Feature/rtodling/nc diag rev1

[rtodling]

The following have been changed as minor updates to nc-diag generation:

  modified:   analyzer             - protect use of external sfc file; prep for handling pe files as directories 
   modified:   etc/gsi.rc.tmpl      - define plat for profl; prep for pe-files handling as dirs
   modified:   etc/gsi_fdda_1.rc.tmpl - prep for pe-files handling as dirs
   modified:   etc/gsi_fdda_2.rc.tmpl - prep for pe-files handling as dirs
   modified:   etc/gsi_fgat_1.rc.tmpl - prep for pe-files handling as dirs
   modified:   etc/gsi_fgat_2.rc.tmpl - prep for pe-files handling as dirs
   modified:   etc/gsi_sens.rc.tmpl - prep for pe-files handling as dirs
   modified:   etc/gsidiags.rc - prep for pe-files handling as dirs
   modified:   etc/obs.rc.tmpl - prep for pe-files handling as dirs
   modified:   gsidiags - prep for pe-files handling as dirs
   deleted:    nc_diag_read_mod.f90   - this does not belog here 
   deleted:    nc_diag_write_mod.f90  - this does not belog here
   modified:   setupdw.f90  - no nc4 file for pe w/o data (nobs=0)
   modified:   setupps.f90  - no nc4 file for pe w/o data (nobs=0)
   modified:   setuppw.f90  - no nc4 file for pe w/o data (nobs=0)
   modified:   setupq.f90   - no nc4 file for pe w/o data (nobs=0)
   modified:   setuprad.f90 - equalized bin/nc4 output within reason
   modified:   setuprw.f90  - no nc4 file for pe w/o data (nobs=0)
   modified:   setupspd.f90 - no nc4 file for pe w/o data (nobs=0)
   modified:   setupt.f90  - no nc4 file for pe w/o data (nobs=0); split file w/ defined dplat
   modified:   setupw.f90  - no nc4 file for pe w/o data (nobs=0); split file w/ defined dplat

Other changes in this branch are minor:

• add handle to tap on berror covariance
• add handle to have diag flies written in directories (as done at NCEP) - still to check if pays off.
• optional output of external sfc fields
• memory management in hybrid component

[rtodling]

Sorry - not sure what happened to the font! I tried adding comments to the change and I got the garbage!! Sorry - in any case, the changes are ok.

[mathomp4]

Sorry - not sure what happened to the font! I tried adding comments to the change and I got the garbage!! Sorry - in any case, the changes are ok.

Ah. You have a "#" character. If that's at the start of a line in Markdown, you get a Class 1 Heading. Big Font!

You can change it to a "-" or "*" to get a list (like you have at the end). Or you could surround it with triple-backticks to get a code block

[rtodling]

Thanks Matt - I will be observant of "#" characters in the future.

[rtodling]

I am closing this ticket as the ultimate changes to the NC4 files stuff are in upcoming update.

[rtodling]

Actually all changes are here.

[jing-guo]

After its conflicts with develop HEAD in active_channels.tbl by taking what are already in the develop HEAD, and with a green check "All checks have passed". It still shows "merging is blocked".

[jing-guo]

I add Meta for the next "1 approving review", if that is why the merge is blocked.

1. sometime we don't want automatic merging to take place, then it does.
2. sometime we do want automatic merging to take place, then it won't. Not have a clude.

[jing-guo]

Not have a clue. :-)

[rtodling]

We don't do auto-merge/update Jing. I don't think Meta is a person w/ approval powers in the case here - perhaps she should be at some point.

[jing-guo]

We don't do auto-merge/update Jing. I don't think Meta is a person w/ approval powers in the case here - perhaps she should be at some point.

But this page is showing the old reason "merging can be performed automatically with 1 ..." for why "Merging is blocked", while I do have some approval powers.

So taking your advice, I removed Meta for this purpose, and assign you as the next assignee, with the understanding that "the one made change, can not approve the change", since I did make some minor but additional changes to resolve conflicts.

Regarding "auto-merge", I still feel there are something in the process that I don't have a full control over the outcome. I will keep observing.

[rtodling]

The branch is now in sync w/ Meta's changes ... Jing can you approve it please?

[jing-guo]

There must be some step missing in this PR process. My later code-review or approval won't make a difference.

May be we should send to Matt for advice.