In computer science, all data is bits With n bits one can encode at most 2^n values, numbered from 0 to 2^n - 1
With 8 bits one can encode 256 different values, variying from 0 to 255 or, if we interpret those 8 bits as signed integers, those vary from -128 to 127 and if we interpret those 8 bits as a character, we have (most usualliy) 128 ASCII values and for the 8th bit it depends on the character encoding : cp1252 latin_1 iso8859_15 mac_roman and so on
For IEEE754 numbers we have sign, biased or unbiased exponent and mantisssa and special values : nan inf +0 -0 and more
The same 64 bit values can mean a lot of different things :
- unsigned long long value from 0 to 2^64 - 1
- signed long long value from 6 2^63 to 2^63 - 1
- 2 utf-32 characters
- 8 cp1252 / latin_1 characters
- 1 IEEE754 64 bits double (float with double precision)
- 2 IEEE754 32 bits floats
- ... and more
Both programs aim at taking some binary (or string) input, packing it into bytes and unpacking it into a lot of different formats
The IEEE754 class also allows to change one bit of the number to watch the difference
This has been roughly tested manually, not deeply It should work for "standard" input and "standard" bit ordering but some edge cases are not yet neither tested nor well implemented