UCS-2 and UTF-8

AIX® provides a set of code sets that address the needs of a particular language or a language group. None of the code sets represented in the ISO8859 family of code sets, the PC code sets, nor the Extended UNIX Code (EUC) code sets allow the mixing of characters from different scripts. With ISO8859-1, you can mix and represent the Latin 1 characters (languages principally spoken in the U.S., Canada, Western Europe, and Latin America). ISO8859-2 covers Eastern European languages; ISO8859-5 covers Cyrillic, ISO8859-6 covers Arabic, ISO8859-7 covers Greek, ISO8859-8 covers Hebrew, ISO8859-9 covers Turkish, IBM-eucJP covers Japanese, IBM-eucKR covers Korean, IBM-eucTW covers Traditional Chinese. The point is that none of the above code sets covers all of the languages.

The International Organization for Standardization (ISO) addressed the limited language coverage by code sets by adopting Unicode as the encoding for the 2-octet form of the ISO10646 Universal Multiple-Octet Coded Character Set (UCS-2). The 32-bit form of ISO10646 is known as UCS-4 for 4-octet form. AIX uses the 16-bit form of ISO10646 and uses the standard label UCS-2 to describe this encoding.

Although UCS-2 is ideal for an internal process code, it is not suitable for encoding plain text on traditional byte-oriented systems, such as AIX. Therefore, the external file code is The Open Group's File System Safe UCS Transformation Format (FSS-UTF). This transformation format encoding is also known as UTF-8, and UTF-8 is the label that is used for this encoding on AIX.