Improvements in handling bytes encoding

In R, a string can be declared to be in bytes encoding. According to ?Encoding, it must be a non-ASCII string which should be manipulated as bytes and never converted to a character encoding (e.g. Latin 1, UTF-8).