MFEM Poisson example does not produce the correct result for higher order elements in release 2.8
The poisson-mfem in the examples folder works as expected for lowest order Raviart-Thomas elements. But when increasing the order, e.g., to RT1 x P1 elements, the "solution" is completely off. Something is going wrong when changing the order, but I have not yet found out what it is. Based on a quick check, in the 2.7 release (modulo the wrong interpolate methods) this issue does not arise; in the 2.8 release it does. I continue investigating but appreciate also any ideas if present.
Does anyone have a clue what is going on? @oliver.sander did you check higher order elements when testing the new design of H(div) conforming methods?