[aprssig] IS-to-RF packet Weighting

[Bob Bruninga]

> If the IGate software author wants to do proportional pathing for 
> gated 3rd party packets, so be it, but it's not the responsibility of 
> the (non-RF) packet originator to determine what paths would be 
> suitable for RF transmission.

Good feedback.  I should not have re-used the "proportional pathing" name
for this new ORIGINATOR function.  This new function is entirely an
indication from the ORIGINATOR.

Therefore, I am changing all previous references from "proportional pathing"
in this new proposal to "RATE WEIGHTING".  This better matches its
definition.  

The ORIGINATOR's "weighting" of his outgoing packets will be indicated by
his PATH assignment of RGATE0, RGATE1, or RGATE2.  These indicate a desired
RGATE weight of ZERO, or ONE or of TWO, where ZERO is the lowest weighting,
and 2 is his highest weight.

RGATE0 - His highest rate but least important packets
RGATE1 - Half his highest rate but twice as valuable to the originator
RGATE2 - one-forth his highest rate but four times as valuable to the
originator

These indications go in the path after the TCPIP* position in the IS packet:

WB4APR>APvvvv,TCPIP*,RGATE1:packet_data_goes_here...

These weights are used by the local RGATE operator to selectively apply his
RGATING rules so that he can automatically control the load on his local
network. The highest rate for a mobile in motion is 1 per minute of RGATE0
packets.  Half that for RGATE1 and one forth that for RGATE2 packets. IN
addition, normal APRS factor-of-2-DECAY works on all non-moving positions.

Hope that clarifies it.  The updated proposal is on:
http://aprs.org/aprs12/IS-to-RF.txt

Bob, WB4APR