Hi all, attached patch ensures that all locales are available in the UTF-8 variant. The content is derived from the Unicode Common Locale Data Repository (CLDR), which is the nearest thing to an authoritive source. There are three logical changes as a side effect: - no_NO is now an alias for nb_NO, not the other way around - Serbian locales are aliases for either the cyrillic or the latin script version, prefering the former for the UTF-8 variant. - I've used "EUR" as currency_symbol in ISO8859-1, this seems to be more correct than using "Eu" as done before. The only locale not touched is ja_JP.ct since I don't know a way for creating it. Does someone have a recipe for that? Is this locale variant still used? Please test any locales you are comfortable with. Joerg
Attachment:
cldr.diff.xz
Description: Binary data