			Introduction: wgrib2

Wgrib2 is a program for decoding and making inventories of GRIB2 files.  Using wgrib2
is similar to using wgrib except the number of options (flags) has increased to a huge
number and is expected to triple before things settle down.  Why has the number of
options increased?  The idea was to make wgrib2 (1) more configurable, (2)
easier to customize and (3) GRIB2 contains a huge amount of new options for
the metadata.  Maybe modularity and ease of development/maintaince could be
mentioned.

			Is it an option or a language?

If you look at unix programs like awk, sed and even find, you see that the
command line contains options that get "run" for every line or file in the
case of "find".  Wgrib2 takes that concept a little bit farther.  Each
option on the command line is now a call to a function and each
option-function gets called N+2 times.  First in an initialization phase,
once for each GRIB record (N times) and finally in a cleanup phase.  Now we
can have functions that can compute things like averages or manipulate files.
The only exception to this rule are "init" functions that only get called in 
the initialzation phase; for example, open an output file.  

		Order of calling option functions.

Initialization phase:
	each function is called in order (as on the commmand line) (mode=1)

Execution phase:
	each non-init function is called in order, mode=verbosity level >= 0

Cleanup phase:
	each non-init function is called in order, mode=-2

