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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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.
Roberto, The results you sent from config.log make it appear as if the configure script was edited, interrupted, or killed before it could finish running. In particular, right after it fails to find a working Fortran-90 compiler because of a license manager problem, a large section of the configure script is skipped that is supposed to determine the names and paths of basic development tools needed to build netCDF, such as the "ar", "nm", and "grep" utilities. As a result, the shell variables defining these utilities are all defined as empty strings: AR='' AS='' CPP='' EGREP='' FGREP='' GREP='' LD='' NM='' SED='' STRIP='' and that prevents the generation of Makefiles. It's possible you could get beyond the point where the license manager problems occur by specifying not to create a Fortran-90 interface by giving the configure script the argument "--disable-f90", as in: $ ./configure --disable-f90 If that doesn't work, you might instead try to correct the license manager problem with the pgf90 compiler or try to build with a gcc-based Fortran compiler such as gfortran. I have tried to duplicate the error you are seeing on our similar Linux platform, but the configure script runs to completion here and there is no such problem. Our expert on configure script problems is on vacation for two more weeks, so if the above suggestions don't help, maybe he will have a better suggestion when he returns. I'm CC:ing him on this reply. --Russ Russ Rew UCAR Unidata Program address@hidden http://www.unidata.ucar.edu Ticket Details =================== Ticket ID: TZL-450905 Department: Support netCDF Priority: Normal Status: Closed