BendP thoughts #126
Description
The current BendP implementation ties some "reference magnetic field" to the coordinate system curvature. I am not the biggest fan of this, one reason being that it would be very awkward try to define an electric dipole, for example. My personal preference is that bending parameters should only give the coordinate system curvature and edge angles, but no information about the actual magnetic field in the magnet. The magnetic field would then be defined by MagneticMultipoleP or whatever other bending multipoles parameters