Re: Inserting/Updating Unicode Data

From: Adrian <>
Date: Thu, 11 Sep 2008 13:51:52 -0700 (PDT)
Message-ID: <>

Try to create database with characterset set to the AL32UTF8 and nchar_characterset to AL16UTF16. It will resolve all problems with a conversion made by OracleNet.
The problem can be seen when one application run on a machine need to store and display data from multiple languages. It seems that to do it properly a characterset should be set accordingly to the operation on any single word in a different languages and it is simply not possible. Setting characterset to AL32UTF8 will force to use unicode as a 'primary' characterset on all clients and by avoiding conversion it frees you from fighting with NLS variables on all clients... Keep it mind that some of the globalization parameters should be set explicitly on all clients (i.e. NLS_LANG - especially on machines where the you run imp/exp)

Adrian Received on Thu Sep 11 2008 - 15:51:52 CDT

Original text of this message