when one prints out a string with some extended(I guess it's unicode), writeln prints out the ascii versions that do not correspond to what they really are. e.g., an umlaut is printed out as 1/2 or something. how to get it to print the correct codes?