The original description was slightly misleading by saying that utf-8 will encode and find the length, whereas in reality the length was produced by len.