[UDUNITS #RTT-423050]: netcdf and udunits software for NOAA-CIRES data sets

Dear Hans-Joachim Pistorius,

> I want to use your enclosed sample programm to read "netCDF" data files
> from the NOAA-CIRES Climate Diagnostics Center on a computer with the
> "AIX 1.5" system. I copied your files "netcdf.tar.gz" and
> "udunits-1.12.9.tar.gz".
> 
> I used "udunits-1.12.9" instead of the "UDUNITS-2"-Version because I
> read, that the actual version (UDUNITS-2) does not work in a
> FORTRAN-Program.

You are correct: the UDUNITS-2 package does not have a Fortran interface.

> I unzipped and unpacked the datasets and installed
> them as discribed in your instructions. In "udunits-1.12.9" I dropped
> the "PERL"-Extensions because I have not the rights to adress
> system-datasets and I don't need them. The installation endet
> successfull without mistake.
> 
> Then I tried to compile the above metioned sample program. My compiler
> (f77) didn't like the string '' (without blank in between  the '-signs)
> in the program. I inserted a blank. Besides this the compiler marked the lines
> 
> "#define  UD_POINTER    integer*4' and
> "UD_POINTER utmake"
> 
> in the include-dataset "udunits.inc" as mistakes. Because all other
> statements with "UD_POINTER" were commented out, I also commented the
> two mentioned lines out.

The Fortran include-file "udunits.inc" is designed to be included by the 
preprocessor utility before the resulting code is compiled by the Fortran 
compiler.  The xlf(1) Fortran compiler will do this automatically if the file 
that it is processing ends in a ".F" extension.  I suspect, therefore, that 
your file doesn't end in a ".F" extension.

Rename your file to have a ".F" extension and re-activate the lines in the file 
"udunits.inc" that you commented-out.

> Then my program was compiled correctly, but
> the loader didn't find the programs "utopen", "utmake",  "utdec",
> "uttime", "utcaltime" and "utfree" (e.g.: "ld:  0711-317 ERROR:
> Undefined sybol:
> .utopen").
> 
> Indeed I could not find such programs in the library "libudunits.a".
> Can you perhaps explain the reasons of my problems?
> 
> The problem, I'm working on,  is a distribution of wind-directions and
> speeds in Afghanistan as a function of the month and the time of the
> day.  Because we have not enough data, my idea was, to use the
> reanalysis of the u- and v- windcomponents neaer surface of NCEP-NCAR.
> I'm employed by the "Budeswehr Geoinformation Office" and we assist our
> military forces with geographical, meteorological and climalogical
> informations.

Sounds important.  Please contact us if you have any problems.

> Your answer is very much appreciated
> 
> Sincerely yours
> 
> Hans-Joachim Pistorius

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: RTT-423050
Department: Support UDUNITS
Priority: Urgent
Status: Closed