This is an old revision of this page, as edited by Wernher (talk | contribs) at 00:02, 21 March 2004 (lk fixes). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.
Revision as of 00:02, 21 March 2004 by Wernher (talk | contribs) (lk fixes)(diff) ← Previous revision | Latest revision (diff) | Newer revision → (diff)
When an integer or any other data is represented with multiple bytes, the actual ordering of those bytes in memory, or the sequence in which they are transmitted over some medium, is subject to convention. This is similar to the situation in written languages, where some are written left-to-right, while others are written right-to-left. The convention is called endianness, describing the method either big-endian or little-endian. Endianness is also referred to as byte sex.
Note that the endianness does not matter in dealing with a sequence of single bytes. This is the case of a single-byte string literal where one byte corresponds to one character. Strings encoded with unicode UTF-16 or UTF-32 are affected by endianness because in those, a set of two or four bytes represent one character.
Endianness in computers
When some computers store a 32-bit integer value in memory, for example 0xA0B70708 (in hexadecimal notation), they store it as bytes in the following order: A0 B7 07 08. That is, the most significant byte (A0 in our example) is stored at the memory location with the lowest address, the next significant byte B7 is stored at the next memory location and so on.
Architectures that follow this rule are called big-endian and include Motorola 68000, SPARC and IBM 370.
Other computers store 0xA0B70708 as 08 07 B7 A0, that is, least significant byte first. Architectures that follow this rule are called little-endian and include the MOS Technologies 6502, Intel x86 and DEC VAX.
Some architectures can be configured either way; these include ARM, PowerPC, DEC Alpha and MIPS. The word bytesexual, said of hardware, denotes willingness to compute or pass data in either big-endian or little-endian format (depending, presumably, on a mode bit somewhere).
Still other (generally older) architectures, called middle-endian, may have a more complicated ordering such that the bytes within a 16-bit unit are ordered differently from the 16-bit units within a 32-bit word. For instance, 07 08 A0 B7. Middle-endian architectures include the PDP family of processors.
Endianness in communications
In general, the NUXI problem is the problem of transferring data between computers with differing byte order. For example, the string "UNIX" might look like "NUXI" on a machine with a different "byte sex". The problem is caused by the difference in endianness.
The Internet Protocol defines a standard "big-endian" network byte order, where binary values are in general encoded into packets, and sent out over the network, most significant byte first. This occurs regardless of the native endianness of the host CPU.
Serial devices also have bit-endianness: the bits in a byte can be sent little-endian (least significant bit first) or big-endian (most significant bit first). This decision is made in the very bottom of the data link layer of the OSI model.
Endianness, software, and portability
Endianness has implications in software portability. For example, in interpreting data stored in binary format and using an appropriate bitmask, the endianness is important because different endianness will lead to different results from the mask.
Writing binary data from software to a common format leads to a concern of the proper endianness. For example saving data in the BMP bitmap format requires little endian integers - if the data is stored using big endian integers then the data will be corrupted since it does not match the format.
The OPENSTEP operating system has software that swaps the bytes of integers and other C datatypes in order to preserve the correct endianness, since software running on OPENSTEP for PA-RISC is intended to be portable to OPENSTEP running on Mach/i386.
Discussion, background
Big-endian numbers are easier to read when debugging a program but less intuitive (because the high byte is at the smaller address); similarly little-endian numbers are more intuitive but harder to debug. The choice of big-endian vs. little-endian for a CPU design has begun a lot of flame wars. Emphasizing the futility of this argument, the very terms big-endian and little-endian were taken from the Big-Endians and Little-Endians of Jonathan Swift's Gulliver's Travels, two peoples in conflict over which end to crack an egg in the voyage to Lilliput and Blefuscu.
See the Endian FAQ (external link, below), including the significant essay "On holy wars and a plea for peace" by Danny Cohen (1980).
The written system of arabic numerals is used world-wide and is such that the most significant digits are always written to the left of the less significant ones. In languages that write text left-to-right, this system is therefore big-endian, in languages that write right-to-left, this numeral system is little-endian. The spoken numeral system in English is big endian (with minor exceptions: we say "seventeen" instead of "ten-seven"). German uses a strange mixture of big- and little-endianness: 376 is pronounced as "Dreihundertsechsundsiebzig", i.e. "three hundred six and seventy".
External links
Parts of this article were originally based on material from FOLDOC, used with permission.