-
Notifications
You must be signed in to change notification settings - Fork 64
CCPP Framework Meeting Minutes 2024 06 27
Courtney Peverley edited this page Jul 15, 2024
·
4 revisions
Attendees: Mike Kavulich,
CCPP Framework (issues, PRs, discussions)
-
- Waiting on https://github.com/ufs-community/ufs-weather-model/pull/2290
- There has been a big to-do about this PR, several rounds of test failures due to staging data etc.
- It seems like it's close to being merged, but it also seemed that way a week ago. Do we want to just pull the plug and create the develop branch and deal with merges/rebases later?
- Waiting on https://github.com/ufs-community/ufs-weather-model/pull/2290
-
Remove requirement that suite definition file names begin with "suite_" PR#569
- PRs are open for SCM and UFS weather model/FV3
- This one won't hold up creation of the develop branch
- PRs are open for SCM and UFS weather model/FV3
-
Implement new constituent property/field to track "above the top of model" values #573
Standard names (issues, PRs, discussions)
- Creating develop branch?
- Merge strategy?
- I prefer "squash-and-merge" due to the easier-to-follow tracking of history, easier cherry-picking between branches
- Merge strategy?
Mike K is in the process of updating the "Development Workflow" wiki page. It's not done yet, but you're welcome to spy on my progress and raise objections to anything you see there.
- Issue backlog cleaning
- Mike K hasn't had time to prepare anything, will try to find time to review old issues before next meeting.
CCPP Framework
- Constituent updates
- Test failures unrelated to framework stuff
- Grant checked in - didn’t stage data correctly; seg fault from physics
- Grant testing fix now
- Grant checked in - didn’t stage data correctly; seg fault from physics
- Jesse - where can we see status of tests?
- Discussion on PR
- If we can get it in by next week - CAM folks are ok with that
- Dustin - hope we get it in before the 4th
- Dom - removal of blocked data structures PR will be merged first
- Test failures unrelated to framework stuff
- Equivalent units
- Prebuild needs a bit more logic
- Register phase
- Not going into prebuild
- Grant - we should present the new features of capgen to the UFS developers
- Dustin - has tried to mention lots of the new stuff; developers are a little resistant to handing over responsibility to the framework
- Hopefully, in August, Dustin will have time to devote to integrating capgen into the SCM
- Dom - should focus on:
- Good & precise error handling
- Concise output of what framework is doing (right now we get either nothing or 100000 lines of debug output)
- Need to classify the messages we want to output when
- Dustin - has tried to mention lots of the new stuff; developers are a little resistant to handing over responsibility to the framework
- Above top of model constituent values
- Chemistry needs, for example, the amount of ozone above the model top for low-top models
- Dom - similar to LTP value in RRTMG
- Dustin - everything that needs extra layers added - computed beforehand, then passed into radiation (w/ standard names), then passed back out
- Jesse - can be combined with register phase
- Register phase can specify the above model top is needed for radiatively-active quantities
- For radiation, we will need more than one layer above the model top - something to keep in mind during design
- Cheryl - issue in standard names about above model top?
Standard Names
- Rename turbulent_kinetic_energy and change units
- Will keep name as is (no “specific”), but change the units
Discussion
- Develop branch
- Michael K working on development workflow draft
- Merge strategy?
- Michael K proposes “squash and merge”
- Everyone agrees
- We should configure the repo to default to squash
- Commit message?
- Michael W - would like consistency, clarity
- Use PR description - can configure this in github
- Michael K proposes “squash and merge”
- Dom - should not forbid regular merges
- Ex: would want to retain history when merging back into release branch
- Jesse - should update PR template/documentation to specify what we are doing