[aprssig] [APRSSIG] Universal APRS Messaging
bruninga at usna.edu
Tue Oct 28 10:07:23 EDT 2008
>> Decay Algorithm... something like
>> Now, then 15 sec later, then 30,
>> then 1 min, then 2, then 4 then 8
>> then 16 and then stop.
> Unfortunately, this decay algorithm is
> incompatible with the dupe checking
> algorithm implemented... on APRS-IS.
> ... a sliding 30 second window...
> ... would, at a minimum, miss the first
> and second retries (15 and 30 seconds).
> Just food for thought...
Yes, thanks. We considered that and think that is OK. On the
digipeaters, if the first or second retry is ignored due to
being a dupe, then that is OK, because it shows that the packet
was successful on the first transmission and so the packet had a
good chance of being digipeated far and wide. So at least the
packet initially got into the system. Then the 3rd and
subsequent longer period retries are only providing backup in
case there were further collisions down the line.
Same thinking goes for Igates. If it got to the Igate first,
then it is ok for the first two rapid retries to be ignored due
to the dupe filter. Subsequent retries do help for any
collisions when the packet was igated back to RF.
Those first two (8 sec and 16 sec) retries are just for the case
when there is an initial collision on the first packet. This is
why only APRS clients that fully implemented the Decay Algorithm
are the ones that provide significanly more reliable and more
rapid exchange of data. Those that only implemented a
fixed-rate-retry method were clearly not designed for the
real-time intent of the original APRS.
More information about the aprssig