[Got what I needed, thanks – PLEASE DO NOT boost this further.] This is an open call for insight and help: I’m looking for guidance from folks who are familiar with #Meshtastic to help me understand what would be involved in setting up and maintaining a national-scale #mutualaid and disaster-response (relief/recovery/“resilience”) network in the UK.
Post
Replies:
29
As jonty said ... It's complicated, and quite quirky. I encourage you to have a long chat with your LLM about your idea. Definitely get a couple of cheap nodes, and test, experiment, get to know your mesh neighbors. Try mqtt. Ask your LLM about personal security on mesh. Have fun and good luck!
@adamgreenfield I see you've had a decent number of responses that cover the issue in a lot of detail.
Re: the nullagent post, I'm sure the personal position that they've taken is evident, as is the breach of the traditional protocol on security disclosure (i.e. don't be a script kiddy).
Re: the user experience, I suggest grabbing a demo of the devices in a relatively open area and using the phone app. It's very accessible, and behaves like a standard chat app.
@adamgreenfield You might want to look at @tek 's FrozenBBS. It's a message board, you could use this for coordinating people and logistics.
@adamgreenfield@social.coop don’t bother, it doesn’t fucking work
@adamgreenfield meshtastic isn't ready for prime time especially not at that scale. Have a look at Meshcore instead, though that also needs a bit more work I believe to become sufficiently user friendly
@oseiler @adamgreenfield
"Meshtastic isn't ready for prime time, try this even more obscure system". Really?
@adamgreenfield just poking my head in to say I’ve been lightly looking at these things too , recently-ish got my foundation amateur radio license, played around with lora early on (and things network where a group of us volunteered to look at a London network, quickly reported that yeah, this won’t work as they imagine , this environment is different from your example network in Amsterdam, but ignored as they got their mate in who was all, yeah let’s do London and then nothing ….
@adamgreenfield I realise this will not be the answer in many areas..but round here in rural #YorkshireDales the police, cave and mountain rescue use the same sort of radios. We have had discussions about getting a few more to be used in emergencies. These would be held at village halls or churches which would be focal point in a crisis so could use existing infrastructure and not reinvent the wheel. Rural areas are at much more risk of frequent or long power cuts, poor or no mobile or radio signal so a lot of the more high tech options are not likely to work out here. #EmergencyResilience#EmergencyCommunication#DisasterPlanning
@adamgreenfield
There was a hackspace in Greenwich, SE London that created a meshnet that covered the area around Deptford Creek.
https://dek.spc.org/ is their website.
https://dek.spc.org/julian/consume/ is one of the write-ups that they used for one talk about their experiences.
https://www.tandfonline.com/doi/full/10.1080/13604813.2017.1325236 is one of the papers written about the meshnet.
https://web.archive.org/web/20170629105032/https://www.kiddingthecity.org/open-wireless-network/ is the archived version of an article about how it took place.
@adamgreenfield heya there’s been some really good responses that I won’t rehash. Something you might find is reaching out to a few established mesh groups and asking them about what they’ve been able to achieve: https://bayme.sh/ and https://nyme.sh/. A slight tangent but equally helpful was a recent “survey” of UK meshes by an individual who runs a town mesh https://buffalora.org/2025/08/04/uk-holiday/. I’m friendly with the Cornish mesh contingent and can put you in touch with them if you want to chat. I’ll be watching how this unfolds with interest !
@adamgreenfield I cannot state this strongly enough: Meshtastic is *not* suitable for this as-is, nor is it likely to be anytime soon. It is riding on a huge wave of hype but the underlying technology does not scale well or offer the kinds of guarantees that are needed in disaster situations.
@jonty @adamgreenfield Same observations here at my end. #Meshtastic is very indeterministic currently and I don't see a very active development community, either. 🤷♂️
@jonty @adamgreenfield I, unfortunately, came to the same conclusions after spending awhile (like most of 2024) messing around with meshtastic
Right now it works for things like: two people (both of whom are pretty good at troubleshooting) hiking off-grid together and keeping in touch.
IMO it's just too hinky for the kind of static network it advertises/envisions.
@jonty That’s gold, thank you. If you were going to stand up this kind of network in the next three to six months, what would *you* use?
@adamgreenfield @jonty Three to six months? It will take longer than that just to complete a propagation survey.
I'm having fun with Meshtastic but would never consider it suitable as a trusted emergency/contingency 'service'. A helpful adjunct, perhaps.
Wanna see what comes back from here...
@adamgreenfield Strongly depends on the constraints and the purpose. Is it just comms? Is it specifically chat? Is it broadcast or direct?
Some options:
If you have a bit of money and access to places to put them, point-to-point radio links across the city, with local wifi to access it. Host a forum/chat server somewhere on the network. People just use their phones.
If you're looking at no shared infrastructure and cheap as possible: A lot of shitty Baofeng radios.
@jonty (shitty-Baofeng-net is the “we jam econo” of this space & I approve on GP.)
@adamgreenfield I say this after several years playing with it, desperately trying to make it work well in London.
A group in Amsterdam just set out to do just what you describe, didn't listen to anyone saying "this isn't going to work well", and after a month or two has concluded that ...it doesn't work. They are now trying to pay people to figure out why it doesn't work, rather than reevaluating the technology choice.
How many devices would such a network need, to ensure adequate coverage everywhere it might be required? How rich a degree of communication is afforded by the current technical stack? What kind of budget would a local node need to have available, what sort of equipment and training would they have to arrange in order to access the network? These are the questions I’m looking to sketch out some answers for. Your help is *very* much appreciated. 👊
@adamgreenfield Hey Adam, I see you got some good answers. I'm not sure if anybody highlighted the biggest problem with trying to do this in the UK: The ISM band there has a 10% duty cycle per device. This seems OK at first glance, but it hits repeaters particularly hard. It's been a huge challenge for our users trying to build a mesh there.
@JP_Bennett Very useful, thanks. Is that limit mandated in law? I assume there are ways of circumventing it.
@adamgreenfield Yes, it is the law in the UK. It is possible to circumvent, but do so at your own legal peril.
@adamgreenfield Most of the answers will be very localised around geographical constraints. MT/LoRa has formidable range, but is on a (LoS) Line of Sight radio band. Here in the far more mountainous Aotearoa New Zealand we are building it very much for that purpose, leveraging peaks to deploy repeaters and routers to conjoin whole otherwise disparate regions.
There in the UK you already have a fairly active MT scene, I would look to rally that around the cause.
2 media
@adamgreenfield Very interested in the answers. I have been seriously thinking about this in Glasgow as I have access to an excellent location for a Meshtastic repeater node. Regrettably it looks like there are some security problems with the current implementation, but the overall approach is still useful and hopefully updated software can fix those issues. edit: on reading the links others posted much less exited about Meshtastic.
@adamgreenfield Having been nerdsniped into reading the documentation for Meshcore and Reticulum I think that Reticulum is closest to the Walkaway Network described in @pluralistic's Walkaway book. It is more complex though and requires cleverer nodes that can run python rather than working on purely embedded hardware and so has higher power requirements. 1/2
@adamgreenfield #meshtastic is LoRaWan based, isn't it? Then, can it piggyback on existing LoRa networks (e.g. for IoT https://www.thethingsnetwork.org/ )? I have a LoRaWan gateway running at the top floor servicing a bunch of other people's sensors around the neighbourhood. Perhaps existing docs from Thingsnetwork and other (commercial) LoRaWan service providers can give insight into coverage, range and throughput?
@adamgreenfield i'm not sure meshtastic is really going to work for this but i'm interested to see what you find. ok to boost?
