[Paraview] netCDF reader

[Brockmann Patrick]

Hi all,

Testing the netCDF reader with "spherical coordinates" set, I get
a vtkStructured Grid incomplete. There is no circular topology.
I was expecting a complete sphere where points from first column would be
connected to points from last column in the vtkStructuredGrid created.

Is there a way to do this from an existing filter or from
the python programmable filter ?
Or better could the netCDF reader be corrected 
(VTK/IO/vtkNetCDFCOARDSReader.cxx) ?

By the past, I have installed and used the external plugin CSCSnetCDF 
reader
from John Biddiscombe of the Swiss National Supercomputing Centre.
I was quite happy with it.
Today, I discover that paraview binairies are distributed with the 
netCDFCOARDS reader
with some features I don't like. For example, that fact that the vtk 
structure created
are different following the netCDF grid topology (regular spacing or 
not, spherical coordinates or not).

I think it would be attractive for all the General Circulation Model 
(GCM)/ /community 
to provide a netCDF reader, closer to the CF (Climate and Forecast 
netCDF convention) widely adopted.
This reader would create for example a same vtkPolydata structure 
whatever the netCDF grid read
as in the vtk application described from 
http://www.prism.enes.org/Publications/Reports/Report19.pdf

Is there any interest to develop this netCDFCF reader among offcial dev 
team ?

Thanks
Patrick