Issues:
 - Texts are stored in ANSI with a codepage deducted from file extension (*.locale.libx), means that if we open it with Notepad it won't detect locale
 - Some files have text in UTF8 stored as ANSI (probably due to translators using text editors with poor Unicode support), leading to text corruption

Suggestions:
 - Convert all libx files into unicode (UTF8 with BOM?)
 - Keep code and internal error strings in ANSI-English