As global ebusiness continues its growth into every aspect of industry as an infrastructure for information and business management, it is becoming very crucial to make internet application with multilingual capability. Unicode is widely used as a base to provide this capability.

Oracle supports Unicode as UTF-8 encoding since Oracle7. In the newest Oracle9i release, this support is further enhanced to better serve the global ebusiness needs. Codepoint semantics is introduced for text data so that UTF-16 semantics can be built upon UTF-8 encoding which will easily support application server that is built upon UTF-16. The benefit of this solution is to reduce the migration effort and to increase storage efficiency for Latin data. A Unicode data type is introduced to build Unicode application independent of database character set. This data type enables existing application to be gradually migrated into Unicode and it can pick either UTF-8 or UTF-16 as its encoding for more storage efficiency based on data distribution.

This paper describes the functionality of codepoint semantics and the new Unicode data type in Oracle9i database release. Design choices, such as codepoint semantics vice Unicode data type, UTF-8 vice UTF-16, will be discussed. A brief description of new Unicode access interface will be given.