[mark-r]

UTF-8 was originally designed to handle codepoints up to a full 32 bits. It wasn't until later that the codepoint range was restricted so that 4 octets would be sufficient.