Such devices would need to return land on a regular basis even if helped by solar panels. The costs of maintaining a significant number of this nodes floating would be considerable.
Also the delay on the communications would also be significant compared to a normal server. If you take the low computational power of a Raspberry Pi and put it with a 100Mbps radio (effective data rate would be less), the result would be a very slow website.
Agreed. I wonder if the low orbit solution could be actually more achievable than a standard drone one. Both because it doesn't have to fight the wind all the time and because it can be powered by sun more easily. You need enough power to maintain ~6mph (depends on the area, but I believe that would be the long term average in many places). Sure, there's lift and everything, but you're going to drift in one direction most of the time.
I'm not sure about their availability claims either. "With modern radio transmitters we can get over 100Mbps per node up to 50km away". With modern radio transmitters, I think you can also quickly localise the earth-bound part of the system and either close it down, or create enough interference to make the transfer almost impossible... I'm not an expert on radio though - can someone confirm/refute this?
If they actually do this it could revolutionize the internet. If everyone gets a internet drone we could build internet networks in the harshes place in the world.
Such devices would need to return land on a regular basis even if helped by solar panels. The costs of maintaining a significant number of this nodes floating would be considerable.
Also the delay on the communications would also be significant compared to a normal server. If you take the low computational power of a Raspberry Pi and put it with a 100Mbps radio (effective data rate would be less), the result would be a very slow website.