I propose that STOMP 1.2 also support being access via telnet on systems
where hitting Enter generates a \r\n pair (like windows).
Right now the way the 1.1 spec is written, telnet access is supported but
only on systems where hitting enter generates a \n.
I think we can support this by updating the grammer so that:
LF = [<US-ASCII carriage return (octet 13)>]
<US-ASCII new line (line feed) (octet 10)>
And adding the escape \r to the value encoding section.