The multilbyte character routines of libc are not really supported, and as defined in ansi_C.doc: "There are no multi-byte characters.".
It would be good to support UTF-8 as multibyte encoding with wchar_t equal to an unsigned short which would give minimal support to the Unicode BMP. Why not UCS-4 for wchar_t like UNIX systems, because ACK is made for smaller systems in general, so lets go to minimalist especially on memory side.
I am writing this here, as a personal reminder also so its not forgotten.
Functions that need to be implemented: mblen.c, mbstowcs.c, mbtowc.c, wcstombs.c, wctomb.c