Please could somebody give me advice on settings for PGSQL with a database with UTF-8 strings in a large number of languages? I've had no problems so far in storing/retrieving UTF-8 strings but can't find clear answers to other issues. What locale would be best for sorting this dataset? Does it have to be the same as the one set for the OS (these seem very limited for Solaris)? As well as European accented characters, I need to handle some others used for transcription. I have to produce simple to use, sorted lists by initial letter for users and staff. The database is for a library which takes journals from all over the world and has to store titles in all european langauges (broadly defined) plus western transcriptions of others. We are in the process of moving to unicode. The DBMS currently runs on Solaris 9 with a backup system on Fedora 3 Linux and user/editor access is via servlets (Java). Thanks for any help. Janet Bagg, CSAC, University of Kent, UK
