[aprssig] APRS Message Idea

[Curt Mills]

On Thu, 3 Mar 2005, Jason Winningham wrote:

> I just took a look at the spec (1.0.1 was the latest I found at tapr.org) and
> googled for "aprs message decay algorithm" and found nothing that would be
> useful for me if I were implementing a client.  If it's not in the spec, you
> can't blame clients for not implementing it.

Page 10 of the spec, "Packet Timing".  There are four techniques
listed:

    Decay Algorithm
    Fixed Rate
    Message-on-Heard
    Time-Out

"Decay Algorithm" is the one Bob was referring to (exponential
decay).  "Fixed Rate" is what several clients use, and what Xastir
used until a few months ago.  The spec doesn't recommend one over
the other.  A little on-the-air testing will quickly show that the
exponential decay algorithm is far superior.

Also note that Chapter 14 on messaging talks about repeatedly
sending a message until it gets an ack, is canceled, or times out.
The "Time-Out" section says that time-outs are not to be used for
transmit algorithms, Bob seems seems to agree with this, saying that
messages should not time out, but should stay at the 30-minute rate
until they are ack'ed or cancelled.  So... The two sections of the
spec conflict a little bit.

Bob also mentioned that having no indication that there are unacked
messages in a running client is a bad idea.  One could certainly get
into that condition with Xastir by closing the message windows.
Xastir currently times out messages if they're not acked, so we
avoid that problem as well as others.

Note that there's also a revision to the spec that's not on the TAPR
site, and there's no link to it from the TAPR site.
http://www.xastir.org has links to both of them, plus a link to the
OpenTrac spec.

-- 
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!"