bov subarray bug

[tpeterka]

in bov.hpp, should the order in MPI_Type_create_subarray be MPI_ORDER_FORTRAN instead of MPI_ORDER_C? [x][y][z] order is column-major, ie, fortran order.

[mrzv]

Well, it really depends whether you want your array to be in Fortran or C order. I have several codes that assume that the array is in C order (including numpy.hpp), so I'm going to back out your change of C to Fortran. If you really want Fortran support, we can discuss it.

[tpeterka]

It's not a matter of supporting Fortran, but whether we want the user to input sizes in [x][y][z] or [z][y][x] order in the API, whether all our codes use it consistently, and whether it is documented. I like the former, which happens to be Fortran order, but that need not be the case. When I saw it, I presumed that it was being called in [x][y][z] order, which would have been wrong for C order. You might want to double-check, or you may already know that you are using it correctly.

Let's discuss on next skype when I'm back, after SC deadline.