I'm surprised more hobbyists haven't developed public mesh networks. There are so many wireless routers and devices that this seems like a great opportunity. I would think makerspaces would be all over this idea.
Back in the dark ages, there were hobbyists who would order alarm circuits ($6/month) and pay for the DSL equipment on each side. This let them connect to the internet back as it was in 1992 or so.
The problem with this was that you had to know somebody who would be willing to be the other side of your connection. That obviously didn't scale very well.
Portland tried having a [free WiFi](https://www.oregonlive.com/breakingnews/2008/06/portlands_wi...) setup with antennas mounted on streetlights. The goal was to have 95% of the city covered but most people needed a signal booster to reach the antennas and very few people did that. My guess is that even if they did that the bandwidth would have been really slow.
There have been a couple of attempts to do a mesh network in Portland with hobbyists but it's really hard. People have moved from computers to phones, tablets, and dedicated devices like Rokus or smart TVs.
A tech support manager I worked with said "these people don't want to be taught how to catch a fish. They want us to catch the fish, cook it, and serve it with a smile."
Some people are OK with amateur performance - there's no internet because Betty went away for the weekend and isn't there to re-boot the router - but most people aren't.
Building and running a network that is useful (fast and reliable enough) for home internet use is a full time job, and is hard work. This is why most hacker space mesh networks have not achieved anything close to this. NYC mesh is the one notable exception, but this is because it is someone’s full time job, and it is run in a pretty professional manner, like a small wireless ISP.
The alternative, a network that is not fast and reliable enough for home internet use, is another option. But in that case, what IS it useful for? These efforts usually just end up turning into weekly hack nights / hangouts for the mesh group, since there’s no real point in putting up hardware.
A third hurdle is that if you put together 12 random nerds in a hacker space, the odds that any of them can see any of the other nerds houses from their house is basically zero. Since wireless internet radios require line of sight, it’s very hard to even create the first link in the network in this scenario. Getting critical mass to form any kind of network in a neighborhood requires door to door sales, something that most mesh networking nerds are very bad at.
You lay out some really good challenges for establishing a decent public net
1) Maintenance. What are the challenges of maintaining? Hardware? Balancing network? Software integration?
2) Speed. What are the barriers for making fast and affordable?
3) Distance. What are the hurdles for last mile?
This has been telecom's unicorn for a long time. Fiber, wireless, satellite, cable where all promised to deliver. I think there is something with mesh networks and free software that can break this barrier. I don't know exactly what it would be but it still seems like a possibility to me.
Infrastructure is another issue. There's a lot of it between me and my nearest hacker space and while it would be possible to bridge it with guerilla relay transmitters that means deploying a lot of quasi-illegal radio units that are going to need battery changes and the cost could mount quickly. I feel like $10/unit is the magic price point for potential orphan hardware.
I think there's a strong argument to be made in California for having community-operated communications infrastructure in case of earthquake or disaster but it's got to be completely turnkey, with something people can just take home, plug in, and forget about.
https://www.nycmesh.net/blog/how/