All nongeeks please ignore this post, I just thought I'd share this in case someone comes across the same issue:

Here's the scenario: I had to switch a J2EE application from using iso-8859-1 to utf-8 because of some conversion problems when writing data to the DB(here is some useful info and what needs to be done to tomcat and jsp pages to fully work in utf-8). I was getting strange name matching issues with umlaut characters and spent quite a while trying to debug my own code. Eventually I found out that it wasn't really my fault. There seems to be a bug in postgresql < 8.0: character manipulation functions like lower() and upper() don't work correctly with unicode! They just drop special characters after the conversion. The workaround I came up with was to just not do any character conversions in java but use lower() on input and column values.

Technorati Tags: bugs, DB, postgresql, unicode