Improvements at layer 1 and at layer 2 would both be useful. If done "right", both could be developed apart from each other. A better layer 1 protocol should work just fine with AX.25 as a data layer, or with some improved "future" layer 2 implementation. Any improved layer 2 should work fine whether it's using a bell-tone physical layer or some improved magical future layer.
While it is true that certain implementation choices at one layer can affect how another layer works, there are broad choices that can be made without affecting other layers. An example of a bad choice would be a layer 1 protocol that required a long leader for sync coupled with a short-timeout/short burst non-windowed ARQ layer 2. I'd love to say that this was a theoretical example, but I've seen people who expected this to work. -- Regards, Robert Thompson ==================================================== ~ Concise, Complete, Correct: Pick Two ~ Faster, Cheaper, Better: Pick Two ~ Pervasive, Powerful, Trustworthy: Pick One ~ "Whom the computers would destroy, they first drive mad." ~ -- Anonymous ====================================================
