[aprssig] odd siting in these here parts

[andrewemt]

I've seen these sorts of messages before, and they seem to be due to buffer overruns in some station's receive code. I had horrible overrun issues when first writing an APRS client due to the highly inefficient API's in Microsoft Windows and then Java on top of it, where I actually lost sequences of characters between adjacent frames on a busy channel, such that the end of one frame and the beginning of the next was lost.

This shouldn't be an issue for I-gates, because TCP/IP networking can back-pressure over-fast senders; it's generally only a problem for RF receivers in poorly written clients on slow computers. Are there any really badly-written I-gates with buffer overrun problems (or I-gate clients who use non-blocking sockets to busy APRS-IS servers)?

Andrew, KA2DDO

-------- Original message --------
From: Keith VE7GDH <ve7gdh at rac.ca>
Date:09/14/2014  13:50  (GMT-05:00)
To: TAPR APRS Mailing List <aprssig at tapr.org>
Cc:
Subject: Re: [aprssig] odd siting in these here parts

Steve K4HG wrote…

> It is not unusual to see merged packets like this on the APRS Internet System. This particular packet also appears on findU, so it is not a problem in aprs.fi

I should have looked there! Yes, if you are seeing it there too, the XASTIR Raspberry PI IGate must be responsible for mangling the packet.

--
73 Keith VE7GDH
UI-View32
www.ui-view.net


_______________________________________________
aprssig mailing list
aprssig at tapr.org
http://www.tapr.org/mailman/listinfo/aprssig
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/aprssig_lists.tapr.org/attachments/20140914/f7b32657/attachment.html>