Posted by no_creativity_ 1/19/2026
* I see LocalSend and LANDrop frequently suggested on HN but in my experience they rely on having a central Wifi router. No good.
* Android's QuickShare comes included by default, but it's buggy. Just yesterday it failed on me (I'm on an uncommunicated boat): it was defaulting to Bluetooth, so I had to reboot both phones to finally make it work over Wifi Direct. Not to speak about the "oh damn, you have an iPhone" scenario. Not ideal.
Anything else? (to remark: for airplane-like situations so no access to Internet and no central router)
Cross test with Meshcore doesn't show any issues. Chats over 5 hops have almost a 100% success rate.
Long time I avoided MC because of its closed source client - but a Opensource Flutter app for Apple/iPhone is slowly getting usable and stable. (https://github.com/zjs81/meshcore-open)
IMHO MT has a fundamental algorithmic flaw when it comes to dealing with very unreliable and lossy links.
Anyway how would either network fare in the Iranian situation where the authorities are actively trying to shutdown communications? Sure the authorities could simply flood the network with traffic.
Even without something fancy (e.g WiFi Direct, iptables on a rooted phone) you could have phones alternating between offering a network and promiscuously connecting to offered networks, then routing between these.
It's simple enough that I'd be surprised if nobody has done it, maybe because it's slow and power-hungry? I haven't tested setting up hotspots and switching networks from inside app logic, but afaik it's fine as long as you don't do both at the same time.
edit: Having thought about it for a minute, a DTN over WiFi Direct is probably the way to go. Establishing identity for signing||encryption might be tricky, but if you can arrange that in advance or just yolo it in plain text then should be straightforward. Can't find any prior art though. I'll let Codex have a go and report back.
A bit unfortunate naming, indeed.
We've had interesting mesh network experiments in the past (maybe some here remember Fonera), and some are trying on various bands, e.g. World Mobile, but none of these can realistically work unless prepared and deployed in advance, which happens through public choices, meaning public networks built to be truly resilient, rather than centrally controlled.
So, while technically interesting, they are not realistically usable in civil war situations. Instead, it's interesting to think about how vulnerable surveillance devices are in these situations, like modern connected cars and smartphones, which can operate a mesh centrally, for example, to guide and block cars at strategic road junctions and centrally acquire location data from the "meat-bots" carrying smart devices with them.
If I were a citizen in a civil war, I'd be afraid of the connected car and would stay far away from my smartphone if I decided to take action. If I were the ruler of a country that can't make its own cars and smart devices, I'd block them by any means necessary due to the serious national security risk they pose.
We need open hardware and FLOSS imposed by law, making it ILLEGAL to sell black boxes and fund research for verifiable hardware. Not to believe that the latest mesh app is good for anything without giving a single thought to real-world use.
I think they just ran out of funding and died with a whimper.