Java's float and double types provide basic IEEE 754 binary floating-point arithmetic. In many circumstances, such as writing the math libraries, it is useful to be able to perform calculations with increased precision and range. BigBinary would provide that capability for the float and double types. BigDecimal does not interopate as well since it uses base 10 instead of base 2. Additionally, BigDecimal does not include the IEEE 754 special values NaN and infinity.

Name: jd38982 Date: 02/04/2002


Java's float and double types provide basic IEEE 754 binary floating-point
arithmetic. In many circumstances, such as writing the math libraries, it is
useful to be able to perform calculations with increased precision and range.
BigBinary would provide that capability for the float and double types.
BigDecimal does not interopate as well since it uses base 10 instead of base 2.
Additionally, BigDecimal does not include the IEEE 754 special values NaN and
infinity.

======================================================================