[aprssig] APRS routing strategies

[Henk de Groot]

Some additions I wrote in a private reply that may be interesting for 
others too (slightly modified for general use):

For mobile stations I used paths like RELAY,LOCAL. At the time the 
propodal was written it was envisioned that the DIGI knows the paths to 
replace the aliases, but homestations do not. This proposal still uses the 
assumption that every home station is a RELAY in order to help the weak 
mobile signal to reach the digipeater.

Note that the digipeater is taking these special calls out of order. So an 
intermediate compatible solution would be for example to use:

PE1DNN>APRS,WIDE2-2,REGION

When this packet hits an "old" digi it will act on WIDE2-2, if it hits a 
new digi then the digi will act on REGION an rewrite the path with a path 
suitable to cover the whole region.

Note also that the next digi in the chain doesn't need any changes.

For sending messages this will also work. For example you want to send a 
message to somebody in your region. The path would then be:

PE1DNN>APRS,REGION (for fixed station)

Output of 1st digi:

PE1DNN>APRS,DIGI*,WIDE2-2

And another hop:

PE1DNN>APRS,DIGI*,WIDE2-1

Now most modern clients will reverse this path to get back to you. Doesn't 
make a difference with today. In this proposal the aliases were envisioned 
as addition, so WIDEn-N would still work.

Now with LNK en SS this could even work better, but the nice thing of it 
all is that changing from WIDEn-N to LNKn-N would be a change at the 
entry-digi, but not a change for the end user. The user can still continue 
to use REGION and expect his packet to be heard in the whole region. So no 
more need to send out BLN's to please change the path and no more pi**ed 
off users that do not follow the SIG and are confronted with mails to 
please stop QRMing the APRS network.

All you need is to define a few standard aliases that work everywhere and 
where the user can rely on, regardless how many other paradigm shifts the 
digipeater network has to endure.

Since August 30, 2001 nothing has changed. Still the argument is that this 
proposal obsoletes 1000s of digis (dispite there is a migration scenario 
that is safely ignored) and that some day there will be new digis (how 
they would suddenly appear I don't know, apparently magic).

Unfortunately defending this proposal is a bit like fighting windmills...

Kind regards,

Henk.