I recently wanted to use ncdiff on two files that had a
large number of variables in common (approx 200) and a
few that were not (approx 20).
The command failed (v3.0.1) because of the variables
not in common.
What I wanted was for ncdiff to operate on the
variables that were in common, and ignore the rest.
This is what I intuitively expected it to do.
In order to work around this problem I had to use ncks
to create new input files without the offending
variables, and then use ncdiff on the new files. It was
tedious to find and exclude all the variables not in
I therefore suggest ncdiff automatically ignore
variables not in common (except perhaps for a warning
message). If that is inadvisable for some reason, then
I suggest a new flag be created to provide this