I have an issue with welcome texts. Diacritic characters of European languages show like cyrillic, because the text is written in single-byte characters. Host owners see their welcome.txt in their codepage (French, Italian, Swedish), but my local LFS takes that text as cp1251 and Cyrillic. I see "piщ" instead of "più", "fцr" instead of "für" and so on. Other users must also have the same problem.
BTW, what happens if someone makes a multibyte welcome.txt?
IMO, it´s better to detect codepage (or decode unicode txt) and tag the welcome message on host.