[f2py] SOLVED: segfaults

Hi list,

I have no more segfaults when running the fortran routines. It turns out
that I have to create enough space for some data arrays, whereas they
are only declared as:

integer          foo(1)
double precision bar(0:1)

in the fortran source. Is this technically allowed? Through some magic I
didn't follow through they are filled with values at much larger
indices. Since the f2py conversion routines wouldn't allow me to pass
larger arrays than size 1 and 2 respectively, I changed the declarations
to:

integer		 foo(*)
double precision bar(0:*)

and now everything works. Thanks again for this great piece of software!
I wouldn't want to think about writing my whole project in fortran or
writing out intermediary files or something like that.

David

PS: the might-be bugs in my test.f script from my last mail still stand.