Version 4.4.3 of the NetCDF-Fortran library is now available.
This is primarily a compatibility release that addresses a
false negative introduced by netCDF-C 4.4.0 with
netCDF-Fortran 4.4.2. If you have tried to use
netCDF-Fortran 4.4.2 with netCDF-C 4.4.0 and received a
test failure in nf_test and nf03_test,
you can safely ignore these failures.
Information about the release is available in the Unidata's NetCDF-Fortran GitHub site.
Source-code zip and tar.gz archives can be found here.
compiling amber
Posted by antonio florencio on January 22, 2016 at 07:16 AM MST #
Good
Posted by Richard on January 22, 2016 at 05:23 PM MST #
ok
Posted by yasin on January 30, 2016 at 09:33 PM MST #
TANKOU
Posted by plantren on January 31, 2016 at 12:56 PM MST #
investigation
Posted by victor on February 09, 2016 at 02:56 PM MST #
very good
Posted by Renato Rodrigues on February 14, 2016 at 06:26 PM MST #
CoLM2014
Posted by lihuoqing on March 08, 2016 at 04:31 AM MST #
ok
Posted by lenin on March 09, 2016 at 08:41 AM MST #
NF90_GET_VAR causes compiler gfortran failure when the variable (dataset) is of compound user defined type, as for example
TYPE binSST
real(kind=4) :: sum
real(kind=4) :: sum2
end type binSST
TYPE(binSST), dimension (5000) :: sst_data
Assuming UNIDATA keeps supporting fortran interface to netcdf-4,
does some netcdf-4/F90 expert know if such failure is caused by the compiler gfortran (under Linux ubuntu), or if it results from some feature not fixed yet in the fortran interface release. In this case, when is expected to have available a new fortran interface release with the fixed problem, please ?
Some specific suggestions are welcome as that problem is a bloc road on my work.
Posted by Arlindo Arriaga on April 21, 2016 at 04:20 AM MDT #