While most of the time Jaikoz displays non-English (mostly Greek) characters correctly, sometimes it shows just gibberish. The very same files display correctly on Winamp and Windows Explorer. If the files are updated from MusicBrainz, the new non-English characters show up fine.
I have kept some sample files unprocessed so if you want to duplicate the problem I can arrange to send them to you.
On my computer I get the incorrect values in both JaIkoz AND Windows Explorer. The problem lies with the program that orginally created these files. The ID3 (for MP3) specification defines that ISO-8850-1 is used as the charset encoding to use by default , although you can use UTF16 (and with ID3v24) UTF8 as well. ISO-8859-1 uses one byte per character, and works fine for English and other latin languaes such as French, Italian and Spanish, but doesnt work well for language with other scripts such as Greek. I think Im right in saying greek is a cyrillic script and this is better served by another one byte per character encoding charset Windows-Cp1251 which is probably the default encoding on your computer which is why it looks ok on Windows Explorer for you, BUT use of this within Mp3s is illegal and means the mp3 is not transferable to other computers.
What I could do in Jaikoz is provide an option to 'Convert values from Charset x' to correct incorrectly encoded files. It cant automatically identify and fix these cases, you would have to select the fields and the charset yourself.