Either the Linux Standards Base
or, more commonly, Least
. See MSB
In little-endian architectures such as x86 and newer Alphas in stupid mode, the address of the LSByte is the same as the address of the value itself. This leads to free typecasting (well, downcasting anyway) of pure-integer types, at the expense of some code being possibly broken in weird ways which will make it unportable to a big-endian machine.
The LSBit isn't a very interesting beast. It's always equal to 1 (assuming either 2's complement or unsigned pure integers). Yawn. Even with negative values it's equal to -1 - not much more exciting, except that at least in 2's complement it's -1 when it's unset and 0 when set. It's only really interesting when you're dealing with non-integer types. LSByte isn't very interesting either; it's just equal to i, where i is the value of the byte (or 256-i if it's in a negative 2's complement number, still no BFD).