Unicode character problems with Oracle 8i

If the Oracle connection is used with Oracle 8i, some Unicode related issues might occur. These issues occur due to limitations of the interfaces offered by the Oracle 8i:

• W_CHAR keys in entry/subentry do not work.
• W_CHAR characters that are not also present in the NLS_LANG codepage are retrieved incorrectly from the database. W_CHAR fields are retrieved in the NLS_LANG codepage and then 'upgraded' to Unicode. This loses characters in the translation.
• C_CHAR fields are retrieved in the NLS_LANG codepage, while ITP expects them to be in the Microsoft Windows 1252 codepage.

A partial workaround would be to set the NLS_LANG codepage to the Microsoft Windows 1252 codepage. Then all C_CHAR fields will be correctly retrieved. The W_CHAR fields will be correctly retrieved, as long as they do not contain any characters that are not part of the Microsoft Windows 1252 codepage.

From Oracle 9.2 and up these issues have been fixed.

By default ITP/Server uses the Oracle 8i OCI client. Refer to the Administrator Manual of the ITP/Server Administrator for details how to configure the version of Oracle OCI client to be used. Existing models depending on the 8i behavior can be broken if this setting is changed.