It soon became obvious that the ARPANET was becoming a human- communication medium with very important advantages over normal U.S. mail and over telephone calls. One of the advantages of the message systems over letter mail was that, in an ARPANET message, one could write tersely and type imperfectly, even to an older person in a superior position and even to a person one did not know very well, and the recipient took no offense. The formality and perfection that most people expect in a typed letter did not become associated with network messages, probably because the network was so much faster, so much more like the telephone.
– J.C.R. Licklider, Albert Vezza, Applications of Information Networks, Proc of the IEEE, 66(11), Nov 1978.
Who invented email? Electronic mail is a natural and perhaps inevitable use of networked communication technology that developed along with the evolution of the Internet. Indeed, message exchange in one form or another has existed from the early days of timesharing computers.
Network capable email was developed for the ARPANET shortly after it’s creation, and has now evolved into the powerful email technology that is the most widely used application on the Internet today. Key events and milestones in the invention of email are described below:
- Timesharing computers. With the development in the early 1960’s of timesharing computers that could run more than one program at once, many research organizations wrote programs to exchange text messages and even real-time chat among users at different terminals. As is often the case, more than one person at the same time noticed that it was a natural use of a new technology to extend human communications. However, these early systems were limited to use by the group of people using one computer.
- SNDMSG & READMAIL. In the early 1970’s, Ray Tomlinson was working on a small team developing the TENEX operating system, with local email programs called SNDMSG and READMAIL. In late 1971, Tomlinson created the first ARPANET email application when he updated SNDMSG by adding a program called CPYNET capable of copying files over the network, and informed his colleagues by sending them an email using the new program with instructions on how to use it.
To extend the addressing to the network, Tomlinson chose the “commercial at” symbol to combine the user and host names, providing the naturally meaningful notation “user@host” that is the standard for email addressing today. These early programs had simple functionality and were command line driven, but established the basic transactional model that still defines the technology — email gets sent to someone’s mailbox.
- MAIL & MLFL. In 1972, the commands MAIL and MLFL were added to the FTP program to provide standard network transport capabilities for email transmission. FTP sent a separate copy of each email to each recipient, and provided the standard ARPANET email functionality until the early 1980’s when the more efficient SMTP protocol was created. Among other improvements, SMTP enabled sending a single message to a domain with more than one addressee, after which the local server would locally copy the message to each recipient.
- RD. The Director of ARPA, Steve Lukasik, asked Lawrence Roberts, then the director of the IPTO, to improve on READMAIL, which required messages to be read in order, and with no ability to save or reply. Roberts wrote RD in one three-day weekend as a collection of macros in the Tenex text editor TECO (Text Editor and COrrector), and called the program called RD.
The new program included capabilities to sort email headers by subject and date, giving users the ability to order the messages in their Inbox, and to read, save, and delete messages in the order they wished. In a sign of the pragmatism associated with much of the email development over the years, RD was developed not as a research effort, but as a practical effort to solve a real-world problem of email management.
- NRD. The DARPA researcher Barry Wessler improved on RD, and called the new program NRD, including several new usability features.
- WRD / BANANARD. Marty Yonke recoded SNDMSG and NRD into an independent program called WRD. This was the first program to integrate reading, sending, and a user-friendly help system in the same application, and was later renamed BANANARD.
- MSG. John Vittal improved on BANANARD and called the new program MSG, with powerful features like message forwarding, a configurable interface, and an Answer command that automatically created properly addressed replies. MSG can fairly be called the first modern email program.
Dave Crocker (see MS below) feels the Answer command was revolutionary: “My subjective sense was that propagation of MSG resulted in an exponential explosion of email use, over roughly a 6-month period. The simplistic explanation is that people could now close the Shannon-Weaver communication loop with a single, simple command, rather than having to formulate each new message. In other words, email moved from the sending of independent messages into having a conversation.”
- MS / MH. In 1975, the DARPA program manager Steve Walker initiated a project at RAND to develop an MSG-like email capability for the Unix operating system. The project was undertaken by Dave Farber, professor at the University of California at Irvine. Dave Crocker, starting graduate school at the University of Southern California’s Annenberg School, designed the functional specifications, and Steve Tepper and Bill Crosby did the programming.
The resulting system supported multiple user interfaces, from the basic Unix email command to the MSG interface, and was called MS. Crocker comments: “The program was very powerful, and very, very slow.” A follow-on project at RAND rebuilt the program to take more advantage of the Unix system environment, breaking the commands out into individual programs that ran in individual Unix shells. Bruce Borden did most of the programming, and named the resulting application MH as an abbreviation of Mail Handler. Since 1982, Marshall Rose and others have upgraded and maintained MH, and it has become the standard email application for the Unix environment.
- RFC 733 & RFC 822. In 1977, Crocker, John Vittal, Kenneth Pogran, and D. Austin Henderson collaborated on a DARPA initiative to collect various email data formats into a single, coherent specification, resulting in RFC 733.
The specification combined existing documentation with a bit of innovation, and was the first RFC explicitly declared an Internet standard in order to try and bring some order to the various email formats in use across the ARPANET — an effort not initially greeted with universal approval among the independent, distributed research community. In 1982, Crocker revised RFC 733 to produce RFC 822, which was the first standard to describe the syntax of domain names.
- MMDF. In 1978, Crocker followed Dave Farber to the University of Delaware, where they took on a project for the U.S. Army Materiel Command (AMC) to develop a capability to relay email over dial-up telephone lines for sites that couldn’t connect directly to the ARPANET. Crocker created the first version of what would become the Multi-purpose Memo Distribution Facility (MMDF) in six months of work, and then set up and operated an experimental relay site at the University of Delaware for various AMC sites.
The MMDF link-level protocol was developed by Ed Szurkowski. Several others worked on the software after Crocker left, including Doug Kingston, Craig Partridge, and Steve Kille, developing enhancement such as creation of a robust TCP/IP layer. Kille adapted the software to support the ISO/CCITT OSI X.400 email standard, one of the first systems to do so, naming the software “PP” after “Postman Pat”, British vernacular for the local postal delivery person. MMDF was also deployed to provide the initial email relay capability for the CSNET.
- Sendmail. In the early 1980’s, email relaying was also being performed using the simple UUCP technology at the University of California at Berkeley, where the BSD Unix operating system was developed. Eric Allman later created a program called delivermail to cobble together multiple email transport services, creating, in effect, a switch rather than an integrated email store-and-forward capability. Allman then built on this experience to create the sendmail program, which was distributed with BSD Unix, and has gone on to become the the most commonly used SMTP server on the Internet.
- Commercial Email. In 1988, Vinton Cerf arranged for the connection of MCI Mail to the NSFNET through the Corporation for the National Research Initiative (CNRI) for “experimental use”, providing the first sanctioned commercial use of the Internet. Shortly thereafter, in 1989, the Compuserve mail system also connected to the NSFNET, through the Ohio State University network.
- Online Services. In 1993, the large network service providers America Online and Delphi started to connect their proprietary email systems to the internet. Soon after, internet providers themselves would begin offering email domains, beginning the large scale adoption of internet email as a global standard.
Resources. The following resources provide additional information about the history of email:
- Early forms of email addresses.
- A list of early email systems on the ARPANET can be found in Appendix A of RFC 808.
- Claims by Shiva Ayyadurai to have invented email.