Brian, the character set of the catalog file should
be fine as iso-8859-1 (it will be converted properly to utf-8 when inserted in the db). The reasoning for having those
in iso-8859-1 was that most people editing them directly are
probably using iso-8859-1 (or -15 or something close) rather
than utf-8 so were less likely to mess them up and insert a
character in the wrong encoding).
The translation server for example runs utf-8 but still spits
out iso-8859-1 catalog files.
You would need to change the code if you wanted to have utf-8 catalog files for all locales (and then you could not merge changes from the translation server or newer openacs releases
as easily).
If there is a good reason to switch all catalog files to utf-8 its pretty easy; just make the case...
