[Ncep.list.wwatch3.discussion-group] WWIII Notes - Weds, Aug 15, 2018

[Jessica Meixner - NOAA Federal]

Hi Everyone,


Thanks to those of you who joined us.  It looks like we are still on track
for an October public release.  The status of the milestone projects (
https://vlab.ncep.noaa.gov/redmine/projects/emc-ww3/roadmap) are:

   -

   6.XX Namelist for remaining programs (no update, a few more programs to
   go)
   -

   6.XX Bit reproducibility: The NCEP operational system with the grid
   redesign no longer has bit reproducibility as an issue.  However, when
   having two grids of equal rank partitioning the MPI to run them is still an
   unresolved issue. In addition we still have several regression tests that
   are not bit-reproducible (See list of them here:
   https://vlab.ncep.noaa.gov/redmine/projects/emc-ww3/wiki/How_to_use_matrixcomp_to_compare_regtests_with_master#5-Look-at-results
   )
   -

   6.XX Memory management: no update
   -

   6.XX IC5: Update from Qingxiang: IC5 is ready to be merged back to
   master, ST6 updates soon after.
   -

   6.XX Directional Singularity at NPole: works in serial, having MPI issues
   -

   6.XX BT Output: Update from Qingxiang: work will start soon on this
   -

   6.XX GFDL FMS coupling (Waiting on Directional singularity)
   -

   6.XX Bug Fixes and Documentation
   -

      Ugdev2bugfixes just went back into the master.
      -

      Fabrice has bug fix for W3SBS1MD
      -

      Mickael has many bug fixes and small updates which will be merged
      back in September.
      -

      UKMet has some updates for SMC grid with netcdf which will be ready
      soon.
      -

      Stefan wants to add some output parameters (might not be ready for
      public release)
      -

      There are 50 open bug fixes and documentation issues
      <https://vlab.ncep.noaa.gov/redmine/projects/emc-ww3/issues?utf8=%E2%9C%93&set_filter=1&f%5B%5D=status_id&op%5Bstatus_id%5D=o&f%5B%5D=fixed_version_id&op%5Bfixed_version_id%5D=%3D&v%5Bfixed_version_id%5D%5B%5D=1380&f%5B%5D=&c%5B%5D=tracker&c%5B%5D=status&c%5B%5D=priority&c%5B%5D=subject&c%5B%5D=assigned_to&c%5B%5D=updated_on&group_by=&t%5B%5D=>.

      -

      You can see what open tickets you have by going to:
      https://vlab.ncep.noaa.gov/redmine/projects/emc-ww3/issues?query_id=259


Some other issues that were discussed is an overall potential problem of
uninitialized variables and arrays.  We might need to systematically
address this as a wider group, but please keep your eye out for them and we
can fix as we find them.  While some are never used before being set to a
value, it is possible this is part of the bit reproducibility issues we are
seeing for regression tests.

Another issue that was discussed is the manual.  In general, it’s time for
manual and documentation updates for the public release.  In addition,
there are several sections that could use attention and updates. Two in
particular:

   -

   The timestepping section is out of date and could use improvement.  We
   should give examples of time-stepped used in operations and expand on the
   discussion of why we might say “do as we say not as we do” (as it’s
   currently written).  NCEP technical note 259 (
   http://polar.ncep.noaa.gov/mmab/papers/tn259/mmab259.pdf) has equations
   to determine timesteps and that would be useful to include in the manual.
   https://vlab.ncep.noaa.gov/redmine/issues/52842
   -

   Another point that was brought up is to update the output parameter
   definitions, making sure they are consistent with what is actually
   calculated in the code.  In addition, possibly just presented in a easier
   to find & use way for new users. (Issues:
   https://vlab.ncep.noaa.gov/redmine/issues/54027,
   https://vlab.ncep.noaa.gov/redmine/issues/54013)


Lastly, there is a desire for being able to choose the output parameters
calculated for point output instead of being limited to the different table
options.  Currently, this is not being planned for the upcoming public
release.

Last, a few GIT reminders

   -

   Please do not put netcdf or any other binary file in the repo.  Also
   please try to keep large files in general out of the repo. If you have a
   large regression test file, we will add it to the ftp instead.
   -

   Please use remote branches instead of pushing directly to the master
   unless your update is super small.  It’s easier to run the regression tests
   and make updates on a remote branch than if you push directly to
   code-review.  It will also help with timing as many of you might all be
   pushing to the master at the same time.
   -

   Checklist for submitting to master:
   https://vlab.ncep.noaa.gov/redmine/projects/emc-ww3/wiki/Checklist_for_submitting_to_master



If you have any presentations for the group this fall please let me know.
Our next meeting will be Wednesday, September 19th.

Best,

Jessica


-- 
Jessica Meixner
NCEP/EMC
NOAA Center for Weather and Climate Prediction
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.lstsrv.ncep.noaa.gov/pipermail/ncep.list.wwatch3.discussion-group/attachments/20180816/0e36814e/attachment-0001.html