In areas from blood flow to biosensor applications, it is essential to predict fluid flow. The standard model states that fluid velocity is zero at solid surfaces, but evidence is accumulating against this in many situations, especially regarding aqueous solutions and surfaces coated with polymer cilia. We have studied flow of aqueous solutions containing variable amounts of monovalent and divalent electrolyte past solid surfaces whose charge was varied and whose 'softness' was varied by polymer cilia. Deviations from the standard model are observed when the wall shear stress exceeds a critical level whose magnitude depends on the system studied. In some respects this is understood, in other respects it is not. The puzzles will be emphasized.