[aprssig] APRS ACK process

[Robert Bruninga]

>>> "David L. Martin" <dlmarti at kc2lcf.net> 5/11/2005 11:26:09 AM >My
concern was the number of packets that a 
>single message generates.  If my first packet 
>generates ten packets, and the destination 
>then sends twenty ACK's, and those twenty 
>ACK's generate ten packets each....

Which is exactly why we must obsolete the RELAY
and WIDE dupe-generating paths.

>The only reason we don't see two hundred ACK's 
>from a single message is that the system breaks 
>before it gets to that amount.

Which will all go away under the New-N paradigm
when the digis will no longer support dupe-generating
paths and users will no longer be able to use them.

But on the flip-side, the receiving station of a message
is supposed to intentionally send back an ACK for
each copy he receives.  This is just one of the ack-
enhancements that is required in the system to help
improve the lower  probability of acks.

This helps assure an ACK gets back.  But the digis,
once it sees an ACK and digipeats it, then is supposed
to ignore additional copies like it does with all other
packets.  This improves the originator of an ACKS's
probability of getting the ACK heard.

Some client software does not do this and that
is why that software has poor ACK reliability.

Bob, WB4APR