NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
All, Previously on the list: Ø Version 4.4-beta1 is the first release to support the Fortran 2003 standard for calling C code from Fortran. It is used in this new release to replace the older C code wrappers. I would like to suggest that the ?old style? not be completely deprecated. We support a number of operating systems where we will not have Fortran 2003 language support. (Yes I am certainly aware of the GNU project, but that is NOT an option for producing commercial products anywhere but Linux.) I am happy to report we have successfully built netcdf-fortran-4.2 for x64 where NF_INTEGER is 64-bit. I have a set of proposed changes I would like to submit for review. Here is a short list of issues: [1] Modified cfortran.h for direct support of ?intelVisualFortran?. [2] Split EXTERNL into NC_EXTERNL and NF_EXTERNL to properly handle DLL import/export. [3] ncfortran.h: Expanded to support NF_INT_IS_C_LONGLONG, NF_INTEGER8, and _int64 cfortran macros. [4] fort-nc4.c: Used _int64 cfortran macros for the Fortran/C interface. [5] Added support for NF_INT_IS_C_LONGLONG everywhere to call the nc_...longlong variants. This requires NO changes to the C lib (with the exception of the shared ?config.h? and ?nfconfig.inc?). All changes hinge on the fact that ?Integer*8 == long long?. These changes should be easy to follow and completely safe, since they only come into play when nfconfig.inc defines NF_INT_IS_C_LONGLONG. Could the person handling the Fortran library please get in touch with me? I would like to submit these changes and ask they be included in future releases. Regards, Roy Dennington Semichem, Inc.
netcdfgroup
archives: