Reflections on PTTH

[Maitland Bottoms]

One thing I didn't say in the PTTH document was that I was thinking
that the digital procedural packets would take the form of AX.25
datagrams. Maybe even at 300 baud AFSK.

So my examples were conceptually digital AX.25-like procedural packets
interspersed with analog FM messages. My concept for a radio that
implements this is that the data tones would be surpressed when
detected, and the user would only get the audio payload via the "user
interface". 

What didn't even cross my mind back then was 802.11. Now I must point
to something like
http://www.commsdesign.com/design_corner/OEG20020201S0035 
which talks about an RTS/CTS handshaking taking place at a baseline
common protocol, while possibly delivering packet payloads via faster
enhanced capability signalling.

With Paul's idea of a common HF signalling mode negotiation protocol,
I should add examples for a MEMI types which will cause the endpoints
to switch to clover or pactor kinds of modulation.

The Internet protocols formed via the exchange of lines of text (like
SMTP, HTTP, POP... ok, most of them) depend upon a reliable data
delivery of the TCP/IP protocol. With a radio handshake protocol, the
reliablility cannot be assumed. So the handshake procedurals should
have checksums at the least, and hopefully redundancy for Forward
Error Correction. Only then can the endpoints act with a resonable
expectation that they are taking the correct actions.

Since things like the world-wide-web and 802.11 seem to be sucessful,
borrowing from their techniques is likely a good plan.

Hope this helps,
-Maitland