> I am trying to install netcdf on a Data General Aviion under
> DGUX 4.32. Has anyone does this before, and would they be
> willing to let me in on how they did it? I seem to be stuck
> on defining the XDR / IEEE infinity values for floats and
> doubles. For all I know that's just the beginning...
You shouldn't be too concerned with getting these infinity values right.
Although there are some programs in the distribution in netcdf/extrema/
designed to help determine these values, there is actually no guarantee of
the existence of values with the required property that XDR maps the value
to/from IEEE infinity. Now that netCDF has better support for
user-specified fill values, use of the default fill values should not be
necessary. Nothing in any of the test programs or utilities will depend on
whether the XDR_D_INFINITY or XDR_F_INFINITY are defined correctly.
--Russ
