FHIR Chat · FHIR vs. Opendata · implementers

Stream: implementers

Topic: FHIR vs. Opendata


view this post on Zulip Simone Heckmann (Mar 17 2016 at 08:33):

I have been (repeatedly) asked if & how FHIR relates to opendata, whether the opendata structures have been considered as basics for fhir and if so, on what grounds they have been dismissed. Can anyone give me some information about that?

view this post on Zulip Ewout Kramer (Mar 17 2016 at 10:02):

Hi Simone, I have had that question before, and now try to look up "the spec" for open data. That's pretty hard, since if you google it, you'll get hits about "open data". Do you have a link?

view this post on Zulip Simone Heckmann (Mar 17 2016 at 10:10):

I think it's this: https://www.oasis-open.org/committees/tc_home.php?wg_abbrev=odata

view this post on Zulip Grahame Grieve (Mar 17 2016 at 10:49):

Simone - I've looked at odata several times

view this post on Zulip Grahame Grieve (Mar 17 2016 at 10:50):

it's certainly influenced us - for instance, the _filter parameter is directly based on the odata one

view this post on Zulip Grahame Grieve (Mar 17 2016 at 10:50):

but the data model hasn't been useful in my view - it's too orientated towards square things. I'm happy to hear from people where they think we should have been informed and haven't been, and maybe that will lead to useful changes

view this post on Zulip Grahame Grieve (Mar 17 2016 at 10:51):

I did have the intent, at one point, to define a way to query against fhir resources, and get back an odata data set, but there was never very much interest in that, and underneath that problem is the need for a full ETL to square-ize things'

view this post on Zulip Simone Heckmann (Mar 17 2016 at 10:57):

Grahame, thanks for the insight. I will forward your request to the asker.

view this post on Zulip Ewout Kramer (Mar 17 2016 at 11:56):

Hold on, that's OData, a standard originating from Microsoft. That has no content models, and is much based on a relational model.

view this post on Zulip Ewout Kramer (Mar 17 2016 at 11:57):

And indeed, I have looked extensively at the spec (being a .NET guy), it's just not a good fit for our deeply nested resources.

view this post on Zulip Ewout Kramer (Mar 17 2016 at 11:57):

I thought maybe there was something really called "open data" standard. Which would be the same as calling your restaurant "The Restaurant", but hey....

view this post on Zulip Jason Walonoski (Mar 17 2016 at 12:02):

I've seen so many unrelated Italian dining places here in the US called "La Trattoria" it isn't even funny... which more or less means "The Restaurant"

view this post on Zulip Ewout Kramer (Mar 17 2016 at 12:21):

Yes, we actually have a restaurant called "the restaurant" in Amsterdam. These people obviously did not think about SEO....

view this post on Zulip Marten Smits (Mar 17 2016 at 12:24):

We also have one called "worst" (which means sausage in Dutch), also not a very good idea...

view this post on Zulip John Moehrke (Mar 17 2016 at 12:26):

Yet "Someplace Else" is a good one... too bad it burned down, here in Milwaukee

view this post on Zulip Eric Haas (Mar 17 2016 at 15:40):

@GG "it's too orientated towards square things. " ?? Bricks, ice cubes?

view this post on Zulip Grahame Grieve (Mar 17 2016 at 18:15):

tables with rows and columns


Last updated: Apr 12 2022 at 19:14 UTC