google wifi drones

i meant the concept of flying wifi routers buzzing around the skies of a town. (not specifically requesting them to be google owned)

This is necessary for Bitcoin AND Internet itself. With both decentralized internet and payments the people are close to immune to totalitarian governments. How does a government control its people? 3 things: through control of money, information and guns.  Thanks to internet, it's much harder to spread propaganda and to hide the truth. Thanks to Bitcoin they are loosing the control over money. Thanks to Silkroad clones that depends on Bitcoin and Tor, (or 3d printers) its much harder to prevent people from buying guns (or print them). But all of this would be nothing without universal access to internet.

The Internet in its present form is far too vulnerable, who know what a cornered US government would or could do to save its existence. Probably a lot of damage. A decentralized network (mesh) is next to impossible to destroy, and a decentralized currency like is next to impossible to destroy as long as it can rely on such a network.

router + solar panel attached to a balloon is cheaper/easier to manage yes. especially where there is not excess electric draw from the solar batteries, purely to fuel the drone. as balloons wont need motors. but then again, why need balloons. when you can just mount them to walls of buildings... without a balloon.

few moments pondering the feasibility

so now im thinking cheap, reliable service is wifi routers attached to walls. closer to street level, which then send signals to a large relay station for the larger distances between different towns/cities.

final idea. satalite dishes..

sorry to offer different options as if im changing my preference. but feasibility wise its easier to get people to connect existing technology to a wireless router then it is to have new specific device that everyone has to buy. and drones/balloons normally have to be at a certain height (eg drones have to be above building height to avoid crashing, which may cause signal strength issues due to distance from the ground.)

clever idea indeed

I'm a member of a not for profit private wireless network that operates in the 172.16.0.0/12 address space and is made up of individually run nodes that are connected.    It's a bit like the OP envisaged, but not really targeted at a doomsday scenario.  This is also not a true dynamic mesh network.  More fixed point-to-point, but the topology is mesh like.

I have set up a bitcoin node on this network and 1 or two guys connect to me.  The problem with using this space is that the standard bitcoin code severely limits discovery on this address space as it is considered private ips.  My node happens to be also internet connected so acts like a gateway on this network.

Some issues I recall:
- node doesn't advertise a private ip
- node doesn't allow multiple connections in the same16 subnet
- node doesn't get it's own ip from other nodes.  So my node only advertises it's public internet facing address onto the private network.

I would love to be able to make auto discovery possible on this network but it is not.  At the moment I have to manually advertise the existence of my node and manually addnode other nodes I become aware of (and get them to addnode me) in my bitcoin.conf file.

If the network code of bitcoind implemented some changes things would be easier:
- allow it so send the ip as reported by the node it is connected to.  (as per https://github.com/bitcoin/bitcoin/pull/3088)
- allow the node to advertise a private ip
- the ability to allow multiple connections on some "trusted" networks
- multicast/broadcast discovery (see https://github.com/bitcoin/bitcoin/issues/3802)

I suspect some of these changes will be helpful in other use cases also.  E.g. assisting nodes on the same private lan finding each other and thus to reduce combined bandwidth usage.

The problem with these remains the outdated technology.
Batteries.

Add two more 0's on that 2nd picture and then it will work nicely. 

The great thing about BitPools is that if enough people pledge their money toward it (keeping their coins in their own cold storage) the solution will come to us.

Then it's just a matter of voting with your bitcoins on the best solution.

Maybe safecoin is the answer to the bandwidth issue:
The resources in question are storage, CPU, online time and bandwidth.

I'm not sure how they deal with paying for bandwidth and online time but it might be a good way to reward each node.

http://www.safecoin.io/