unicode NLS_CHARACTER_SET question 8i on NT

Hi:

I (like several others) am confused about what I need to do in order to have a multilingual capability for my database (8.1.6). The online docs indicate that one USUALLY has to create the database with the proper NLS_LANGUAGE and/or NLS_TERRITORY, etc. There appears to be one exception: When the language you want to switch to is a superset of the current. My current territory is AMERICAN, the current NLS_CHARACTER_SET is WE8ISO8859P1. I want (I think) to make everything stored in UTF8. I THINK UTF8 has to be a superset of WE8... and GB2312, and BIG5, etc. and most anything else I want to store in my Oracle database.

1. Is this correct?
2. Does this mean I can get away with an ALTER DATABASE command?
3. What must I alter? Territory AND CHARACTER_SET AND CURRENCY,etc. ?

Part B:

I have gotten a Java 2 application to read/write GB2312 strings to/from a file using buffered readers and writers. It reads in GB2312, writes out UTF8, and can read in the UTF8 again and make it look correctly in the textarea. I am trying to store the UTF8 in Oracle now.

Q:Am I correct in thinking Java 2 String types are UTF8 unicode?

Thanks for any help you can give.

Mike Received on Tue Aug 29 2000 - 00:34:45 CDT