}
      <!--<li>
        <a href="#tabs-ttl">Pseudo-TTL</a>
      </li>-->
{% endunless %}

Jose Costa Teixeira (Jun 19 2020 at 16:23):

seems it's doing nothing, can I remove?

Jose Costa Teixeira (Jun 19 2020 at 16:24):

or is it needed?

Lloyd McKenzie (Jun 19 2020 at 16:30):

Check w/ Grahame to see if there's any intention to ever have these for profiles

Jose Costa Teixeira (Jun 19 2020 at 16:33):

Ill do that in a separate PR then, I thought these were leftovers

Jose Costa Teixeira (Jun 19 2020 at 16:34):

@Oliver Egger I don't know if I'm missing something but where is the generation of StructureDefinition-{{[id]}}-snapshot-by-mustsupport ?

Jose Costa Teixeira (Jun 19 2020 at 16:35):

the layout that you showed just creates a tab, but the content is not correct

Jose Costa Teixeira (Jun 19 2020 at 16:35):

image.png

Jose Costa Teixeira (Jun 19 2020 at 16:36):

image.png

Jose Costa Teixeira (Jun 19 2020 at 16:36):

should these MSs be in there? Or the fact that they are in a slice fools the logic?

Oliver Egger (Jun 19 2020 at 19:29):

the content is created by the ig publisher but i'm not familiar with the implementation if it should show childrens MustSupport, but the graphics you provided would look like it should :sunglasses:

Jose Costa Teixeira (Jun 19 2020 at 19:49):

Oliver Egger said:

the content is created by the ig publisher

Is there no end to the number of rabbits inside this hat ?

Grahame Grieve (Jun 19 2020 at 19:50):

Check w/ Grahame to see if there's any intention to ever have these for profiles

yes there is. I thought I had done that

Jose Costa Teixeira (Jun 19 2020 at 19:55):

I confirmed - it does work now, except for that issue.

1. https://github.com/HL7/ig-template-base/pull/84 ready to merge
2. https://github.com/HL7/fhir-ig-publisher/issues/115 created

Michaela Ziegler (Aug 21 2020 at 07:11):

I have two questions:
1.) Is there a possibility to display only the defined elements from the PatternCodeableConcept in the mustSupportView (here system, code, display) and not all elements from the element type?
2.) Which is best Practice? Must Support be defined on the element type or on the elements system and code? Or both? mustSupportView.png DifferentialTable.png

Grahame Grieve (Aug 27 2020 at 06:02):

well, I think this will happen with the next release. If it doesn't, let me know where I can reproduce this

Grahame Grieve (Aug 27 2020 at 06:02):

best practice is on the element type

Michaela Ziegler (Aug 28 2020 at 07:23):

Grahame Grieve said:

well, I think this will happen with the next release. If it doesn't, let me know where I can reproduce this

Unfortunately there has been no change in the view with the new publisher. In the Must-Support-View all sub-elements are still displayed. As an example here Composition.type
must-support.png

Grahame Grieve (Aug 31 2020 at 00:55):

ok fixed next release

Last updated: Apr 12 2022 at 19:14 UTC