| ||||||||||||
|
|
Precision Limitation, versions prior to 1.9If you do not understand the following paragraph, then you will likely not be bothered by any precision limitation. There is a limitiation in Visual C++ which limits the precision possible for math functions. The maximum precision is a 64-bit floating point, or double, which has 52 bits of integer (or significand) precision. The other bits are used for exponent and sign. Because there is inherent loss of information beyond 52-bits, conversion between one base to another will be limited to 52-bits. Beyond that, you will lose the least significant bits of accuracy beyond 52 bits. To work around this limitation, when converting directly from a hex/binary to binary/hex, a different algorithm is used which is accurate up to 500 bits. Once a mathematical operation is performed such as addition or subtraction, the result is again limited to 52 bits of precision. Precision Limitation, version 1.9Precision was dramatically increased to 80 significant figures, which is 269-bits of integer (significand or mantissa) accuracy. This astronomical accuracy is way beyond the 64-bit double floating point. It is better than 128 bits and better than 256 bits. I can't imagine needing more precision than this. Ingeters can be accurately represented up to 99,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999,999. | |||||||||||