Need to formalize how decimal numbers are parsed into integers. Do we truncate or round? Currently rounding is performed, but should this be the default behavior? I think there should be a compile ...