Headless aprx APRS server

In earlier articles I documented a trial of aprx v2.08r593:

Encouraged by Kenneth Finnegan who merged some fixes by WB4XO (which still didn't work properly), I set about resolving some problems.

HeadlessJavaprssrvrAbove is the RPi sitting on a TNC-X KISS TNC connected to the TNC's USB port. In this case, aprx uses the supplied KISS interface, it does not use Linux kernel support for AX.25.

Don't take this as a recommendation for TNC-X, it is only a partial KISS implementation.

WB4XO had attached the digi problems, but it didn't fully address the defects.

Further fixes including to the digi function, KISS/SMACK negotiation and logging have been made and pushed to the source fork at owenduffy/aprx. This was forked from PhirePhly/aprx which in time may pull the changes.

The digi now works better, and filtering seems to work to the limited extent tested.

The conclusion and commencement of this second trial is that aprx may be the basis of an intelligent digi that could facilitate fixing the woes of APRS caused by enslavement to source routing.

Trial is of direct only digipeating and rx iGate, so the fact that the TNC-X does not reliably decode VK2RHR-1 (possibly because VK2RHR-1 is grossly overdriven) is not of great consequence as far as iGate is concerned though it is inclined to transmit over VK2RHR-1. (The TNC-X does decode my own OT-USB which is properly adjusted.)

The trial may progress to rx-only iGate as there is little point digipeating packets that have been iGated, indeed not transmitting reduces channel packet loss.

I have a deb package for those who might like to try my revised aprx on RPi.

More as it unfolds…

Update 23/09/14: The MFJ-1270B with SMACK v1.3 has been installed for some statistical studies and extended proving of the SMACK capability.