Stream: implementers
Topic: Genomics resource/profiles GitHub repo
Gaston Fiore (May 13 2016 at 22:44):
The way the FHIR Clinical Genomics Subgroup works means that I cannot commit to the FHIR SVN repo until changes are approved through majority voting. At the same time, people have been asking for a way to see logs of weekly or periodic changes, the latest code, etc. As such, I'd like to create a public GitHub repo with the genomics resource/profiles and work there on a continuous basis. Then, after the couple of times per year that everything is subjected to voting, I'll be able to commit the approved code to the FHIR SVN repo.
My question is the following: do I need to include any special statements, licenses, or the like, in the public GitHub repo that will contain these "unofficial" FHIR genomics resource/profiles?
Do you have any other suggestions or comments? Thanks a lot!
Grahame Grieve (May 15 2016 at 08:53):
If you use FHIR in the name of the repo, you'll need to ask for trademark approval as soon as the trademark system is up (this week, hopefully). Permission will be granted for this use.
Grahame Grieve (May 15 2016 at 08:54):
But I recommend to the committee nut to
Grahame Grieve (May 15 2016 at 08:54):
Not to get so out of sync with the current version
Lloyd McKenzie (May 16 2016 at 02:24):
Most work groups make changes to resources on the basis of tracker items that are approved by the work group. Such approvals are normally done as part of the work groups' weekly or bi-weekly calls. Occasionally, changes might be applied prior to approval to give the WG an idea of what the outcome of the change will be. What's the reason for diverging from that typical process for Clinical Genomics?
Gaston Fiore (May 16 2016 at 13:41):
Thanks @Grahame Grieve and @Lloyd McKenzie for the background on how things work usually at WGs. I'm brand new to FHIR and the CG WG and thus I don't have answers as to why the CG WG works differently from what you described. I'll let @ Bob Milius and/or @Gil Alterovitz explain this. Thanks a lot for your input!
Bob Milius (May 16 2016 at 15:24):
CG hasn't been using the tracker to manage changes, but I'm thinking that we should. So, to submit a change request, I would click on the "Propose a change" link on the bottom of the resource/profile page? Is that how it would be done? Typically in the past, our work group would have a discussion on issues/proposed changes/use cases, and Gil's group would propose a way to address the issue and implement it on his server which doesn't require a WG vote. It would be a straw man approach. Once the WG agrees with his changes (not always easy) and votes it up, it gets put into the current build. Maybe we should go straight to the current build and us the tracker to keep us on track? @Gil Alterovitz
Gil Alterovitz (May 16 2016 at 16:24):
When we set up the FHIR subgroup, we decided as a group to go with the approach of having the subgroup look through different models- and then the subgroup would recommend a solution for the overall main group to evaluate and vote on. Since there were often a number of interlinked issues in the modeling, with need look at different options/examples quickly, we did that in exploratory approach on the calls. We can certainly revisit that as we get toward non-draft version that won't have as many inter-related changes/models over time. I will bring this up at future subgroup meeting.
Grahame Grieve (May 16 2016 at 17:33):
There is a maturity aspect to this. and as more involvement comes, the cost of a sub-group agility grows
Last updated: Apr 12 2022 at 19:14 UTC
