Boundedness of the range of a strategy-proof social choice function

For the provision of $m\ge 1$ divisible public goods, relatively weak restrictions on the domain of a strategy-proof social choice function are identified that ensure that its range is bounded. Domain restrictions are also identified for which strategy-proofness implies that the range and the option sets of a social choice function are compact. To illustrate the usefulness of these results, it is shown how a theorem about generalized median voter schemes due to Barberà, Massó, and Serizawa can be established without their assumption that the range of a social choice function is compact provided that the tops of the preferences are not restricted to be finite.