Comment (by Roland Haas):

@Miguel Zilhão it does not change any value. It is just its returned values will differ on different MPI ranks since a given boundary (of a grid patch) may be of a different type on different MPI ranks, eg in a situation with 2 ranks and a grid split like this:

| P1 | P2 |
x->

then for process 1 the left x-boundary is an outer boundary, but for process 2 it is an interprocessor boundary. Thus GetBoundarySizesAndTypes will correct report that fact on each patch.

The point is that when selecting boundary conditions you do not want a per-patch information, but a per-level information (this is the reason for it being n LEVEL mode and not LOCAL mode).

The GetBoundarySpecification on the other hand is not concerned with per-patch information and just returns the values that were set in CoordBase’s parameters (repos/cactusbase/CoordBase/src/Domain.c):

CCTK_INT CoordBase_GetBoundarySpecification(CCTK_INT const size,
                                            CCTK_INT *const nboundaryzones,
                                            CCTK_INT *const is_internal,
                                            CCTK_INT *const is_staggered,
                                            CCTK_INT *const<
--

Ticket URL: https://bitbucket.org/einsteintoolkit/tickets/issues/2288/leans-boundary-condition-setting-for-bssn