Stream: fmg
Topic: R4B + R5
Grahame Grieve (Dec 15 2021 at 18:51):
A discussion topic for todays FMG (which I won’t be on): I’m publishing frozen versions of FHIR for the connectathon - both R4B and R5. typically, we publish them with a date on them - the date any associated ballot closes. So it would be 2022JAN for this publication. Only, I’m publishing both. And neither is associated with a ballot. So what name for the folder they publish in? 2021DEC and 2021DECB ?
Grahame Grieve (Dec 15 2021 at 18:53):
An alternative approach we could consider is that since these ballots make the beginning of using version release tags, as previously discussed, these versions will be 4.3.0-DRAFT1 and 5.0.0-DRAFT1, so we could stop using dates and use the full version from now on
Rob Hausam (Dec 15 2021 at 18:56):
The alternative approach sounds pretty good to me.
Lloyd McKenzie (Dec 15 2021 at 20:09):
It'd be a nice test to see if the new naming approach blows up any tools with a release that's not super-critical.
Josh Mandel (Dec 15 2021 at 21:19):
Agree that in abstract, the alternative approach seems clearer
Lloyd McKenzie (Dec 15 2021 at 21:56):
FMG outcome: Preference is 4.3.0-SNAPSHOT1 and 5.0.0-SNAPSHOT1, though we can live with 4.3.0-SNAP1 and 5.0.0-SNAP1 if we're overly concerned about length. ('Draft' is a common name for 'for-comment' ballots and so we want to avoid confusion)
Josh Mandel (Dec 15 2021 at 22:16):
(I missed the capitalization in our discussion -- I don't know that I'd capitalize the names.)
Grahame Grieve (Dec 16 2021 at 17:29):
isn't semver practice to uppercase them?
Josh Mandel (Dec 16 2021 at 19:47):
No, not per the grammar and not per the examples in https://semver.org/
<pre-release identifier> ::= <alphanumeric identifier>
| <numeric identifier>
Examples: 1.0.0-alpha, 1.0.0-alpha.1, 1.0.0-0.3.7, 1.0.0-x.7.z.92, 1.0.0-x-y-z.–.
Grahame Grieve (Dec 16 2021 at 20:00):
lowercase it will be then
Grahame Grieve (Dec 22 2021 at 20:59):
referred item from MnM for our next meeting: the product pattern (http://build.fhir.org/product.html) is defined, but has no owner, nor any direction, and is not properly integrated as a pattern. MnM just met with an implementer who is feeling pain because of the lack of consistency (both in access and content) in this area
Lloyd McKenzie (Dec 22 2021 at 22:43):
In theory, it could be owned by FHIR-I workflow, as they own the Event/Request/Definition patterns and part of their scope is increasing consistency across resources. That said, it's not strictly workflow.
Grahame Grieve (Dec 22 2021 at 23:11):
FHIR-I will be invited to the meeting to consider it's future
Last updated: Apr 12 2022 at 19:14 UTC
