Tiegcm development group:

We have several modifications to v1.92, so I would like to request
that the group test and evaluate this code in preparation for a release 
of v1.93.  Suggestions for additional changes or improvements are of
course also welcome. The attached file Release_1.93 summarizes the main
changes to the code.

Some of these modifications have not been included in $TGCMROOT/tiegcm1.92/src/modsrc, 
(mainly the new KP namelist parameter and the new SABER/TIDI lbc option).
This is to avoid unexpected and confusing changes for people who are using the
modified files in this directory. In fact, I would like to discuss the possibility
of eliminating the "official" modsrc directory in upcoming versions,
and using SVN instead for accessing pre-release developments of the code.

A copy of the attached proposed release notes is in ~foster/tgcm/tiegcm/trunk/doc.
Also in this directory, the file tiegcm1.92-trunk.diffs contains *all* the
gory details of code differences between the SVN trunk (revision 296), and the 
v1.92 fixed code.  This file also shows which files are identical, i.e., have 
not been changed.

I have saved SVN tag tiegcm1.92_r296 of the new code. To access the model
execute the following svn command:
  svn checkout $SVN/tiegcm/tags/tiegcm1.92_r296
Where the environment variable SVN is: 
  file:///home/tgcm/svn at hao, or 
  svn+ssh://arc.hao.ucar.edu/home/tgcm/svn on bluefire
(see ~foster/.cshrc at hao or bluefire for an example of setting SVN)

The svn checkout command should result in the appearance of directory 
tiegcm1.92_r296 in the working dir, which contains the usual src and scripts 
subdirectories.  To build and execute a short default run, copy the appropriate 
job script (linux or ibm) from tiegcm1.92_r296/scripts, and set the following:
  set model       = tiegcm1.92_r296
  set wrkdir      = .
  set tgcmroot    = .
Then either submit (ibm) or execute (linux) the job script as usual.
After the run, rename and modify the default input file for your own run,
set the input file name in the job script, and resubmit. I strongly recommend
that you do NOT use any old versions of job scripts or namelist input files.

Please note that it is now possible to build and execute the code on hao 64-bit 
Linux systems (e.g., arc), for either serial or parallel MPI runs. See instructions 
at the bottom of the release notes.

FYI, I am preparing a "flow chart" of potential model and solar input options
(Weimer, Heelis, GPI, data runs, hpower, ctpoten, etc.), including the new KP 
namelist read parameter. I'll send this out in a later email or present it at 
our next meeting.

Thanks,

--Ben