Apparently NASA's also considering standardizing communication protocols to use Internet standards. That's a good idea, of course, but there is a hitch: The protocol degrades rapidly when latency goes up, which is a fancy way of saying that across large distances the speed of light becomes a problem. The TCP (Transmission Control Protocol) requires the receiving end to send an acknowledgment packet back for each packet received, so that dropped packets can be detected and sent again. When the speed of light becomes a significant factor, the acks can pile up and the sender can end up retransmitting lots of packets it doesn't have to because it thinks it never received an acknowledgment.

This is why, for example, satellite Internet connections work well when streaming large quantities of data in one direction, but become frustratingly slow when you're exchanging short packets of data and awaiting a response. Imagine, worst case, you're using a simple terminal emulator, and you have to wait a noticeable fraction of a second after you press a key for the other end to echo it back so that your terminal displays it on the screen. That's just the situation with geosync comsats; three seconds roundtrip to the Moon would be intolerable.

It'll probably turn out that NASA can get away with using Internet protocols locally on the Moon, but the pipeline from the Earth to the Moon will need to use another protocol better suited for the environment, and perform protocol conversion at each end.

Nothing in realworld engineering is ever as simple as it seems.