Nervous System: The Inventor, the Barbershop Quartet, and the Origin of the Email Attachment

David Kalat writes about email technology and how one man’s fixation on “transmitting photos of grandkids” changed the face of communications.

Download the article.

Email technology has not changed materially since it was devised in the early 1960s—and many of the quirks of email can be attributed to our determination in keeping one of the world’s primary channels of communications backward compatible with email systems devised generations ago.

Almost thirty years ago, a man who wanted to one day have pictures of his grandchildren sent to him digitally figured out a way to stitch together ancient, disparate, and often clunky email systems to send more than just text. Close to three decades after his innovation, email users transmit an estimated trillion email attachments a day. The fact that this happens is now so commonplace as to be beyond remark, but the technology underlying it is a magic trick.

The essence of the problem is that, when the concept of email was first developed in the 1960s, the standard unit of computing power was a seven-bit byte. Because each bit is a binary value of either a 1 or a 0, a byte made up of seven such bits has a total of 128 possible combinations. In the early days of computing, these byte values were mapped onto 128 different alphanumeric characters, collectively called the “US ASCII characters.” Subsequent computer scientists would expand the byte to eight-bit values, doubling the available range of possible characters, and then expand it further by compounding multiple bytes into larger sequences—but the world of email remained stubbornly resistant to these changes. To this day, the technical standard describing an email message, RFC 5322, is built around old-timey seven-bit bytes. That same standard also specifies the formatting rules that define which parts of an email are the header (with the addressing information) and the body (with the user-created communication).

Henry Ford once famously remarked that his Model T car could be purchased in any color, so long as it was black. Email is a similarly robust platform—it can be used to send any message, so long as it is made from the 128-character ASCII alphabet.

Enter Nathaniel Borenstein. As he put it, he wanted to one day have pictures of his grandchildren sent to him by email. In 1992, the idea was laughable. Who in their right mind would wait for an interminable data transmission over a dial-up modem rather than just print a picture and mail it properly?

There was also the minor point that Borenstein was a young man who had no children, much less grandchildren. He had focused on the idea of transmitting photos of grandkids, though, to overcome the perception that wasting that amount of bandwidth on a single message was antisocial. To convince email users to embrace using the medium to transmit larger files, he had to make that seem friendly, homely, and routine.

He also had to find a way to make it work.

The trick was to take whatever file you wanted to attach to an email and re-encode its binary data into ASCII characters, which the email standard would accept and treat as if they were text. Then, once the encoded version was received on the recipient’s side, a reverse process would decode that text back into the original file. Borenstein called it “Multipurpose Internet Mail Extensions,” or MIME for short.

In March 1992, to test the efficacy of MIME, Borenstein sent the first email attachment successfully. Borenstein copied one hundred contacts on this test message, to which he attached two files. One was a photograph of Borenstein’s barbershop quartet, the Telephone Chords. The other was a small audio file of the Telephone Chords’ a capella rendering of “Let Me Call You Sweetheart.” These were the first functional cross-platform email attachments. Previous proprietary systems had offered similar possibilities, but only if the sender and recipient were using the same platform. Borenstein’s MIME was agnostic about the user’s technology, and worked regardless.

As an added benefit, the implementation of MIME types expanded the functionality of email. Messages were no longer confined to the ASCII alphabet, because MIME allowed any eight-bit bytes to be converted into ASCII text for transmission. This opened up the options of using HTML to give email more pizzazz or deploying foreign-language character sets beyond what is offered in ASCII.

MIME also opened up new risks for information security. Not only could email attachments be used to transmit documents or photos, but also executable software. Bad-intentioned actors could send malicious software (“malware”) to their victims, hijacking the recipient’s computer in various ways. To protect against this new scourge, many email providers started filtering attachments to identify and block likely malware.

Providers also filtered attachments to prevent users from overloading the system with large files. The act of encoding binary files into ASCII text adds bloat, meaning that an attachment actually occupies over thirty percent more storage space than the original file itself does. The MIME standard does not place any limits on the size or volume of files that can be encoded and sent as attachments, but in practice most users’ email providers impose limits of their own.

Borenstein devised a way to convert other document types, like pictures, into text representations, enabling them to be transmitted as email “attachments” that could be translated back into their original form at the recipient’s end. Thanks to this invention, he and every other email user today can receive pictures of grandchildren, documents, spreadsheets, videos, and other types of attachments.

 

The views and opinions expressed in this article are those of the author and do not necessarily reflect the opinions, position, or policy of Berkeley Research Group, LLC or its other employees and affiliates.

Read the full article. (subscription required)