-------------
- Added many new helper tools in the Yellowstone testing directory.
- Perform a sort of the time-series variables by size before partitioning
them across processors. Since the partition strides through the list of
variables names, by sorting the variables by size first, we do a reasonable
job of selecting as many small variables as large on every processor
- A few minor typo bugfixes.
- Added the ability to output all metadata to a "once" file. This metadata
includes ALL variables that are NOT the time-series variables being written
to file. Hence, this can produce time-series files that are not entirely
self-describing.
- Fixed a bug in which a job hangs when using more ranks than variables
- Switched to using the identified 'unlimited' dimension name instead of
explicitly using the string 'time'.
- Added the ability to process time-slice files with multiple time steps
in each slice file.
- Added new plotting utility and module. Some changes to the getsizes
utility in the Yellowstone testing area. Some more tweaks here, too.
- Updated the PyReshaper to use the newly created PyTools/trunk. (Will
modify the externals to use a tag once the new PyTools has been tested and
verified).
- Renamed any ATM/LND 0.1 degree tests in the tests/yellowstone directory to
0.25 degree, which they actually are. Similarly, any 0.25 degree OCN/ICE
tests were named 0.1 degree.
- Updated the Specifier and Reshaper Doxygen comments to use Doxypypy
formatting.