#797: N13CarpetRegrid211snap_coarseE does not hold after being enforced in cell centering ---------------------+------------------------------------------------------ Reporter: rhaas | Owner: eschnett Type: defect | Status: review Priority: major | Milestone: Component: Carpet | Version: Resolution: | Keywords: ---------------------+------------------------------------------------------
Comment (by rhaas):
Yes, I meant "returns an empty box". It's actually in the posted text but since the code snippet is not word-wrapped one has to scroll to the right to see it.
I'll give the suggestion of adding/subtracting on reffact a try. There seemed to be sufficiently many cases (vertex/cell centered, reffact even/odd, reffact > 2, location of the edge) to make asking for some help a good idea. I'll also add a test for empty at the very beginning in the current calculation and compare both variants (code both and compare the resulting boxes).
Turns out this would have been so much easier to find had I turned on CARPET_DEBUG :-) We live and learn it seems.
I think I'll also pick up my footer of asking for this trac spam issue to be fixed again (would be nice to fix this before the next release) :-)