[standards-jig] A Proposal to the DTCP/JOBS/PASS madness

Justin Karneges justin-jdev at affinix.com
Sat Dec 14 07:22:07 UTC 2002


Hi Dave,

On Friday 13 December 2002 08:31 pm, Dave Smith wrote:
> With this in mind, I am going to adjust my proposal such that when the
> TCP connection is established from the Target to a Initiator/Proxy, a
> four-byte integer in network byte order is sent from the
> Initiator/Proxy to the Target. That integer will be the remote port
> which the Target will use to identify itself.
>
> Obviously, it's not ideal to have _any_ line protocol, but if we must,
> I'd prefer it to be as minimal as possible.

If you're going to add a line protocol, then I suggest looking back to DTCP 
where it is already very minimal.  The novelty in your proposal was that it 
claimed to have no line protocol, but if that is not the case then we don't 
need another DTCP, do we?  I've already solved the proxy case in my latest 
variation.

Anyhow, the other issue at hand is the "proprietary" aspect of these line 
protocols, including your idea of a 4 byte binary header.  IMO, a small 
non-standard protocol is the way to go, but temas, pgm, and others insist on 
using HTTP for a handshake.

-Justin




More information about the Standards mailing list