Jianping,
I knew that the Oracle implementation of UTF-8 was limited to 3 bytes. It
is not the only software that has this limitation. The problem is that I
had expected Oracle to fix the code to extend the encode/decode logic to
upgrade the UCS-2 to UTF-8 encoder to be a UTF-16 to UCS- encoder. I
expected that they would be OK as long as they did not encode any
surrogates. It expected that the change would be like UCS-2 to UTF-16
upgrades.
I also know that you can sometimes change the database encoding as long as
the data is a subset of the encoded data. For example if you have pure 7
bit ASCII you can change the encoding to UTF8 because the characters are
UTF-8 compatible.
If there are no surrogates in the database, is there any reason that I can
not change the database from UTF8 to AL32UTF8?
Carl
