Meshtastic: A Review | Hacker News Reader

Meshtastic: A Review

tylercipriani.com

101 pointsthcipriani39 comments

Would love to see a power efficient device streamlined for texting with this. A smartphone is so overkill and battery wont last that long.

Would also be nice not to connect with bluetooth. Either connect with a cable, make it an all-in-one texter/transmitter, or make it all-in-one but detachable. I guess this is starting to reach, but bluetooth is just too finicky for something with this use case IMO. I can see the value in wireless pairing though - you could put the transmitter up a bit higher and text on the ground.

There is also the Armachat: https://hackaday.io/project/171790-armawatch-armachat-long-r...

That’s cool.

Looks like they don’t encrypt, though.

Here's a really interesting platform:

https://pocket.popcorncomputer.com/

Not much activity during last year, however.

This is a commercial version of the concept. They have been doing it for a while: https://gotenna.com/

The concept is of interest to me, as I like writing code to help people that help people. I’ll bet I could write a really nice Apple SDK and iOS app. I noticed the reviewer wasn’t too thrilled with their iOS app.

I did try working with goTenna, but they won’t open up their low-level API, and I don’t like using closed-source SDKs (that other people wrote —I have issues), for the most part.

I’m having difficulty with their documentation. It seems a bit scattershot. I’ll keep looking, but it appears as if there isn’t a place that simply lists the interface parameters. I would probably use BLE, and I suspect they do serial over BLE (I think goTenna does the same). Not a problem, as long as the AT commands are listed and explained, somewhere.

If you wanted to work on Meshtastic instead, everything is fully open and the community is nice and helpful. There's an iOS app too...

I still haven't found an AT command manifest.

It appears as if we are supposed to figure out what various commands are, by reading the firmware spec, at different parts of the docs.

Without that unified AT command reference, I'm not even gonna try (for now, at least). My dance card is too full, to spend a bunch of time, with my BLE explorer, testing out commands.

If they published it, I’d probably scare up a cross-[Apple]-platform BLE-based SDK. I already have the BLE part written[0]. This is used in my BLE Explorer App[1], or the test harnesses could be used to explore the device. It would just be a matter of writing the command interpreter.

[0] https://github.com/RiftValleySoftware/RVS_BlueThoth

[1] https://riftvalleysoftware.com/work/ios-apps/bluevanclef/

I'm asking if we have one already (even if it's someones napkin notes).

Heh, I had the same complaint about the gotenna low-level APIs, and apps using their SDK being second class citizens in their environment.

I wanted to enable global delivery by having gateways that would join a DHT and help different mesh's to have global delivery with via a raspberry pi or similar gateway. That received a hard no on their mailing list. Was even considering using the bittorrent DHT, there's a BEP to use up to 1KB of storage on DHT peers.

A few months later I saw they were offering global delivery to their enterprise customers.

Interesting - With how cheap hardware is getting, would there be some potential use case for backcountry hiking/biking/adventuring with a few of these, and leaving them as breadcrumbs/mesh repeaters?

It would allow a cheaper emergency connection to society than satphones, and could potentially allow for some waypoint following with modern gps trail navigating phone apps.

Curious if anyone has experience with Meshtastic and/or outdoor adventuring and could weigh in.

I like the idea of a mesh network but anyone considering doing something like this, please be super aware of fire risks and rather don't do anything if you are not 100% sure that what you are doing is 100% safe.

Yup, you can even get kits that come in waterproof enclosures with solar cells. Please mark your nodes as radio equipment and with your contact info on them.

If you use the default meshtastic build you can forego a screen and use the cli to tweak the node and even disable it's gps for better battery.

There is work happening on a dedicated "repeater" codebase too.

What's the best online space to catch up with/follow progress?

Development is on github: https://github.com/meshtastic

The discord is pretty active: https://discord.gg/ktMAKGBnBs

Can you link one of these waterproof+solar kits?

Disclaimer, I know this seller through the Meshtastic discord.

https://www.etsy.com/listing/1048791528/solar-powered-off-gr...

Nice review! I too wish for a solarpunk future with mesh communications infrastructure. I see a Pine64 box in your pictures, will you try out PineRadio in the future?

> The T-Beams run off big honkin’ 18650 batteries—the same lithium-ion cells used in Tesla battery packs.

While this is true, it seems unnecessarily alarming. 18650 are also the type of cells inside old-ish laptop removable batteries, which are far more common and don't frighten people.

My cultural touchpoint is vapes and flashlights.

"Yes, they literally wire 'em together to power a lawn mower."

https://youtu.be/L_VManypHV4

18650s are quite pervasive in large capacity external battery packs and there are a plethora of battery chemistry varieties too, all depending on specific use case.

I have the Heltec LoRa little dev ESP32 with it installed but no one to talk to yet haha. The display on it makes it a decent solution though.

Back in the day I really enjoyed using the Ricochet wireless modems which allowed up to 128k BPS communication, link to link.

I wonder if Meshtastic can bring back some of that experience, and be used as a way to access my own bandwidth from miles away? Anyone know if SLIP/PPP is usable over LoRA?

Cool concept.

I don't know what it is, but from the article it appears to be a secondary device (Meshtastic) you pair with your iPhone/Android phone via Bluetooth. This secondary device has the giant Lithium cells and microcontroller with the LoRa antenna setup.

Some software package on the phone, perhaps a messaging like app, sends message data (AES encrypted) between your phone and the Meshtastic device. The Meshtastic device somehow knows which other Meshtastic devices to send/receive from?

Feels like the phone shouldn't even be needed is a simple LCD display and tiny keyboard were added to the Meshtastic? (And I suppose the en(de)cryption would have to be done on the device now.

It's bridging to digital traffic over long range radio.

The phone provides a familar interface AND the GPS cords for texting locations (and possibly pinning on maps already present on phone when WEB-Maps are offline (google, bing, other, etc).

Once matured it will have some application for hikers on trails away from cell tower coverage sans Starlink access.

I'm used to rural fire brigade trucks and SES rescue vehicles that have a slew of standalone radio gear and multiple GPS units already fitted .. but I can see some scope for this when looking for lost tourists and city folk.

Some of the LoRa boards do have built-in GPS so you do not have to rely on the phone's GPS. It also means accurate time, so you can display how long ago a message was received. When I tried meshtastic, the LCD had the option to point towards other receivers.

It also took 20 minutes for first GPS sync, which was less fun.

> It also took 20 minutes for first GPS sync, which was less fun.

This is true in a common base case (LilyGO T-Beam neo6m with stock antenna) but is easily fixed. The stock antennas are absolutely pathetic and something like https://www.mouser.com/c/?q=GP.1575.25.4.A.02 is much much better. The other significant improvement is using a board with a better GPS module, like the M8N version of the T-Beam.

This is correct. It just sends all messages via LORA, which nearby receivers can pick up. These will broadcast it again (with a initial TTL of 3).

"Meshtastic" is a perfect example for the observation that as long as the first and last few letters of a word are in the correct order, the rest can be scrambled and your brain can still process the word normally. Unfortunately my brain read "Metastatic" (as in cancer) instead of "Meshtastic"...

If you want a device but don't want to have to buy batteries or make a case, there are several sellers on etsy who have pre-finished units.

While the T-beams are nice for having screens and gps, the RAK based devices have better battery life.

What is the range of this thing? How dense meshastic devices have to be around us for this to be usable?

Between 2km (dense urban area) and up to 100km, usually probably around 10km. It is basically a replacement for CB radio, and works similarly. The Meshtastic devices are set for a specific (chat-) channel, and only mesh with devices which have the same configuration/channel.

My experience is about 10% of that, km-range in really good conditions. LoRa is low power, long range wifi protocol, with low bandwidth, designed for sensors running on a battery.

10% of 2km? I feel like you had something set up wrong if you were trying to go long range. I did a basic test with a mestastic ttgo T-Beam, with the stock antenna no modifications and was getting 2miles (about 3km) through dense forest. Out in the open it was much longer but I didn't test it fully. Anyway I was super impressed, it might be worth trying again.

> In testing, LoRa works up to a few miles away with a good line of sight.

Why not stick with AX.25 and packet radio instead of reinventing the wheel?

Encryption is one of the stated goals, which is illegal for packet radio in most jurisdictions. Also requiring a licence to transmit (no matter how easy/cheap it is to obtain) is a barrier to adoption that must be considered.

AX.25 is anything but modern. No FEC, floods easily and has other constraints.

I don't know why but there doesn't really seem to be a push to so better on FM digital modes compared to what you see with FT-8.

VARA FM offers high speeds on FM from what I’ve seen. Haven’t actually played with it myself though. Might be worth checking out!

I play around with APRS here and there, and you either have a cheap setup, like a radio plugged into your smartphone which is bulky and messy with wires hanging around. Or you shell out a large amount of money for a (still bulky) radio that can do APRS natively.

It feels like outdated tech to me, all the hardware is generally big, slow, power hungry, and expensive.

The APRS system also has issues with timing and flooding a channel if there are too many stations in one area.

Not sure what you mean by bulky, but these seem pretty small to me: https://www.wimo.com/en/picoaprs

Seems pretty common to put similar bare boards in people playing with high altitude balloons.

Sure APRS has limits, but keep in mind that you don't have to use the "APRS" frequencies, for high density uses like say a bike/horse/foot race it's pretty common for support teams to have APRS to help coordinate and to use a non-standard frequency.

So tracking 1000s of devices second by second, sure that's not going to work well.

Reply on news.ycombinator.com