Jul 17 2008, 11:14 AM
We are going to change our database character set from WE8ISO8859P1 to AL32UTF8. We know that this could cause problems related to performance or disk space because we will change from a single-byte to a multibyte encoding.
We have an Oracle 10g database.
Is there any other restriction or problem? For example: using Oracle Text or other application?
Jul 17 2008, 11:48 AM
Ever see those messages about possible character loss when using the old export/import and the character sets are different between the server and database? An "a" with an umlaut can turn into a regular "a" and then the meaning or context of a word can change. Then there is client versus server compatibility, and set/superset of the character set to consider. Those are some issues you need to investigate ahead of time, definitely check out the globalization guide.http://download.oracle.com/docs/cd/B19306_...t.htm#sthref157
Jul 18 2008, 08:21 AM
"We know that this could cause problems related to performance or disk space"
Because rows store twice as large, watch out for row chaining/relocation on import.