Double-double Packages

Pure double-double precision functions *with strict error bounds*.

Emulated float128 or double-double arithmetic. A floating point expansion with 31 accurate decimal digits.