This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Michael, > The program name is CDFconvert (the source code can be found here > http://www.atmos.albany.edu/facstaff/rmctc/cdf_cvt/) and yes it does require > a Fortran compiler as most (if not all) of the code is written in F90. > > I guess this is the problem with using a less well known compiler. I'm afraid you're right. It's possible -- even likely -- that the calling-conventions of your Fortran compiler are identical to those of a supported compiler. The problem is determining which one. One crucial calling-convention deals with the passing of strings (CHARACTER* variables). Do you know how your Fortran compiler does it? Some compilers pass a pointer to the string and add an additional argument that specifies the length; other compilers pass a structure that encapsulates the string. You could also just try the various calling-conventions, one-at-a-time, to see if one works with "CDFconvert". The C macros are listed in the UDUNITS file "port/cfortran/cfortran.h". I would start with "f2cFortran". You might also try contacting the vendor of the Fortran compiler to ask if there calling-conventions are similar to the well-known ones listed in the previously-mentioned file. Good luck. Regards, Steve Emmerson Ticket Details =================== Ticket ID: GYX-412442 Department: Support UDUNITS Priority: Normal Status: Closed