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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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.
>To: "'address@hidden'" <address@hidden> >cc: "Cook, Mr. John (NRLMRY)" <address@hidden> >From: "Mantri, Mr. Ramesh" <address@hidden> >Subject: some observations on netCDF API >Organization: Navy >Keywords: 200303281619.h2SGJPB2009260 Hi Ramesh, > very recently we decided on storing some of our data in netCDF format. > This data is written inside of a C++ program and read from a Java > program. We haven't yet tried out the Java API, but we have tried out > the C++ API and we would like to share some of our observations with > you. It is possible that some of these observations stem from our > inexperience with netCDF API, but some might be useful to you in > upgrading the software. > > 1) The C++ API is certainly easier to use than the C API, but there are > problems in making use of the C++ API. > > 2) Apparently some of the header files included in some of the C++ code > follow conventions that are no longer relevant. For e.g., the use of the > include files > #include <iostream.h> > #include <strstream.h> > > 3) We tried to edit the header files to fix the compilation errors. The > compiler no longer complained, but it resulted in unresolved references. > java.lang.UnsatisfiedLinkError: > > /d/USERS/local/mantri/swdir/resin-2.1.7/doc/webapps/evis/WEB-INF/lib/libTedsProxy.so: > > /d/USERS/local/mantri/swdir/resin-2.1.7/doc/webapps/evis/WEB-INF/lib/libTedsProxy.so: > > undefined symbol: _ZN6NcFileC1EPKcNS_8FileModeEPjj > > 4) So next we tried to get a recent version of the netCDF software from your > web site and tried to configure it and build it (on a Linux RedHat 8.0 > machine with gcc 3.2). It resulted in simply too many errrors. I ended > up removing the Fortran subdirectories (which we don't need anyway), but > sadly the C++ directory too. > > 5) In short, we have been able to use the C API successfully. We would like > to use the C++ API, if possible, because it is simpler to use. Hopefully, > some time in future, that might be possible. Please try netCDF version 3.5.1-beta10, available from ftp://ftp.unidata.ucar.edu/pub/netcdf/netcdf-3.5.1-beta10.tar.Z It has improvements to the C++ interface and I think it addresses all your concerns. If you still have problems, please let us know. Thanks. --Russ _____________________________________________________________________ Russ Rew UCAR Unidata Program address@hidden http://my.unidata.ucar.edu