This edition of techblog  is about the initial days of internet – right from where all the ‘interconnections’ started!

1954

Back in 1950s, computers could only perform one process at a time  and  during that time no other process can use the CPU resources. This was termed as batch processing. But today we have time sharing based processors. Hence you can virtually perform many ‘works’ simultaneously.

Another aspect was that the main computer which did all the processing was pretty big and it (with the necessary cooling systems) occupied almost the entire building .

1957

Since companies and institutions had to keep the computers in those large building, the work of the developer was cumbersome. But in year 1957, two main ideas were implemented

a. The developer could work from a remote terminal and connect to the main computer using the same 

b. The time sharing based processing was introduced which allowed two or more developers to work on the same machine at the same time.

1958

This year also saw many advancements in satellite research and the Defense Advanced Research Projects Agency (DARPA) adopted a new plan to join the main computers together. Earlier the communication messages were sent from one computer to the other via a messenger.

DARPA solved this problem by creating links between main computers.

1966

By the year 1966, the network links formed by DARPA became large and was termed as the ‘ARPANET’ (the first form of internet).

Around the same time another network called the RAND (Research ANd Development) was formed for military purposes.

Two other major networks were also born during that time. In England, National Physical Laboratory’s (NLP) network was formed for supporting commercial applications.

And in France, CYCLADES network was created for scientific computing and for enabling communication between research institutes.

Many institutions and companies put small computers in front of main frame computers so that these systems can be used for managing the network activities. They were called Interface Message Processors (IMPs). The main frame was only used for the initialization of the tasks and for managing the data stored. The small system also served as interfaces to the mainframe systems.

These IMPs were connected together for communication and formed an IMP subnet.

The Network Control Protocol (NCP) was then developed in order to standardise the communication between them

Later this was replaced by the more efficient Transmission Control Protocol (TCP)

TCP introduced with many new ways to avoid congestions in the network and the protocol supported breaking down of large files into smaller segments for transmission.

This triggered packet switching in networks.

In the earlier days, the communication links had a centralised architecture and if one of the links goes down, then the entire network fails. Also, people used radio waves to communicate between nodes. This would easily be broken in the event of a nuclear attack (as the ionosphere can no longer reflect the rays back).

This forced them to developed a decentralised architecture.

Meanwhile, the CYCLADES  used nodes for interconnection and formed a decentralised inter-network. And the term ‘INTERNET’ was born!

Later the telephone companies came up with the X.25 protocol which allowed customers to communicate with each other.

X.25 essentially linked the customer through the central server of the phone company.

Later the networking protocols were standard by the ISO (ISO standard model). And, in 1992 Lee invented the world wide web (www). After this stage, the growth of internet was exponential!