CCPP Framework Meeting Minutes 2021 05 11

Agenda

PR Use assumed-size arrays in CCPP, Fortran/metadata consistency fixes in CCPP (https://github.com/ufs-community/ufs-weather-model/pull/527) was merged on April 30
- A few fixed array dimensions were forgotten, Grant will fix some in a forthcoming ccpp-physics PR
- Using assumedd-size arrays helped to identify and fix a few bugs already (part of #527)
- We need to watch out for future PRs to follow the new rules/guidelines (sent to physics developers)
  - Order of arguments in subroutine must be same as in metadata (why?)
  - Use assumed-size array dimensions where possible
Next up on our transition plan is to add capgen to the UFS, get build system and code to work, ...

PRs for ccpp-physics and CCPPStandardNames are there:
- https://github.com/NCAR/ccpp-physics/pull/637 - some b4b issues with SCM to resolve, not tested with UFS yet
- https://github.com/ESCOMP/CCPPStandardNames/pull/13
  - Vertical dimension: vertical_dimension, vertical_dimension_plus_one`
  - Interface, layer: layer is the default, thus omit; if it is on interfaces then add interface identifier
Existing PRs cover about 2/3 of the existing standard names in the UFS, 1/3 is still missing
Still need to submit a set of proposed rules to the repository
Spreadsheet: https://meet.google.com/linkredirect?authuser=0&dest=https%3A%2F%2Fdocs.google.com%2Fspreadsheets%2Fd%2F1kWDC7_2sagm-PTP_tXaGyI3O3sJFK8FY9FElQbb6V3o%2Fedit%23gid%3D1778121391
Doc: https://docs.google.com/document/d/1AEtGEqFiXIL2iHUhpZtBadBvIhF4fR4Oru0XlzhTtAs/edit
When do we need to engage NCAR and others? Now?
- Review of current PR should help to settle on the changes, then involve the bigger CCPP community in a few weeks time?
- DTC deliverable is end of June 2021
- NCAR needs a stable list of names to give to the developers, by the end of the month?

CCPP Framework wiki