From multi-network to the Internet. One protocol to rule them all

From multi-network to Internet


We had left our history at the time the computer network that linked some North American universities expanded to other countries using 3 separate networks; Arpanet (telephone connection) PRNET (by radio) and Satnet (via satellite) each using its own protocol. The time has come to unify them through the use of a single protocol.

From multi-network to Internet

In 1973 the now renamed DARPA set up a working group for the creation of this new protocol. This group would be guided by the same principles of consensus and meritocracy with which the original protocols were created.

The new protocol would opt for a completely different approach. Based on the work of Cyclades, the French network communication project and research by Xerox PARC laboratories, it would seek to decentralize control of the network by passing it on to senders and receivers.

Xerox Research Labs (The Inventors of Photocopying) they were responsible for many innovations. Unfortunately, their executives were unable to make them profitable. One of the best known inventions was the Alto computer.

Decades before Apple released its first graphical desktop model, the Altos had a mouse, graphical desktop, and window manager. Xerox needed a way for these computers to communicate with each other and with another invention they had made; laser printers.

To achieve this they hired a young engineer named Robert Metcalfe. Metcalfe was a member of the team charged with developing the common protocol for all networks, but he was not comfortable with the consensus and meritocratic approach so he focused on his work for Xerox.

That work would be the basis of what we know today as Ethernet, the communication protocol for local networks. The basic idea for Ethernet came to him while spending the night at Steve Crocker's house. Crocker was the coordinator of the group that created the first Internet protocols and established the methodology of consensus and meritocracy that would later be the basis of many open source projects.

It is said that not being able to sleep, Metcalfe took a scientific magazine and read an article about the protocol used to connect two computers using radio waves that we discussed in the previous article. The engineer improved the idea by having the transmitter check if the channel is in use before transmitting, in which case it would wait. Additionally, it constantly monitors for problems, in which case streaming stops and resumes later.

From Metcalfe's works The so-called PUP protocol (PARC Universal Packet) emerged in which, unlike ARPANET, the controls over the data flow and the integrity check are not in the interconnection infrastructure but in the sender and receiver.

In May 1974 the DARPA team publishes the first draft of the Transmission Control Protocol (TCP) adopting the PUP approach. Although those responsible cite a technical reason, that it is much easier to scale a decentralized system, some historians believe that the decision had to do with the same philosophy that led to the choice of a meritocratic and consensus working method.

The original ARPANET network it was controlled by a contractor called BBN who had the power to access, monitor and update it remotely.

Anyway, you don't have to look for any conspiracy theory for the original approach. At a time when computing power was low and expensive, it was centralized so as not to take away processing power from the local computers of network participants.

Interoperability between networks using the new protocol was put to the test on November 22, 1977. A pickup truck driving on a California highway radioed to an ARPANET computer that cabled it to another computer located on the US East Coast. From there, they were sent via satellite to the United Kingdom, which sent them back to California where they were received by the truck in which a team generated patterns that allowed it to check the integrity of the data. There were no errors, just interruptions caused by bridges and other construction.

After several revisions, in 1978 the TCP protocol was forked into two parts; TCP would take care of communications between computers and the new IP protocol the connections between different types of networks. What was sought was to avoid duplication of functions between computers that give access to external networks from those that control local networks.

Continued ...


4 comments, leave yours

Leave a Comment

Your email address will not be published. Required fields are marked with *

*

*

  1. Responsible for the data: AB Internet Networks 2008 SL
  2. Purpose of the data: Control SPAM, comment management.
  3. Legitimation: Your consent
  4. Communication of the data: The data will not be communicated to third parties except by legal obligation.
  5. Data storage: Database hosted by Occentus Networks (EU)
  6. Rights: At any time you can limit, recover and delete your information.

  1.   Optimus said

    The second to last paragraph reminded me of the movie Firewall with Harrinson Ford.
    Very nice articles doing the historical panning.
    Salu2

    1.    Diego German Gonzalez said

      Thank you

  2.   St. said

    What an interesting series of articles. I am reading them very carefully. Although I do not quite understand what he means by "a meritocratic and consensus working method."

    Thank you!

    1.    Diego German Gonzalez said

      Instead of taking into account the place that each person occupied in the ladder, their contributions were considered and an agreement was reached