Discussion related to the organization of the ccpp-physics repo and scheme naming conventions

[grantfirl]

• How can we better organize the ccpp-physics repo?
• Do we need to adopt a scheme naming convention, and if so, what should it look like?

[dudhia]

Organized directories (separated physics, separated host-specific interstitials) precludes the need for better naming.

[dudhia]

• With better organization we may also be able to separate out host-specific standard names and physics-specific standard names (internal flags). I don't like that the standard names have large GFS-specific sections that also include namelist options. They are not really "standard" in the sense of the host-agnostic ones or by being specific to a physics option.
• Also I see 3 flags in the state variables section and there should be a rule against that. The rest of the flags are in GFS sections.
• In summary, there is a lower tier of "standard" name that is either host or physics-option specific. The problem with these is they could multiply in number with hosts and physics options requiring constant additions to standard names lists. Even only allowing host-agnostic standard names in the host-agnostic directories doesn't address the physics-specific namelist issue but makes a case that those should be separated from host sections.