Stream: implementers
Topic: Endpoint.connectionType
Grahame Grieve (Nov 10 2016 at 22:11):
@Brian Postlethwaite Endpoint.connectionType is a coding that has no binding. That's a pretty unusual combination. And one that will make EndPoint hard to use. Why shoudn't connectionType be a code or a uri?
Grahame Grieve (Nov 10 2016 at 22:12):
what do we think consumers are going to do with end point, other than examine the connection type to see if they can use it. And how will they know how to use it if they don't know what the codes are for the kind of services they can consume?
Grahame Grieve (Nov 10 2016 at 22:12):
anyway, how many service types can we imagine? I would have thought 8-10:
- dicom variants (3)
- xds variants (2)
- direct/secure messaging
- FHIR end point
Grahame Grieve (Nov 10 2016 at 22:13):
- v2 protocol variant
Grahame Grieve (Nov 10 2016 at 22:13):
what else could we imagine?
Elliot Silver (Nov 10 2016 at 22:14):
Here's the current list: http://build.fhir.org/valueset-endpoint-connection-type.html
Grahame Grieve (Nov 10 2016 at 22:15):
oh, there is a problem in the value set somewhere that leads to a missing link
Grahame Grieve (Nov 10 2016 at 22:17):
but my list looks remarkably like the existing list - that's good huh
Elliot Silver (Nov 10 2016 at 22:18):
great minds thinking alike, and all :-)
Grahame Grieve (Nov 10 2016 at 22:19):
ok. so. I only have a publishing issue then
Brian Postlethwaite (Nov 10 2016 at 23:01):
So we're good here then? (have been under a rock for a few days)
The binding is extensible so that others can add stuff (this wasn't there in the Sept build, we decided on the final struture at the connectathon and reviewed with @Elliot Silver then.
Grahame Grieve (Nov 11 2016 at 00:50):
yes we're gooed except for the publishing issue but that's mine to investigate
Last updated: Apr 12 2022 at 19:14 UTC
