I want to welcome you to explore the fascinating history of how early internet pioneers first figured out how to exchange digital messages between users. It required creative thinking to solve tricky technical problems with the computers and networks back then!
Communication in the Early ARPANET Days
In the 1960s, scientists were connecting university computers together into early "networks" that could share information remotely. The first of these was the ARPANET funded by the U.S. government‘s Advanced Research Projects Agency (ARPA).
However, these connections focused on shuttling data back and forth, not providing ways for the researchers themselves to communicate. Methods for easy person-to-person messaging didn‘t exist yet across networks.
The ARPANET used packet switching, breaking data into addressed chunks sent independently then reassembled. This allowed computers with different architectures to exchange information. While revolutionary, it didn‘t readily support live communication.
Early on, in 1971, some systems had ways for users logged into the same computer to exchange messages locally. For example, SNDMSG on the TENEX operating system let users append notes to each other‘s mailboxes, storage files allocated for each individual.
But these mailboxes were isolated inside that one machine. There was no standardized way yet to send messages between systems across the experimental ARPANET links that connected them.
Ray Tomlinson‘s Inspiration – Bridging Networks via Email
What messages needed was a creative programmer willing to combine existing parts in an innovative way. Ray Tomlinson working on TENEX and ARPANET saw this opportunity.
The network had CPYNET, a file transfer protocol for sending data between hosts. Local computers had SNDMSG for their mailbox messaging needs.
Tomlinson asked – what if mailbox messages could also be sent over the network? This would allow SNDMSG-style direct communication between people using ARPANET-linked machines, not just locally.
His breakthrough was realizing he could append messages to mailboxes remotely if he merged SNDMSG mailbox addressing with CPYNET‘s file transfer abilities.
Some obstacles remained to translate this idea into reality…
Technical Hurdles Towards Network Email
Tomlinson first had to modify CPYNET to allow "appending" data rather than just overwriting files entirely in transfers. Appending was necessary to add messages to the end of mailboxes without deleting existing content.
Next was addressing – there had to be a way for SNDMSG to distinguish if a message should be delivered locally inside TENEX or directed out over ARPANET.
Tomlinson introduced the @ sign for this purpose. The destination mailbox was now identified as username@hostname – indicating which system that user account was on. No ambiguity or overlap, since @ wouldn‘t normally appear in usernames on its own!
First Email Sent Over ARPANET in 1971
With the modifications complete, Tomlinson prepared to test this new email potential. The first messages were sent over ARPANET wires between two TENEX servers in the same room – BBNA and BBNB!
foobar@BBNA => foobar@BBNBThe foam user account sent was simply for demonstration. But when received successfully… networked email had arrived! Tomlinson distributed the capability to ARPANET developers in another message. By 1972, it was being adopted by other systems connected to this early internet.
Email Quickly Changed Communication
Once email spread through networks, digital messaging saw hockey stick growth over the next decades as the internet expanded:
Year: Users (Millions)
1990 ~1
1995 ~100
2000 ~500
2010 ~3000
2020 ~5000Today, over 100 trillion messages sent per year. Spam increased due to privacy issues, but instant email is now indispensable personally and professionally. We owe thanks to pioneering programmers like Raymond Tomlinson who creatively bridged gaps between networks and users.
I hope you‘ve enjoyed this initial tour of the very first email system back in 1971! Let me know if you have any other questions.
