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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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.
Javier, >Date: Tue, 20 Oct 1998 13:48:27 +0100 (WET DST) >From: Javier Maldonado Miranda <address@hidden> >Organization: Instituto Andaluz de Ciencias de la Tierra >To: address@hidden >Subject: Re: 19981014: NetCDF Spanish installation problems >Keywords: 199810201147.FAA27173 In the above message, you wrote: > I'm trying to install GMT3.0 (do you know it?). We've heard of it. > This program needs your program netcdf. But I have found that when > I try to compile this program with your libraries it gives some > problems. GMT3.0 has some years and I don't know if your librarie > netcdf.h is a new one. I've tried to get some help from the people > that designed GMT, but I haven't had any answer yet. Maybe you can > help me. I'm going to send you a file that shows the problems that > appear at compilation time (comp_problems) and the problematic files. > > > Thank you again in advance. > > > Javier. > > P.S: by the way, can I remove the directory from where I installed your > program?. I'm not sure what you're asking. You may remove the "src/" subdirectory after successfully installing the netCDF package. Is that what you're asking? > Javier Maldonado Miranda > > Instituto Andaluz de Ciencias de la Tierra > Facultad de Ciencias > Avda. Fuentenueva s/n > 18022 Granada > Spain > > tlf: + 34 910 50 47 10 > FAX: + 34 958 24 33 84 > e-mail: address@hidden > > ********************************************************************* > > /opt/SUNWspro/bin/cc -O -I/usr/local/netcdf-2.4.2/include -DLIBDIR=3D\"/usr= > /local/gmt/lib\" -DSI -c pslib.c > ar cvr libpsl.a pslib.o > r - pslib.o > ar: writing libpsl.a > ranlib libpsl.a > /opt/SUNWspro/bin/cc -c -O -I/usr/local/netcdf-2.4.2/include -DLIBDIR=3D\"/= > usr/local/gmt/lib\" -DSI gmt_init.c > /opt/SUNWspro/bin/cc -c -O -I/usr/local/netcdf-2.4.2/include -DLIBDIR=3D\"/= > usr/local/gmt/lib\" -DSI gmt_map.c > /opt/SUNWspro/bin/cc -c -O -I/usr/local/netcdf-2.4.2/include -DLIBDIR=3D\"/= > usr/local/gmt/lib\" -DSI gmt_plot.c > /opt/SUNWspro/bin/cc -c -O -I/usr/local/netcdf-2.4.2/include -DLIBDIR=3D\"/= > usr/local/gmt/lib\" -DSI gmt_support.c > "gmt_support.c", line 1381: warning: argument #4 is incompatible with proto= > type: > prototype: pointer to function(pointer to const void, pointer to co= > nst void) returning int : "/usr/include/stdlib.h", line 120 > argument : pointer to function(pointer to int, pointer to int) retu= > rning int > "gmt_support.c", line 1525: warning: argument #4 is incompatible with proto= > type: > prototype: pointer to function(pointer to const void, pointer to co= > nst void) returning int : "/usr/include/stdlib.h", line 120 > argument : pointer to function(pointer to double, pointer to double= > ) returning int ... Although I don't know GMT all that well, I suspect that all the warning messages may be safely ignored because the C compiler is converting (for example) pointers to integers into pointers to voids as required by the netCDF 2 interface. Only the GMT authors, however, would know for sure. > /opt/SUNWspro/bin/cc -O -I/usr/local/netcdf-2.4.2/include -DLIBDIR=3D\"/usr= > /local/gmt/lib\" -DSI blockmean.o -L. -lpsl -lgmt -L/usr/local/netcdf-2.= > 4.2/lib -lnetcdf -lm -o ./blockmean > > xdr_enum /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (var.o) > xdr_float /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (cdf.o) > xdr_int /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (cdf.o) > xdr_vector /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (sharray.o) > xdr_u_long /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (cdf.o) > xdr_double /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (cdf.o) > xdr_long /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (dim.o) > xdr_opaque /usr/local/netcdf-2.4.2/lib/libnetcdf.a= > (cdf.o) > ld: error fatal: Errores de referencia de s=EDmbolo. No se escribi=F3 salid= > a en ./blockmean > *** Error code 1 > make: Fatal error: Command failed for target `blockmean' The above is due to not referencing the network support library on the link-line (which is where the xdr_* functions reside). Try adding a "-lnsl" argument to the link-line. -------- Steve Emmerson <http://www.unidata.ucar.edu>