Continuing the discussion from New at 'Desktop On Fire!": I’ve got two old ARM7 devices that are barely still useful due to the age of the software running on them. My CuBox i4Pro has barely been used, is a perfectly usable device with good workmanship and comes standard with Bluetooth and WiFi hardware. My ODroid XU4 has no on-board wireless support but runs a third-party Android 7.1 install (LineageOS) like a champ.
The Meshcore Flasher appears to install closed source code in its repeater, has a paid model and my hardware is unsupported. Meshtastic’s repeater is open-source and free.
In the event of an actual disaster, the likelihood of an accompanying economic collapse will render Meshcore unusable while free solutions will continue as long as decentralized power consumption is available. Knowing this, what would it take to add Meshcore-style features to Meshtastic’s repeater? Also, is there anything about the repeater that would prevent it from working on an old ARM7 machine?
Edit
Meshcore’s web client only supports Chrome and derivatives, making desktop usage on any operating system that doesn’t have Chrome require a custom app. Recently, in the thread this was spawned from, a custom app was written for Haiku but this would have to be repeated for every operating system that doesn’t support Google’s browser APIs. UGH!
Edit2
I2P seems to have economic independence also if encryption needs to be added to Meshtastic, especially if the same governments that were funding Tor go belly-up at the same time as Meshcore. All encryption software needs some beefy CPU time but I2P is at least FOSS all the way through.
Just a small clarification: MeshCore is released under the MIT license — it’s fully open source. The source code is freely available. Only certain specific firmware builds/features are behind a paywall, and it’s really more of a small contribution to support the solo developer behind the project than any real commercial barrier.
On the Chrome-only web client point: that’s a fair criticism, but the community is already working on alternative clients. As for Haiku, well, I’m the one who wrote that “Sestriere” client you’re talking about — so I can confirm firsthand that the open protocol makes it perfectly possible to build native clients without too much trouble!
That said, let’s not make this a religion: MeshCore is certainly younger than Meshtastic and has different characteristics, each has its own strengths. But there’s a certain irony in wanting to build a network designed to work without the internet and then relying heavily on MQTT for data transmission, right?
Maybe something like esptool could be used to flash the devices? We do have esputil in HaikuDepot. I do have 3 Heltec LoRa32 V3’s on their way here. Hopefully the will make it out here by next week.
I don’t think you mean arm7, maybe you mean arm v7 (ARM ISA version 7) or arm cortex A7 (a 32 bit ARM core)? Arm7 are 32 bit ARM cores, but really old, typically less than 100 MHz clock speeds. ARM7 - Wikipedia
As far as the religious implications, I’m prepping for the last days and believe we’re already in the second of the 4 horsemen of the apocalypse. The black horse is a financial famine because the only one of the 5 commodities listed in those 2 verses in Revelation 6 that are measured by mass rather than volume, is the silver denarius: a day’s wage. Let’s keep the balance in the black.
Back on the Topic of Code
If the client were written using the existing source and compiled into WebAssembly, then any fully-featured web browser could run the existing code without needing an actual port! I’ll look into the specs later because if it isn’t implemented in a modern language, my old hardware might not run it parallel enough to be useful.
Hardware Specs of my Old 32-bit Devices
The devices I listed are 32-bit ARM. They each have 2 GB of RAM. They have LPDDR3 because that was all that was available 13 years ago. My main system is a 64-bit, 3rd-gen Intel i7 with 32 GB of DDR3 in a desktop case. I don’t use anything newer because that’s all I need for now. My hardware is always obsolete just like a new car becomes a used car as soon as you drive it home from the dealer. Why waste money on new stuff if the old stuff works well enough?
That looks very interesting.
I’ve read your announcement of the application already,but now with much more background information,I’m really considering to buy a LoRa device and give it a try.
I wonder if there’s anyone I could message within the range.
Probably not,so that would be a show-stopper,but still very interesting project.
The great thing is that if there’s no one today, there will be someone tomorrow—yours! Jokes aside, there are plenty of websites that show maps of public nodes. For example, in Italy, there’s https://nodi.meshcoreitalia.it/, which also displays some of the German nodes.
The Italian map shows only nodes in the south-west part of Germany and stops at Heilbronn,which is still 50km from me.
The map from the MeshCore project shows nodes around me which are all maybe like 20km away from me,but none directly in my city.
Maybe that would work,the chance is pretty low I guess.
50km is a lot. But it mostly depends on where you can install your antenna. At street level, a companion [device] reaches 2km at most if you have apartment buildings or houses around you. It’s a different story if you have the chance to be higher up and can ‘see’ the other repeater, even at 10 or 20km.
My position here is rather good,so it may even make sense to run a router here.
I live in the upper floor of a high building and have a perfect sight over the whole city.
Anyway,that won’t help with other routers in other cities as there are mountains inbetween.
The MeshCore experience is really cool. I couldn’t see anyone on the map at first, but once I turned on my node, I got in touch with others who were also setting theirs up or thinking about it. I can’t quite get a direct connection with them yet since I’m only on the second floor, but from your position, you could probably cover half the city. Maybe down the road, someone else will get curious and install a node just because they see someone is already active. The important thing is just to start; the rest follows naturally. The best part is that they draw so few watts that they don’t even impact the household electric bill.
This is a very interesting project to me. I worked in radio electronics for over 25 years, and have always had an interest in amateur radio, etc. I think I’m going to order a couple of the Lora32 devices. I don’t see anything on the map in my area, but like you mentioned, they could be out there, and if not, someone might see mine and get interested.
The MQTT bridge is definitely something I’d like to look at too. I already have an ESP32 project that I have breadboarded and programmed that will send sensor information from my boat, over MQTT using WiFi, to a dashboard on my phone. I have an MQTT broker on a cheap VPS server. It would be interesting to see if I can merge the two.
Ha. You hit a topic close to my heart. I run several Meshtastic relay nodes and travel with a Meshtastic client.
I want to like Meshcore, however when there are 100’s of meshtastic nodes within 100 miles of me, and only a few Meshcore, it’s hard to justify dedicating hardware to it.
Mix in the paid license for Meshcore, and the rivalry between them, i’ve been sticking with Meshtastic for now.
Thanks for the heads up with rivalry and paid licenses,that doesn’t speak for Meshcore.
I was almost purchasing a device,now I think it’s probably better that I didn’t.
The behavior of the Meshcore maintainers in your feature request is really childish.
After all,there aren’t so many people participating in that network and it can’t work without enough people,so working together would be a logical choice to improve the network for everyone.
Maybe someone needs to create a third project that is compatible with both others.
To be clear, Meshcore is absolutely the better design because of the routing. Meshtastic works well due to it’s simplicity, but at scale it doesn’t really function in a usable way (besides seeing messages go *miles* over Lora over multiple hops and going “neat”)
Yeah, it was kind of my red flag. A quick “hey, it’s technically difficult and we don’t have the people to implement it” would have been fine.
I’d still recommend purchasing a device to be honest. The LoRA modems are cool, and can be used for a wide range of things.
I’m honestly digging the idea of https://reticulum.network , it’s just at too early of a stage to see widely implemented.
I do agree that the response from the devs to your request was strange indeed. Your idea does make a lot of sense. It should be possible for someone else to modify the code and create a custom firmware for the repeater that would rebroadcast the Meshtastic chirps.
Anyway, I have 3 Heltec Lora32 V3 on their way from Amazon. Probably won’t make it our here until next week though.
Joking aside, I could tell you that residency is useless—you will all be assimilated! I’m at the construction site in Venice today; let’s see if the kids will let me get back to you properly tonight.
I’ll look into the specs later because if it isn’t implemented in a modern language, my old hardware might not run it parallel enough to be useful.
Jokes aside, there are plenty of websites that show maps of public nodes. For example, in Italy, there’s 
I’m at the construction site in Venice today; let’s see if the kids will let me get back to you properly tonight.