apparently some mpi implementations don't do mpi_c_float_complex. so I'm
using mpi_complex, even though they claim I shouldn't use fortran
stuff...

I think. I can't really figure out why they have so many different names
for these things, and in principle the C data types should be bit
compatible with the fortran data types, but I figured I may as well make
sure...

I was normalizing many many times because of the previous error, so now
everything seems to make sense, and I get collapse of energy and
enstrophy with those given by vortex

this seems to fix the laminar solution to Kolmogorov flow problem.

still debugging. at the moment, fields are set to zero a bunch of times,
but it doesn't seem to help

still having problems though. I *think* I know where the problems are
coming from though.

now it will be easier to handle functions like correlations etc.