Hi Thomas,
Am 21.08.2018 um 20:38 schrieb Thomas Beale:
That would be the case if only the AQL spec were used for conformance
testing. But conformance also relies on the RM, ADL and other specs,
as you can see here
<https://www.openehr.org/releases/CNF/latest/docs/openehr_platform_conformance/openehr_platform_conformance.html>.
If you go to 6.2.2 and down to 'Directory Operations', you can see
conformance tests for Folders. The first step is to make sure all
systems implement just the basic structure the same way.
This is clear to me. However, if we are talking about vendor-neutral
platform ecosystems with lots of client applications sharing an openEHR
backend, you just need to have (just a stupid guess) ~98% conformance
(and AQL is in practice just as important to devs as the more
fundamental specs), otherwise it is getting too painful and expensive to
change the vendor. Even with small variances in the implementation, you
might just create a more friendly looking version of vendor lock-in. I
think this is obvious and it is likely I have missed your point.
The worst case right now would be that a query that mentioned some
FOLDER structure was run in Marand, DIPS, Code24, EtherCIS, EhrServer
etc, with different results. This is partly because we have not agreed
on how to use Folders (e.g. mark them in a certain way to represent an
episode etc), and partly because some systems don't use them at all.
Even systems that do have Folders may not use them in the same way (we
know this is true). So Folder is a slight black hole in openEHR which
we are actively working on in the SEC to tighten up, so that every
implementation uses them in the same way.
The Querying conformance table
<https://www.openehr.org/releases/CNF/latest/docs/openehr_platform_conformance/openehr_platform_conformance.html#_querying_component>
also needs to be augmented to include Queries that reference Folders.
A bit more work is needed to get all of this in place.
Jap, I agree that every AQL implementation that wants to meet the
conformance criteria needs to support folders in a defined way. From my
perspective, AQL queries on instances of EHR_ACCESS should also be
considered for the conformance criteria, as consent information might
also be relevant for querying in analytics scenarios (representation of
consent information is something the community might also need to take a
closer look at in the future).
- thomas
Birger
On 21/08/2018 18:28, Birger Haarbrandt wrote:
Hi Thomas,
from my perspective, this approach (by not being explicit about the
RM classes (and semantics) that need to be supported by the Contains
keyword) led to a situation in which two vendors (Marand and DIPS)
can claim that they have a valid implementation of openEHR but are
not compatible. I can't even tell if Marand (not support Folders at
all) or DIPS (using questionable overloading of the semantics) is
more right or wrong with their approaches on this matter...
Birger
--
Thomas Beale
Principal, Ars Semantica <http://www.arssemantica.com>
Consultant, ABD Project, Intermountain Healthcare
<https://intermountainhealthcare.org/>
Management Board, Specifications Program Lead, openEHR Foundation
<http://www.openehr.org>
Chartered IT Professional Fellow, BCS, British Computer Society
<http://www.bcs.org/category/6044>
Health IT blog <http://wolandscat.net/> | Culture blog
<http://wolandsothercat.net/> | The Objective Stance
<https://theobjectivestance.net/>
_______________________________________________
openEHR-technical mailing list
[email protected]
http://lists.openehr.org/mailman/listinfo/openehr-technical_lists.openehr.org
--
*Birger Haarbrandt, M. Sc.
Peter L. Reichertz Institut for Medical Informatics (PLRI)
Technical University Braunschweig and Hannover Medical School
Software Architect HiGHmed Project *
Tel: +49 176 640 94 640, Fax: +49 531/391-9502
[email protected]
www.plri.de
_______________________________________________
openEHR-technical mailing list
[email protected]
http://lists.openehr.org/mailman/listinfo/openehr-technical_lists.openehr.org
