AL32UTF8 is a superset of UTF8. Per the Remedy installation guide, the db character set should be set to AL32UTF8. Run csscan with a target character set of AL32UTF8 and see what it returns as lossy/unconvertible.
I believe what you are running into is a problem with how the data as it is stored in the database. Since the client has been configured with AMERICAN_AMERICA.WE8MSWIN1252 for all this time, it has been reading/writing the data with the codepoints appropriate for that localization. These codepoints are very different than the codepoints for the same character in the character set AL32UTF8. See the following articles for more info: http://en.wikipedia.org/wiki/CESU-8 http://repettas.wordpress.com/2008/05/15/al32utf8utf8-unicode-database-character-set-implications/ UTF8, in the world of Oracle, is not really a UTF-8 compliant character set (bad name). AL32UTF8 is a complaiant UTF8 implementation. See the following: http://forums.oracle.com/forums/thread.jspa?messageID=1139523 http://download.oracle.com/docs/cd/B14117_01/server.101/b10749/applocaledata.htm#i636713 http://oracleappstechnology.blogspot.com/2007/10/difference-between-utf8-and-al32utf8.html What I think you need to do: 1. convert the instance from UTF8 to AL32UTF8. Part of this will require cleaning up data (CESU-8 data, 8-bit ascii data, and other unconvertible data that csscan returns) 2. update the oracle client to use AL32UTF8 While that may look like a short list, the actual process may have complications. In the ARSCHEMA, ACTLINK, and other Remedy meta-data tables, a checksum is calculated and stored. That is based on the name, last modified by, and other values. If any of the values used to generate the checksum contains data that must be converted, the migration path becomes very difficult. The output of the csscan utility should give you an idea of what you are looking at. If you have SQL Developer installed (java based oracle client that comes with 11g), connect to your db and see what kind of data is returned when you look at it from there. This is a java based client and is fully unicode capable. You may be surprised with the results for your 8-bit ascii characters (umlauts, euro symbol, etc.). -- Axton The opinions, statements, and/or suggested courses of action expressed in this E-mail do not necessarily reflect those of BMC Software, Inc. My voluntary participation in this forum is not intended to convey a role as a spokesperson, liaison or public relations representative for BMC Software, Inc. _______________________________________________________________________________ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor:[email protected] ARSlist: "Where the Answers Are"
