1. The weather balloon will keep going up until it bursts (typically in the stratosphere). This will not take long (under two hours). So, if you are going to use a balloon it's going to be tethered. Tether it means that it can easily be shutdown by cutting the tether.
2. WiFi range with omnidirectional antennas is very limited. With directional antennas it is measured in km. So either you have to keep the balloon near the ground (which makes it easy to detect and destroy) or you have to point an antenna at it from somewhere (which means that you can't scan for open WiFis easily).
3. The typical payload for a weather balloon would be about 1kg. The solar panel that you are pointing to already weighs more than that and you'd have to include the Raspberry Pi, storage (Pirate Bay is 21.3Gb), suitable antenna, backup batteries and housing. So you'd be having a hard time and would need multiple balloons.
So suppose you really want an airborne server platform. You are going to need the following:
1. Something that flies under its own power and has sufficient energy to keep flying. That probably means a gasoline or kerosene engine of some sort, unless you keep it really light and can use solar power.
2. The computer bit is easy and everyone is focussing on it but it's not exceptional. Pirate Bay is small and needs a not complex computing environment (especially if there were many of these and any load was spread across them).
3. A way of communicating with this device. This means radio communications with sufficiently high bandwidth and that implies ground stations to pick it up.
While I agree that there's a bit more to it, let me see if I can rebuff a few of your points.
1. Yes, a typical weather balloon will do that, but it's possible to make a balloon reach neutral buoyancy at a specific altitude see:zeplin.
2. WiFi is a poor choice, that I agree. Assuming you're using something else, you can actually use a directional antenna pointed straight down with a very narrow (2-5 degree) beam. As long as there's no obstructions on the way down, it should work fine. It's like of like how a satellite works. You're probably not going to get 100Mbps, but if you multiplex it over enough balloons, it might work.
3. See 1, you can tune buoyancy with weight. Weather balloons are usually really light so that it keeps the costs down and they can go higher faster.
I hope I'm not being pedantic by pointing out that the person to whom you are replying has, in fact, floated a capsule to the upper atmosphere: http://blog.jgc.org/2011/04/gaga-1-flight.html
> but it's possible to make a balloon reach neutral buoyancy at a specific altitude see:zeplin
Only for a short time. Hydrogen leaks, and helium leaks even faster. Plus how do you keep it from blowing away in the wind? (Keep in mind that winds also blow up and down.)
Leaking is a harder problem, I don't see any way around it other than using hot air or having some sort of autonomous in-flight refueling. As for the wind, I vote that it shouldn't matter, as long as it doesn't get stuck in a downdraft that crashes it, or an updraft that brings it above its operational ceiling, then lat/long shouldn't matter too much as long as there's enough of them. This is supposed to be a backup in case TPB gets shut down, some sort of mass backup distribution. On average, the airborne servers should be available once in a while.
But you're right, there's massive problems any persistent airborne presence.
> then lat/long shouldn't matter too much as long as there's enough of them
So you are suggesting let them move randomly yes? Well if you assume a 10km range you would need 1.62 million of these to cover the earth, and since they would probably bunch up due to wind, you'd need even more of them.
And why do you assume they won't be stuck in up or downdrafts? Earth's air pressure varies constantly, so a truly neutral position is impossible. But even if you managed it, the slightest breeze would move it easily - meaning they would constantly crash to the ground or go too high and pop.
1. The weather balloon will keep going up until it bursts (typically in the stratosphere). This will not take long (under two hours). So, if you are going to use a balloon it's going to be tethered. Tether it means that it can easily be shutdown by cutting the tether.
2. WiFi range with omnidirectional antennas is very limited. With directional antennas it is measured in km. So either you have to keep the balloon near the ground (which makes it easy to detect and destroy) or you have to point an antenna at it from somewhere (which means that you can't scan for open WiFis easily).
3. The typical payload for a weather balloon would be about 1kg. The solar panel that you are pointing to already weighs more than that and you'd have to include the Raspberry Pi, storage (Pirate Bay is 21.3Gb), suitable antenna, backup batteries and housing. So you'd be having a hard time and would need multiple balloons.
So suppose you really want an airborne server platform. You are going to need the following:
1. Something that flies under its own power and has sufficient energy to keep flying. That probably means a gasoline or kerosene engine of some sort, unless you keep it really light and can use solar power.
2. The computer bit is easy and everyone is focussing on it but it's not exceptional. Pirate Bay is small and needs a not complex computing environment (especially if there were many of these and any load was spread across them).
3. A way of communicating with this device. This means radio communications with sufficiently high bandwidth and that implies ground stations to pick it up.