In most computers, each byte has its own unique address. Because the
address refers to all eight bits of the byte, it's not useful to talk about
the order of the bits within the byte. We can discuss the significance of
the bits, but not their order.
Since the bytes of multi-byte numbers are individually addressable, we do
need to consider the order of their bytes in memory. Two reasonable
possibilities exist: big-endian and little-endian byte order.
Big-endian machines store the most-significant byte of a multi-byte number
at the lowest-numbered address, with the other bytes following in decreasing
order of significance. Little-endian machines store multi-byte numbers in
the opposite order: from least-significant to most-significant.
Within a single computer, the byte order doesn't matter because the
processor does all its memory loads and stores in the correct order,
whichever that may be. The problem arises when we need to move data between
machines with different endianness.
We could convert all multi-byte numbers to ASCII representations and
transmit those, but that would require extra bandwidth and would complicate
generating and parsing the messages. To avoid these costs, most low-level
protocols use fixed-size binary fields. If these fields are larger than one
byte then their endianness becomes an issue.
This article shows how to use C++ to deal with such mixed-endian systems in
a simple and reliable way.
