Adding a BOM stripper is precisely what I did. One, convert Latin-1 to UTF-8, two, strip the BOM, three, reverse the order of the lines, and the file is ready for import. > if BOMs are supposed to be ignored at the beginning of files Depends who is doing the supposing. Ignoring a BOM at the beginning of a file is not a universal or very widespread convention, except perhaps in the Windows world. (Did I contradict myself just now?) I can't say I blame tail for not including code for it. As far as it is concerned, it was just another odd character at the beginning of the line. If the line moves, so does the character.