After update to 4.3b a new feature emerged: the normal flux at an electrode (viz electric current), for example, is always zero. If my electrode happens to be parallel to xy, xz or yz plane, I can naturally select a proper component of the flux vector, no problem, but if the surface is curved or the surface normal is not along the x, y or z axis, the problem becomes real.
Has anybody else noticed this, or have I done something wrong?
Has anybody else noticed this, or have I done something wrong?