"John Hansen" <john@geeknet.com.au> writes:
> Look at the upper/lower I sent to the list, they should be able to
> replace upper/lower for the utf8 encoding.... (and works independent of
> locale)..
I was under the impression we couldn't use these, precisely because they
weren't locale-aware. ("It works for most people" isn't good enough.)
In any case, don't we need a solution that covers sorting (strcoll) as
well as upper/lower?
regards, tom lane