Is APRS Broken?

As I wrote yesterday, I have been experimenting with APRS on the HF bands (30 metres.) This morning I spotted a couple of mobile stations beaconing position reports, or more likely I was receiving the digipeated copies of them. I can see that position reporting on HF could be useful if you want to be tracked and are out of range of any VHF digipeaters or gateways, but HF stations with their much larger capture area and four times slower data rate really couldn't sustain many fast-moving mobiles sending position updates every minute or so.

APRS is meant to be more than just a system for capturing position reports using radio and I find the facility to exchange messages with ham friends and know that they received them (even if they aren't immediately able to respond) to be very useful. It isn't a substitute for conventional digital contacts, it's an additional way of communicating. However, unlike on VHF where you probably know the people whose calls you see on your screen, on HF there is no way of knowing whether someone wants to chat, or even if a particular station is attended. Unless the intention is to chat direct by radio, using HF seems to me to be an inefficient way of reaching the APRS-IS internet backbone. So apart from giving someone in the middle of nowhere an extra chance that their packets will be received - which would be a rather boring use of an expensive HF radio - I'm still unsure of the value to me personally of running an HF APRS station. The technology is interesting but the practical use eludes me.

Despite this I was still keen to try APRS over HF. As I didn't know whom to contact I decided to send a greeting to Lynn, KJ4ERJ, currently holidaying in Spain, who I knew (from checking aprs.fi) was online at that moment. I disconnected my APRSIS32 client from the internet so I would know that if my message was received, it would have been picked up on the radio.

The message went out on 30m, and was repeated several times as no acknowledgements were received by my station. In the meantime I decided to send a second greeting message to Colin, 2E0XSD. No acks were received for that message either. Eventually I checked aprs.fi and sure enough my messages had made it to APRS-IS through various gateways in Switzerland and France. They had even made it to their destinations and Lynn and Colin had both sent replies. But I never received their replies over the radio, either.

The screengrab shows a section of the raw messages list from aprs.fi which includes the paths of my messages as they were received on the internet. It appears that nearly all were received and digipeated by HB9MM-4. I'm not sure if that station also gated my messages to the internet.

The way I understand APRS to work, any messages sent to me (and any acknowledgements of messages sent by me) should be relayed back to me over RF by any stations that have heard me on RF. The acknowledgements and replies from Lynn and Colin were on APRS-IS and could be received by the HF stations that were hearing me. But they were not transmitted. I was receiving a strong signal from HB9MM-4 so it is very unlikely that the packets were sent and none of them were decoded.

What's the use of a system for messaging if it can't be relied on to work as intended? It seems to me that APRS is broken and is really only useful for collecting position reports where the traffic is all one way - to the internet.