JIRA sometimes uses non-UTF-8 as default and doesn't tell the user

In Setup.java we have the following lines of code

        encoding.put("fr_FR", "ISO-8859-1");
        encoding.put("pt_BR", "ISO-8859-1");
        encoding.put("zh_CN", "Big5");
        encoding.put("ja_JP", "Shift_JIS");

In other words, if a user selects their language as pt_BR then they get ISO-8859-1 as their encoding instead of UTF-8. I don't know if that is wrong per se but

a) it causes problems
b) the admin sections still claims that the default is UTF-8 and there is no indication to the user that this was silently changed on them.

For 4.0.1 we should investigate whether we really need this special casing or if we can just use UTF-8 every where. At the very least we need a better indication either during setup or in the admin section that this has happened.