It seems to reinvent parts of 802.15.4 but not especially better. Is there some writeup on where M17 stands in the larger low data rate radio ecosystem?

It's not really meant for the same applications as 802.15.4. it competes with DMR, DStar, and C4FM for narrow band digital voice over very long distances (single links can exceed 100mi easily).

What it solves relative to its competitors is having an open voice codec instead of relying on a very expensive silicon encoder/decoder chip while also having greater flexibility and performance.

Also, if we just look at data capabilities, Meshtastic is a much more interesting comparison.

Hadn’t heard of M17 before. Thx for sharing.

I don't know about that. I don't know anyone, at least here in the UK or Europe, who even uses digital anything other than mashing in front of WSJT-X all day. On 2m/70cm, it's all voice still.

DMR/DStar is fairly popular where I am, though it's mainly through repeaters.

Also, by far the most heavily trafficked frequency anywhere I've been is APRS which is digital store and forward messaging. It's just extremely antiquated at this point.

From what I can see APRS is mostly used by porcupine vehicles broadcasting where they are. Literally because no one cares, it's the opposite of what what Musk wanted for his jet :)

That’s a pretty narrow perspective. From both the perspective of HF and VHF/UHF.

For HF there’s all sorts of digital stuff going on. Check out FreeDV for a Codec2-derived digital voice mode. We are also doing packet radio (yes, AX.25, still built into the Linux kernel) with BPSK/QPSK modems on HF replacing the 61 year old Bell tones- see https://wiki.oarc.uk/packet:ninotnc#experimental_new_modes_f...

I recently stood up this site to support and track the re-emergence of the / redevelopment UK packet radio network- which is utilising VHF, UHF, a geostationary satellite, as well as HF. https://ukpacketradio.network/

It’s not my bag but I see a lot of VARA going on too.

There’s a lot more going on than just WSJT-X.

Yeah there's stuff there but there are literally bugger all people using it. Even the VHF repeater nets in the UK are pretty dead. I mean I'm in the middle of London and I occasionally get two dudes talking about their greenhouse and that's it. They're not even growing anything interesting - it's all tomatoes and casual racism.

While I'll say M17 looks very interesting - I'm far more interested in analog audio, with both sub audible data steams and leading/trailing data burts. You could easily cram a 50 bits per second data stream into the space below and above the audio passband, for example, giving you things like live GPS.

I have a good ear, and can pick people out of noise, and I think most Hams have that too. I'm less concerned however with the need for spectral efficiency - we have vast amounts of lightly used spectrum, ultra-narrowband is not really needed for us - and for digital specifically, I'm much more interested in modulation schemes that are durable - and will work in very low signal situations (like being able to sustain traffic at lower than -100dBm).

My gripe with digital in the end, is its all or nothing, and often in weak signal, it's nothing. I say this as someone who works in the two way radio world for my day job.

That all said, I've had another unrelated idea for years of being able to build multi-band wide area trunked and non-trunked systems with roaming.

The first part of it is a beacon operating in every area this network exists, I've taken to calling it a localizer - but basically it'd broadcast a band plan on a handful of known frequencies (say allocating 8-10 fixed frequencies in each band, with the idea that only one per band would be used in a given area), with frequencies, which talkgroup ID's they carry, what operating mode they operate with (digital of one or more flavors, analog, etc), where the local control channels are for any trunked systems, what talkgroup ID's those systems carry, what area that localizer covers (a bounded box for example), and what the adjacent localizers are.

So when you turn on your radio and put it in this operating mode, it's going to scan the known frequencies for beacons, then auto-program the local RF properties in.

From there you can chose a channel (your radio would auto-program from the localizer), and you can chose that 'channel' - if its trunked, you'd register to that system, but for conventional analog or digital, you would just tune into it. For systems using dial on demand talkgroups, you could just request that talkgroup be served to you as part of your registration.

The idea is to create a multi-band, multi-mode radio system, that anyone could use with the correct radio. But because it would also include conventional analog repeaters, anyone with any radio made in the last ~60 years could still participate.

If any of this sounds familiar, its because some of it is borrowed from Nextel - its not exactly how Nextel worked, but iDEN didn't have a consistent band plan from city to city so when you'd turn your subscriber on, it'd scan for control channels, then register to one, and as part of that process, you'd be sent the band plan for that market (because Nextel was not built out of a national allocation, but rather individual SMR pairs which varied from market to market).

I'm also really interested in medium bandwidth mesh radio applications (ala Ricochet).

Okay, but one requires reading pages of specifications, and the other would fit on a sticky-note.

Ideally it's a firmware update on your radio, no sticky note needed