Missing Help about Unicode enable

[cgertler]

Hi,
i can not find any documentation about the preference Oracle -> Options -> Unicode enabled
Could you please provide some information about this (missing in the online help and also in the printed version)?

We are running a 10.1.0.3 Server with NLS-Charset UTF8 on Linux. In case this property is enabled, i can not edit CLOBs using PLSQL-Developer. I receive the following error message when posting the clob data.
ORA-24812: characterset conversion from or to UCS2 not succesful.
If the feature is disabled, the CLOB can be edited and posted.
But i think we will loose some characters from poland or czechia when this feature is disabled and content is posted to the database.
Any ideas on this?
thanks in advance
claus

 

[Marco Kalter]

If you disable the preference, the UTF8 characters will be converted to the client character set. You may indeed lose certain characters, depending on the CLOB contents and the client character set.

When enabled, the data will be passed as Unicode, so no character data is lost. However, for CLOB's the data is always passed as UCS2. Apparently something goes wrong with this conversion:

ORA-24812: character set conversion to or from UCS2 failed
Cause: If the database character set is varying-width, the CLOB/NCLOB value is implicitly converted to or from UCS2. This implicit conversion failed.
Action: Contact Oracle Worldwide Support.