aprs at kd4rdb.com
Sun Jan 8 09:13:08 CST 2006
I have seen similar problems with my tracker moving around town. I came
to the conclusion that UI-View acting as an IGATE was delaying delivery
of packets to TCP. I even wrote a small utility program to compare
positions of my station to later reported positions of my station.
Since during the testing period I was moving, each packet was unique for
sure. I found that in each case a ui-view station would deliver a
packet late anywhere from 120seconds to 500 seconds. It was suggested
that the guy's internet connection was queueing packets but I found this
hard to believe. I spoke with the guy who's copy of uiview was the
problem and there was nothing particularly unusual about his DSL
Now just to be fair about all of this.... I can't say for sure that
ui-view is the problem... but everything and everyway I've looked at
this points me that way. And of course we have no way to dig around in
the sources to investigate. That leaves me with the "If it quacks like
a duck" cliche.
> I was going to make the subject "out of order packets" again, but this is a
> bit different. I'm used to seeing mobile stations (in the Vancouver BC
> backtracking from time to time. There has been lots of speculation about
> cause, often with the mention of the possibility of digis with radios that
> had the squelch set too loose so the TNC would hold onto packets until the
> squelch closed. I happened to be watching VE7CUU-14 driving around earlier
> today in Squamish BC and noticed some wild excursions back in time that
> far more extreme than anything that I had seen before. I checked the raw
> data at findu.com and discovered that as well as being gated by his own
> VE7CUU, some beacons were being "re-gated" by DG0ES. This station is
> in Germany with a TCPIP connection. Why on earth would a station in Germany
> be "re-gating" a station that is in Squamish?????
> I was viewing VE7CUU-14's track with a TCPIP feed.
> See www.findu.com/cgi-bin/raw.cgi?call=ve7cuu-14&time=1
> The first position gated by DG0ES was at 2006 01 07 1836 32 UTC. Examining
> the mic-e payload, I could see that it was originally gated by VE7CUU's own
> IGate at 2006 01 07 1824 16 UTC - 12 minutes earlier. In another
> instance, a
> packet "re-gated" by DG0ES was delayed by just 3 minutes. First, DG0ES
> shouldn't be gating a station that is half a world away, but does the
> APRS-IS need to be adjusted to reject dupes from even earlier than it
> currently does?
> I sent an APRS message to DG0ES, but didn't receive an ack. I checked after
> sending the message and DG0ES hadn't been heard in the previous two hours.
> This question was posed on the NWAPRS SIG earlier today.
> 73 es cul - Keith VE7GDH
> "I may be lost, but I know exactly where I am!"
> aprssig mailing list
> aprssig at lists.tapr.org
More information about the aprssig