The real scenario long-term is that quality content like Wikipedia could either be taken offline, be poisoned by AI or taken over or censored by authoritarians or corporations or interest groups. Like social media and parts of the normal internet already.
So archiving is good anyways.
To your actual question - quality non-fiction e-books would be valuable. Wikipedia is a superficial skimming of human knowledge, lots of the real stuff is in books (think medicine, agriculture, algorithms, engineering etc.).
Practically, a home WIFI has a very limited range, so just handing out sheets with instructions to your neighbors would work. And for a wider mesh network, you'd need to make do with whatever evolves in that scenario.
https://www.arednmesh.org/content/supported-devices-0
It's self-configuring, too - as soon as the node spins up, it will automatically find and connect to nearby nodes and start routing.
Because inevitably, you find a bunch of them after you've lost access.
I think this identifies a real problem in western society; there's a crisis situation of sorts, but instead of people talking to each other, this comment worries about whether people can figure it out on their own.
I believe that if there were to be a crisis situation like that, long-term power outage or whatever, people would find each other again instead of the individualism we have right now.
"Congo."
I personally have not known shame for decades. I was ashamed.
Then someone tried to burn down the library.
I would also store it in a steel Oscar the Grouch style trash can for a cheap faraday cage, which gets you protection from solar flares, and EMP blasts.
But--if you don't think of asking Hacker News every single thing you need to know beforehand, I think you still want the LLM to answer questions and help you bootstrap it.
Which militaries are working on battle field AI. All of them.
Could a 64Gb dual xeon run say 50 to 100 users of kiwix?
Having something that you can plainly ask how to start that will point you in the right direction and explain the base concepts is worth a lot more, it turns raw data into genuine information. Yes it can be wrong sometimes, but so can human teachers and you can always verify, which is a good skill to practice in general.
You can build a hotspot and try setting up meshes with any of the available hardware or software packages out there, but you're going to end up being the gatekeeper to the service. HAM radio ends up working out the same way, as I understand. It's just too technical for people to have this spring up collectively without a single person or team doing everything.
Lack of tech experience to even know how to build a mesh let alone prioritize its limited bandwidth is why the general public isn't going to assist.
>And then, could they repeat the hotspot, to build a mesh? I know there are projects to do that, but what do they accomplish exactly?
Yes, pretty much. The problem is poor definition of the problem, though.
What are we trying to solve? A way to send trickles of comms out, like "Mom and Dad, we're alive?" or "We have life-threatening casualties at x',y'?" Emergency kiosk to send emails one at a time? Doable if you have an Internet source like a Starlink, or any other uplink that's still up somehow.
Or is to restore the "Internet" as generally known, which might as well be synonymous with YouTube and Netflix and web browsing for people. You and your system would be overwhelmed as soon as your mesh comes up.
I guess a requirement for that is a sufficiently generalized protocol with a matching hardware stack.
My NAS has 2x2Tn striped and a video 1Tb. It streams 4k to multiple sources. If I needed faster, I'd throw it on the install server. 4x1TB striped, and 8Gb of ram. 2xGbit net. Did it tell you this cost all of $60?
Something like the internet archive, but fully decentralized.
Isn't OSM like 2TB to host if you include all tile scales? I guess you could skip the last few where most of the size explosion lies but that doesn't give you much precision.
That could be easily compressed by having all of these link to one water image though, no?