Jan 10, 2022

Teleprinters, sometimes referred to as teletypes based on the dominance of the Tyletype corporation in their hayday, are devices that send or receive written transmissions over a wire or over radios. Those have evolved over time to include text and images. And while it may seem as though their development corresponds to the telegraph, that’s true only so far as discoveries in electromagnetism led to the ability to send tones or pules over wires once there was a constant current.

That story of the teletype evolved through a number of people in the 1800s. The modern telegraph was invented in 1835 and taken to market a few years later. Soon after that, we were sending written messages encoded and typed on what we called a teletype machine, or teletypewriter if you will. Those were initially invented by a German inventor, Friedrich König in 1837, the same year Cooke and Wheatstone got their patent on telgraphy in England, and a few years before they patented automatic printing.

König figured out how to send messages over about 130 miles. Parts of the telegraph were based on his work. But he used a wire per letter of the alphabet and Samuel Morse used a single wire and encoded messages with the Morse code he developed.

Alexander Bain developed a printing telegraph that used electromagnets that turned clockworks. But keep in mind that these were still considered precision electronics at the time and human labor to encode, signal, receive, and decode communications were still cheaper. Therefore, the Morse telegraph service that went operational in 1846 became the standard.

Meanwhile Royal Earl House built a device that used piano keyboards to send letters, which had a shift register to change characters being sent. Thus predating the modern typewriter, developed in 1878, by decades. Yet, while humans were cheaper, machines were less prone to error, unless of course they broke down.

Then David Edward Hughes developed the first commercial teletype machine known as the Model 11 in 1855 to 1856. A few small telegraph companies then emerged to market the innovation, including Wester Union Telegraph company.

Picking up where Morse left off, Émile Baudot developed a code that consisted of five units, that became popular in France and spread to England in 1897 before spreading to the US. That’s when Donald Murray added punching data into paper tape for transmissions and incremented the Baudot encoding scheme to add control characters like carriage returns and line feeds. And some of the Baudot codes and Murray codes are still in use.

The ideas continued to evolve. In 1902, Charles Krum invented something he called the teletypewriter, picking up on the work started by Frank Pearne and funded by Joy Morton of the Morton Salt company. He filed a patent for his work. He and Morton then formed a new company called the Morkrum Printing Telegraph. Edward Kleinschmidt had filed a similar patent in 1916 so they merged the two companies into the Morkrump-Kleinschmidt Company in 1925 but to more easily market their innovation changed the name to the Teletype Corporation in 1928, then selling to the American Telegraph and Telephone Company, or AT&T, for $30M. And so salt was lucrative, but investing salt money netted a pretty darn good return as well.

Teletype Corporation produced a number of models over the next few decades. The Model 15 through 35 saw an increase in the speed messages could be sent and improved encoding techniques. As the typewriter became a standard, the 8.5 by 11 inch came as a means of being most easily compatible for those typewriters. The A standard was developed so A0 is a square meter, A1 is half that, A2, half that, and so on, with A4 becoming a standard paper size in Europe. But teletypes often had continual feeds and so while they had the same width in many cases, paper moved from a small paper tape to a longer roll of paper cut the same size as letter paper.

Decades after Krum was out of the company, the US Naval Observatory built what they called a Krum TTY to transmit data over radio, naming their device after him. Now, messages could be sent over a telegraph wire and wirelessly.

Thus by 1966 when the Inktronic shipped and printed 1200 characters a minute, it was able to print in baud or ASCII, which Teletype had developed for guess who, the Navy. But they had also developed a Teletype they called the Dataspeed with what we think of as a modem today, which evolved into the Teletype 33, the first Teletype to be consistently used with a computer. The teletype could send data to a computer and receive information that was printed in the same way information would be sent to another teletype operator who would respond in a printout. Another teletype with the same line receives that signal. When hooked to a computer though, the operator presses one of the keys on the teletype keyboard, it transmits an electronic signal.

Over time, those teletypes could be installed on the other side of a phone line. And if a person could talk to a computer, why couldn’t two computers talk to one another? ASCII was initially published in 1963 so computers could exchange information in a standardized fashion. Bell Labs was involved and so it’s no surprise we saw ASCII show up within just a couple of years on the Teletype.

ASCII was a huge win. Teletype sold over 600,000 of the 32s and 33s. Early video screens cost over $10,000 so interactive computing meant sending characters to a computer, which translated the characters into commands, and those into machine code. But the invention of the integrated circuit, MOSFET, and microchip dropped those prices considerably.

When screens dropped in price enough, and Unix came along in 1971, also from the Bell system, it’s no surprised that the first shells were referred to as TTY, short for teletype. After all, the developers and users were often literally using teletypes to connect. As computing companies embraced time sharing and added the ability to handle multiple tasks those evolved into the ability to invoke multiple TTY sessions as a given user, thus while waiting for a task to complete we could do another task. And so we got tty1, tty2, tty3, etc.

The first GUIs were then effectively macros or shell scripts that were called by clicking a button. And those evolved so they weren’t obfuscating the shell but instead now we open a terminal emulator in most modern operating systems not to talk to the shell directly but to send commands to the emulator that interprets them in more modern languages. And yet run tty and we can still see the “return user’s terminal name” to quote the man page.

Today we interact with computers in a very different way than we did over teletypes. We don’t send text and receive the output in a a print-out any longer. Instead we use monitors that allow us to use keyboards to type out messages through the Internet as we do over telnet and then ssh using either binary or ASCII codes.

The Teletype and typewriter evolved into today's keyboard, which offers a faster and more efficient way to communicate. Those early CTSS then Unix C programs that evolved into ls and ssh and cat are now actions performed in graphical interfaces or shells. The last remaining teletypes are now used in airline telephone systems. And following the breakup of AT&T, Teletype Corporation need finally in 1990, as computer terminals evolved into a different direction. Yet we still see their remnants in everyday use.