Jonathan Gregory wrote:
Dear Steve et al.
We can distinguish udunits-2 as a standard for units strings, and udunits-2
as software. If we say that CF uses udunits-2 as its reference, programs in
Fortran can still use udunits-1 to process the units strings, in nearly all
cases, though obviously a ununits-2 Fortran interface would be preferable.
To address Andrew's point about "u" for 1e-6 and Nan's about PSU, we could
make CF-specific xml files available for use with udunits-2. We decided to
do this with udunits.dat, in order to add sverdrup as well as PSU, but didn't
get round to it.
I _think_ to be completely accurate, in order to specify "udunits-2
standard for units strings" you need to also specify which udunits.dat
database to use. Ideally this would be the default one, but as the
discussion shows, its not necessarily the same.
_______________________________________________
CF-metadata mailing list
[email protected]
http://mailman.cgd.ucar.edu/mailman/listinfo/cf-metadata
