[Xastir] Weather Wanings Comment Text

[Tom Russo]

On Thu, Dec 20, 2012 at 07:39:12AM -0800, we recorded a bogon-computron collision of the <aa9vi at yahoo.com> flavor, containing:
> Curt, the fact that LOTWSW is originating from my station seems correct to me and consistent with what was done in the past with the?xastir/data/nws-stations.txt file. ?My understanding is Xastir pulls these warnings that you specify in that file and broadcasts an object over RF and the internet when these warnings take effect. ?I like the feature and it is quite useful to people in the area. ?

No, this is not entirely correct.

When you put something in .xastir/data/nws-station.txt, it allows Xastir
to gate those objects from APRS-IS to RF as third party packets.  Third
party packets come from your station, but are not interpreted as *originating*
from your station.  The LOTWSW object from January (the last time it was 
heard on APRS-IS and therefore aprs.fi) is showing as originating from
your station, which is wrong.

Third party packets that have been plucked from APRS-IS and gated to RF should 
never then be re-gated back to APRS-IS by an IGate.  All compliant igate 
software will decline to upload packets that have already passed through
APRS-IS (otherwise there'd be serious looping problems).

Xastir does not munge the objects from APRS-IS before it gates them as 
third party.  It adds third party headers and then just transmits.

Therefore, the old LOTWSW packet we're seeing on APRS-IS has almost
certainly been improperly gated back to APRS-IS after you transmitted it as 
third party traffic.

Are you actually seeing your station transmitting LOTWSW objects to RF right 
now that are truncated?  Or are you just looking on aprs.fi to see the object?

If you're only looking at aprs.fi, be aware that the single 
object that's recorded there is almost a year old, and may have been gated 
back to APRS-IS by a misconfigured igate.  The truncation may well have 
happened there, not in your station.  In fact, I'm close to certain of it.

> Just want to fix the comment truncation a bit. 

First be absolutely certain you actually *have* a comment truncation problem.
Remember that lots of people have used Xastir to gate weather alerts and
objects back to RF, and a glaring error like the truncation of almost the
entire comment field (which usually also contains information to allow
plotting of the area to which the object is connected) would have been noticed 
before.

Xastir should be retransmitting the object it hears from APRS-IS with no
modification whatsoever, and in my experience it does so.  To check this, 
don't look at APRS-IS, because your transmitted third party traffic should 
NEVER make it back there unless you have a broken local igate, and if it 
does you can't trust the gate that put it there.

Watch your outgoing packets window for packets going out to your TNC.  If you 
see truncated packets in the outgoing object there, then  make certain that 
the truncation is happening locally --- i.e., compare the outgoing third party 
packet with the incoming weather object from LOTWSW.    Unless you see that
Xastir has received an unmolested object from LOTWSW and then transmitted it
to RF completely mangled, the problem is elsewhere.

> Do you think the object should not have an originating ham station?

The third party packet originates from your station, but the actual payload
originates from LOTWSW, and properly written APRS clients read them that way.
Basically, the third-party format encapsulates the entire packet from the 
originating station (including originating call sign) and wraps them in an 
extra bit of data that shows the *transmitting* (in addition to the
originating) station.

That there is a single packet stored at aprs.fi from a year ago with your
station showing as the originator suggests that some other igate may have
stripped the third-party header the wrong way and ignored the fact that the 
original had already come from APRS-IS.  Doing so, it probably gated the 
packet back to APRS-IS as if it came from you, not the originating station.  
It could have done this due to a bug, or there could have been some sort of 
corrupted packet issue that confused it.  That there is only one recorded 
position in the aprs.fi database further suggests a fluke rather than a 
completely broken igate.


-- 
Tom Russo    KM5VY   SAR502   DM64ux          http://www.swcp.com/~russo/
Tijeras, NM  QRPL#1592 K2#398  SOC#236        http://kevan.org/brain.cgi?DDTNM
 echo "prpv_a'rfg_cnf_har_cvcr" | sed -e 's/_/ /g' | tr [a-m][n-z] [n-z][a-m]