Alien SUPER-EARTH: If there is LIFE up there, it'll be moaning about the weather

Re: @zmodem

i will be blipping about with hyperdrive with http://www.youtube.com/watch?v=Pen24y_FaW8 booming out my onboard PA system, while you still doing somersaults on a flying saucer

Re: @zmodem

"in space you use as high as you can so packets go futher(sic), faster"

Quite mad, it's as if the speed of light was just a number and space wasn't so stuffed with noise on the complete spectrum that doing radio even over Terra/Luna distances is far from trivial. Just 93 million miles away is a quite large unshielded fusion reactor which pumps out all sorts of junk, luckily we have a magnetosphere to deflect most of it, not so in open space.

Re: @zmodem

". . . leave omni directional nodes every few light years, you can just send out controls through the network array and it will always reach it destination, unless a node fails . . . packets have CRC and timestamps and so on."

CRC. Over a 'few light years'. Sure - why not?

Let's make the maths easy and say you drop your nodes every 1 ly. Node 1 (N1) relays its packet onto N2. Upon being recieved at N2, one year later, the packet fails the CRC. N2 then sends a NAK[1] to N1, which takes another year to get there.

If you're keeping track, two years have elapsed.

N1 must now re-send the packet, but where does that packet come from? The quickest option would be to have N1 resend the packet directly, rather than pass the NAK all the way back to the origin[2]. To do this, however, N1 would need to have a copy stored in its buffer. Given that two years have elapsed since it first relayed the packet, it stands to reason that each node would require enough storage to hold two full years worth of all data transmissions. Storing data in space, by the way, is not a trivial matter.

But, this being a thought-experiment, let's assume that the nodes have enough storage and that the storage is robust enough to survive without corruption for the requisite 2 year period[3]. The replacement packet is duly dispatched[4] and this time, the CRC suceeds.

Three years have now elapsed.

Error correction is only useful in situations where the time for that correction to take place is less than the useful lifetime of the data you are transmitting. A storm warning, for example, is only useful until the storm hits.

Further, it would take an inordinate amount of time to get confirmation that your message was received - at least the RTT over the entire distance. This, along with the usefulness of the data after the time elapsed is relevant as you propose using it for 'controls'. Even assuming that a control signal taking several years to transmit is still relevant when received, a device being controlled will need to relay telemetry data back to the controller so that it knows that its previous commands were received and executed correctly and what commands to send next.

There are numerous other small and not so small complications here but the point is that communication of any type over interstellar distances is limited in utility by the speed of light. Any message more urgent than 36 years is simply not worth sending to or from any prospective colonists on GJ 436b.

I can imagine a future where scattered human colonies send daily news journals between each other as a way of keeping in touch but meaningful two-way communication between such colonies would be impossible for all but the closest of them. By the time we received a message, we wouldn't know if the person who sent it was still alive.

The depressing fact of the universe is not that we are alone, but that we likely aren't but yet will almost certainly never meet any of our 'neighbours'.

1 - You could set it up with a standard ACK and RTT but that doesn't save any time, though it would prevent problems with a NAK being lost in transit.

2 - Which may, itself, be several nodes (and therefore ly) away.

3 - Even then, given this is proposed as an alternative to long-life storage, one must assume that the storage being used for the send/recieve buffers will need to be replaced periodically. Logistically, that would be difficult. If you imagine some colony ship dropping these nodes as it goes, by the time it gets to its destination, some of those nodes will certainly have failed.

4 - Of course, once retransmitted, that packet will need to be stored for another two years, just in case, increasing the required storage capacity of the nodes.