Constraints on the generalized Chaplygin gas from supernovae observations