On distributed communications networks: Paul Baran 1961

The first word button in the following lines is regarding distributed communications networks. It talks about redundant networks that aren't completely efficient and what happens if one node or one part of the network is hit by a nuclear blast. This was done in 1961 (before the Cuban Missile Crisis when Kennedy was President still) (I was 13 then)  when everyone still thought it was inevitable that the the whole world or (many of it's cities) were going to be blown out of existence at any moment through nukes. At that time this was designed to maintain our government's information even if multiple cities were blown out of existence at that time. Some nodes or mainframe computers with multiple redundant information in each of them would allow our governmental system to continue no matter what places were blown up.

One of the more interesting points for me is that he mentions a "decentralized" network where all nodes are equal. By having this sort of decentralized network there is no central node to be blown up so it can work with any or most nodes blown up or not working. This describes pretty clearly what the Internet is where often no one node  not working or destroyed would destroy all information because of multiple redundancies and multiple nodes in existence all carrying basically the same information.
This allows many or most nodes "servers" to be down(from an EMP or for servicing) and still all useful information might not be lost if you have information on enough servers around the world redundantly distributed. This is what allows the Internet to be superior to  any previously used network structure by having multiple redundancies around the world as well as no one central node in existence without which the whole Internet goes down.

However, to make this sort of thing useful, "memory" had to cost much much less than it did in 1961. But as memory became less and less expensive for the general public between 1961 and the 1990s this Internet could exist with multiple redundant nodes and work like it does now.
 RAND Paper P-2626

On Distributed Communications

Baran developed the concept of message block switching during his research at the RAND Corporation for the US Air Force into survivable communications networks, first presented to the Air Force in the summer of 1961 as briefing B-265^[1] then published as RAND Paper P-2626 in 1962 and then including and expanding somewhat within a series of eleven papers titled On Distributed Communications in 1964. Baran's P-2626 paper described a general architecture for a large-scale, distributed, survivable communications network. The paper focuses on three key ideas: first, use of a decentralized network with multiple paths between any two points; and second, dividing complete user messages into what he called message blocks (later called packets); then third, delivery of these messages by store and forward switching.

end  partial quote from:

Packet switching