I have a database in pg 8.0.4 that is encoded as UNICODE. Somehow, some “strange” characters made it in there. Mostly, several instances of ‘ (apostrophe) are really some extended character. They display in a text editor as \222 (or something similar). I’m not sure how that happened, but they display properly in a web application; in psql they seem to be ignored.
In any event, if I pg_dump the database and attempt to load it into 8.1beta3, I get an error for each of these characters. The beta3 database is encoded in UTF8, which I understand is the same as UNICODE. Perhaps this is incorrect?
I’m not familiar on the character encoding issues, so any help is appreciated!