Hi! I wonder why APFloat::convertToDouble asserts if it is not an APFloat::IEEEdouble while the name "convertToDouble" suggests that it converts when it is not a double. -Jochen