CJKV Information Processing, 2nd Edition

Gaiji Handling

|

453

SSCs are those characters that are considered standard on one OS, but not across multiple

OSes. In a closed system or environment, there is eectively no dierence between SDCs

and SSCs, but the moment one attempts to interoperate between OSes, they become very

dierent entities, with very dierent results. For example, the traditional-form kanji 󰤎

(meaning “black”) was considered an SDC under Windows 3.1J and Windows 95J, but

because it is also an SSC, and specic to Windows, it was not available under other OSes,

such as KanjiTalk or JLK (Mac OS). is particular kanji was a gaiji, specically a UDC,

as far as Mac OS is concerned. is has changed with Mac OS X, and this character is now

considered standard. And, to be absolutely clear, gaiji are not limited to ideographs—

many are symbols or logos.

Both types of gaiji pose problems when information interchange is necessary. e target

encoding or character set may not support certain characters that are used in a document.

is is especially true for user-dened characters, which are usually specic to a single

person’s environment. Even JIS X 0212-1990 characters can be considered gaiji if the tar-

get system does not support their use.

*

Some might imagine that Unicode is a solution

to this problem. While this appears to be true at rst glance (aer all, there are now over

70,000 ideographs from which to ...