In that case, one thing you can do is to make the font follow system locale, and not force it to Western or something.
So if he's Russian, he'll see all accented characters (of filenames and mp3 id3) in cyrillic.
The other solution would require some sort of smart auto-detection algorithm, which will prolly fail a couple times (hebrew and russian codepages use a rather similar range of characters in the 0x80-0xff interval)
Third solution: extend playlist or .SET ("dotSET" à la .NET
) with an option to save the file's locale (for the case of non-unicode mess, like old filesystems (FATxx), plain crappy filenames, ID3) and then, for that file, that decoding will be used by windows.
Like, you have the playlist, and the person set tahoma for pls area font: one file is in cyrillic or supposedly so, and we set that for later use in dotset, and then that one line will be FONT:Tahoma:n:8:Cyr, the others will be FONT:Tahoma:n:8:Wes since why would we default to anything else
For the case of Unicode stuff (properly named files in an NTFS file system, Ogg tags), I think displaying Unicode is not a problem in any post-95 Windows (i.e. 98 and above), but input of non-locale-range Unicode characters is not possible in non-NT systems. The only place where we could need the input of weirdo Unicode characters is the Ogg Tag Editor. The only thing displaying depends on is the fonts, but I'll gladly supply and host some free fonts that have 32000 glyphs if you guys need it