开云体育

Hi all,

I am currently on a road trip through southwest US.
Family and friends us APRS to look where I am.

Lately they told me : your track has some weird jumps and cros crosses.

I analyzed my data and came out to the conclusion that all of these issues are caused by sur based igates running outdated direwolf.

The symptoms are outdated packets are sent to APRS-IS sometimes hours after they were originally transmitted.
I had the case were some of my packets were resent to APRS-IS 3 hours after their original transmission.



--
Geoffrey F4FXL /? KC3FRA
My Direwolf stuff :

Are you sure these packets weren't digipeated by the infamous buggy Kantronics KPC-3, which also had the habit of delaying digipeats? If the digipeater delayed the packets before it got to the I-gate, it wouldn't matter what the I-gate tech was.

Andrew, KA2DDO

________________________________________
From: [email protected] <[email protected]> on behalf of Geoffrey Merck <f4fxl@...>
Sent: Friday, May 3, 2024 10:35 AM
To: [email protected]
Subject: [direwolf] Buggy SDR IGates....

Hi all,

I am currently on a road trip through southwest US.
Family and friends us APRS to look where I am.

Lately they told me : your track has some weird jumps and cros crosses.

I analyzed my data and came out to the conclusion that all of these issues are caused by sur based igates running outdated direwolf.

The symptoms are outdated packets are sent to APRS-IS sometimes hours after they were originally transmitted.
I had the case were some of my packets were resent to APRS-IS 3 hours after their original transmission.



--
Geoffrey F4FXL / KC3FRA
My Direwolf stuff :

On Fri, May 3, 2024 at 10:35 AM, Geoffrey Merck wrote:

I analyzed my data and came out to the conclusion that all of these issues are caused by sur based igates running outdated direwolf.

What evidence caused you to arrive at that theory?

73,
John WB2OSZ

On Fri, May 3, 2024 at 07:43 AM, Andrew P. wrote:

Are you sure these packets weren't digipeated by the infamous buggy Kantronics
KPC-3, which also had the habit of delaying digipeats?

Many KPC-3Plus TNC still have this issue today, because it is not widely known that the KISS-mode Buffer Bug was fixed by Kantronics in a 2022 firmware update (it doesn't help that the new firmware is still called 9.1, you have to go by the date).

The problem did not effect the KPC-3Plus if it was doing the digipeating internally (the serial interface was in Newuser or Terminal mode) but many APRS I-gates connect to the TNC in KISS-mode.

This may be the wrong forum to spread the word, but if you operate a KPC-3Plus (any variety: original or MX or USB) in KISS-mode then you want to make sure that your firmware is up to date. No need to upgrade if you are not using KISS-mode and you are already on an earlier 9.1 version.

73,
Thomas
KK6FPP

Here are a couple places where we see the position jump abruptly, apparently by an older position report being delayed.

?

?

2024-05-02 12:06:34 EDT: >OPITRK,,,WIDE2,qAO,:=/<'I,3-3s>NVG/A=006469_Geoffrey,?QRV?Voice?Alert?100.0Hz [Duplicate position packet]
2024-05-02 12:06:55 EDT: >OPITRK,,WIDE2-1,qAR,:=/<)5X3-4$>NVG/A=006368_Geoffrey,?QRV?Voice?Alert?100.0Hz [Location changes too fast (adaptive limit)]

…

2024-05-02 17:47:07 EDT: >OPITRK,,WIDE1,,WIDE2,qAO,:=/<0,j3-3]>K?G/A=005782_Geoffrey,?QRV?Voice?Alert?100.0Hz

2024-05-02 17:48:32 EDT: >OPITRK,,WIDE1*,WIDE2-1,qAR,:=/<VL#3+h^>jNG/A=007289_ [Location changes too fast (adaptive limit)]

In the delayed packet cases,? digipeater and IGate are involved.

The IGate, , is running Direwolf 1.5.

Digipeater ? is running a Kantronics KPC-3+ with ROM version 8.2 from the previous century.


There are many other cases of the same thing where direwolf was not involved at all.
In all cases, we find digipeaters using the KPC-3+ with very old ROM versions


There is a known problem with older KPC-3+ TNCs.?? ?

?

73,

John WB2OSZ

Fair enough, might be also a KPC3 issue....? I stand corrected.

Maybe I was biased because a couple of month a go, in the rea where I live,? we had an SDR IGate stuck in a sort of loop and it kept replaying the same set of 30 minutes of data.

--
Geoffrey F4FXL /? KC3FRA
My Direwolf stuff :

Hi Again,

What I thought was an occurrence of this SDR issue was actually caused by old KPC3 devices.

Anyway,? I was able to spot the actual SDR issue two times this week


First occurrence:


Frame was sent at 09:43:43 and gated by F5ZEE, the same frame was also digipeated at 09:44:55 (almost a minute after the original one) by DB0TFM-1 and DB0ODE-10 and finally gated by DC1SHM-10 which? is running an SDR igate.


Second occurrence

Original Frame sent a 17:09:40, first time gated by DH1GHL and delayed gated DB0SEL-14 at 17:10:18

Until now, I only observed the issue when 2 or more hops are involved. A possible cause might be the delays introduced by the 2 hops added to delays in the SDR chain (caused by buffering and computation power) are causing packets to be delayed and falling outside the duplicate detection window.

73

Geoffrey


--
Geoffrey F4FXL /? KC3FRA
My Direwolf stuff :