[Xastir] Re: 05 APRS PAIN IMPROVEMENT PLAN

[Curt Mills]

On Tue, 29 Mar 2005, James Ewen wrote:

> > Bletch.  I forgot to mention, the SEA digi does not handle packets sent
> > to RELAY, WIDE or TRACE unless I hear them "first."  In other words I
> > will handle a packet that arrives to a WIDE3-3 but not a WIDE3-2.  I
> > leave it to the higher elevation machines surrounding Puget Sound to
> > handle those packets.
> 
> You mention not acting upon RELAY, WIDE, and TRACE, but then use WIDEn-N as an
> example.

WIDE/TRACE is vastly different than WIDEn-N or TRACEn-N.

WIDE/TRACE are not dupe-checked in a Kantronics TNC, while WIDEn-N
and TRACEn-N are.  WIDE/TRACE are bad ideas around here,
WIDEn-N/TRACEn-N work much better and are easier on the network.

Paths like "wide,wide,wide" do terrible things to the network
because one digi can repeat the same packet multiple times as it
ping-pongs back and forth between the mountaintop digis.  The
dupe-checking in the Kantronics TNC's is bypassed.  You end up with
MANY extra copies of the packet running around the network for a
while.

Paths like "wide2-2" are much nicer to it, as that alias is
dupe-checked at each digi.  If the packet was digipeated within the
last 30 seconds, it won't be digipeated again by that same
digipeater.  It's a flooding protocol that attempts to only flood
through an area once.

-- 
Curt, WE7U.				archer at eskimo dot com
http://www.eskimo.com/~archer
  Lotto:  A tax on people who are bad at math. - unknown
Windows:  Microsoft's tax on computer illiterates. - WE7U.
The world DOES revolve around me:  I picked the coordinate system!"