1982 was a big year for Atari video arcades — with the release of such classics as Gravatar, Millipede, and Space Duel (complimenting the already massive number of popular Atari games filling video game arcades).

In order to make the management (and, primarily, the accounting) of video game arcades easier — and more future-y — Atari developed and released the “Atari Coin Executive”.

And it is incredibly cool.

The central brain of the Atari Coin Executive was an Atari 800 computer (with 48k of RAM) with a number of accessories, including:

• 2 x Atari 810 Disk Drives

• An Atari 850 Interface Module (which added RS232)

• An Atari 825 printer

• An Amdek 13 inch color monitor

How the Atari Coin Executive worked was both simple… and, at the same time, incredibly cool.

The basic process:

1. A “Coin Monitor” was installed in the coin slot of every arcade game.

2. Each Coin Monitor is connected back to the Atari Coin Executive workstation (that Atari 800) via “telephone type wiring”.

3. The arcade manager can then use that Atari 800 to see how much each game is earning.

Fun fact: The Atari Coin Executive software was open source and written in a combination of BASIC and Assembly. Or, as Atari put it in 1982: “In Basic and 6502 Assembler - Source listings and manual supplied”.  You can find images of the Atari Coin Executive software over on the AtariAge Forum.

In addition to the above mentioned setup, the Atari Coin Executive also included a handheld computer called the “Data Recorder”.

The “Atari Coin Executive Data Recorder” was powered by a MOS 6507 CPU with 16K of RAM (8 2k chips), and communicated with the Atari Coin Executive computer via 300 baud serial. It even had a small built-in printer.

This allowed people to manage several arcades, in separate locations, by:

1. Plugging the Data Recorder into each arcade machine equipped with a Coin Monitor.

2. Then taking the Data Recorder back to the Coin Executive computer and downloading the data into the Coin Executive software.

Finally, here’s a color picture of the whole setup — including the custom desk which was used for the Coin Executive.

Floppy disks. Zip disks. Hard disks.

These sorts of spinning, magnetic storage mediums have been critical to several decades of computers. It’s almost hard to imagine the computers of the 1970s, 80s, and 90s without floppy and hard disks (and other magnetic drives).

But how, exactly, did they come into existence?

Let’s take a quick look at the very first of such devices… and their inspiration.

1946 - The Mail-a-Voice

We’ll begin our journey with the 1946 release of the Brush Mail-a-Voice.

A truly fascinating device, the Mail-a-Voice looked like a phonograph… except it used a paper disc that was coated in magnetic material. You could then record up to 3 minutes of audio on a single paper disk (which would spin at 20 rotations per minute)… and then fold the paper disc up and mail it to someone inside a standard envelope.

Thus the “Mail-a-Voice”.

This device didn’t store computer data itself -- it was only for audio -- but it did inspire engineers who were working on cheap data storage for computers…

February, 1948 - Andrew Booth’s Spinning Drum

In a 1947 trip to the USA, Andrew Booth (who was working on his own computer design), had the chance to see the “Mail-a-Voice” in action.

Since Booth needed a good, inexpensive storage medium for his computer… he attempted to build a similar device using a flat, paper, magnetic disk. What was, essentially, a first attempt at what we would now call a “Floppy Disk”.

Unfortunately, it didn’t quite work. Booth found that he needed to spin the paper disk quite a bit faster in order to make it viable as a data storage mechanism… and he had a hard time keeping the paper disk flat.

In fact, as Booth upped the RPM to 3,000 — which is what he determined he needed — the paper disk itself started to disintegrate. Booth would later comment:

“I suppose I really invented the floppy disc, it was a real flop!”

So he abandoned that approach and, instead, decided to use a brass drum. Why brass? Because brass is a bit less likely to disintegrate than… paper.

This system worked. His brass, rotating drum (with a magnetic coating), had a 2 inch diameter and could store 10 bits per inch.

Yeah. 10 bits. Per inch.

Not exactly massive amounts of storage. But, hey, it was a start! And it didn’t disintegrate! Huzzah!

Improving the magnetic drum

With the first prototype working, Booth set about improving his magnetic, rotating drum storage device. The final version ended up being able to store 256 words of either 20 or 21 bits each (different sources cite different values here and there does not appear to be consensus on if it was 20 or 21 bit words).

In modern terms: This would be equivalent to roughly 5 kilobits of data storage.  Give or take.

This storage drum was put to use on the ARC (the Automatic Relay Computer).

When all was said and done, the ARC could utilize that storage drum to handle 50 numbers and could load a program consisting of 300 individual instructions.

It wasn't exactly a “Hard Disk Drive”… more of a “Hard Drum Drive”.

Either way… Pretty darn cool for the 1940s.

1956 - The First “Hard Disk Drive”

Over the years that followed, this idea was refined and improved. The rotating drum was abandoned for hard, magnetic platters — ones sturdy enough to handle much higher RPMs (certainly much sturdier than paper!)... and thus leading to faster data access.

These improvements eventually leading to the 1956 release of the IBM Model 350 Disk Storage Unit for the IBM 305 RAMAC computer.

The Model 350 Hard Disk Drive, in a base configuration, could store roughly 3.75 MB — all contained in a cabinet 5 feet long, 2 1/2 feet deep, and 5 1/2 feet tall — with platters spinning at 1,200 RPM.

And all thanks to a voice dictation device built for mailing 3 minutes of audio on a folded-up piece of paper.

Kilobytes (KB). Megabytes (MB). Gigabytes (GB).

We use these storage measurements every single, gosh-darned day. And most of us feel like we know exactly what they mean. But do we really?

Do we really — truly — know what a “Byte” is… and its origin? I mean… who came up with the term “Byte”, anyway?

Let’s take a moment to look over the history of the term. If, for no other reason, than to feel smarter than most other nerds.

What is a “Byte”?

If you ask Mr. Google, a Byte is exactly 8 Bits.

Ok. Great. 8 Bits = 1 Byte.

So what is a Bit?

That part is simple.

A Bit is the smallest unit of information for a digital computer. A Bit can have two possible values… 0 or 1. It is a boolean. A binary.

Many people believe “Bit” is short for “Bite”. You find this in many computer history books. This little tidbit has been repeated so often, many believe it. However, like many such oft-repeated anecdotes in computing… it’s hogwash.

In fact, “Bit” is an acronym for “Binary Information Digit”. Squish that phrase together and you get “Bit”.

Fun factoids about the origin of the “Bit”

The first usage of the word “bit”, when talking about a type of data in reference to computing, was by Vannevar Bush. He published an articled entitled “Instrumental Analysis” in the October, 1936 issue of “American Mathematical Society”. In it he used the phrase “bits of information” when talking about punch cards.

However… 

“Bit” was commonly used in Middle English to refer to “a mouthful” or a “morsel” of food. (This is the origin of why many believe “Bit” is short for “Bite”… even though it isn’t.) As such, Vannevar Bush may not have actually been thinking about a “Bit” as a “Binary digit”… instead he may simply have thought “this is a morsel of data”. Also worth noting… Bush never actually defines what a “bit” is. Making it likely that he was simply using the word “bit” in the Middle English way.

The first — distinctly verifiable — usage of “Bit” in this way is by John Tukey. From “A Mathematical Theory of Communication” written by C. E. Shannon in 1949:

“The choice of a logarithmic base corresponds to the choice of a unit for measuring information. If the base 2 is used the resulting units may be called binary digits, or more briefly bits, a word suggested by J. W. Tukey. A device with two stable positions, such as a relay or a flip-flop circuit, can store one bit of information.”

There you have it. More information about the origin of the term “bit” than you ever wanted to know.

You’re welcome.

Ok. Great.

So, in short, a Bit is a 0 or 1. And a Byte is a group of 8 Bits. Easy.

Not so fast there, sport!

While the Byte being 8 Bits is commonly accepted today… that was not always the case. Not by a long shot!

In fact, there are two competing stories for who created the term “Byte”… and neither of them were referring to a set of 8 Bits!

Seriously!

Werner Buchholz’s 6 Bits

The most often cited creator of the term “Byte” is Werner Buchholz — who used the term, in 1956, to refer to a grouping of 6 Bits when working on the IBM Stretch Super computer.

A “6 Bit” Byte was common in those days. In fact, Braille was a 6 Bit encoding of characters for the blind. And many of the early computers (from IBM and others) used 6 Bit groupings to encode character data.

6 Bits -- not 8 Bits -- per Byte.

However (you knew there had to be a “however”)…

Louis G. Dooley’s N Bits

Around that same time (1956 or so), Louis Dooley first used the word “Byte” to refer to an undefined grouping of “Bits”. But, typically, used as “4 Bits”.

That's right.  Not 8 Bits.  Not 6 Bits.  But 4 Bits.

Dooley published the following letter in BYTE magazine:

“I would like to get the following on record: The word byte was coined around 1956 to 1957 at MIT Lincoln Laboratories within a project called SAGE (the North American Air Defense System), which was jointly developed by Rand, Lincoln Labs, and IBM. In that era, computer memory structure was already defined in terms of word size. A word consisted of x number of bits; a bit represented a binary notational position in a word. Operations typically operated on all the bits in the full word.

 

We coined the word byte to refer to a logical set of bits less than a full word size. At that time, it was not defined specifically as x bits but typically referred to as a set of 4 bits, as that was the size of most of our coded data items. Shortly afterward, I went on to other responsibilities that removed me from SAGE. After having spent many years in Asia, I returned to the U.S. and was bemused to find out that the word byte was being used in the new microcomputer technology to refer to the basic addressable memory unit.

 

Louis G. Dooley
Ocala, FL”

So… what the heck is a “Byte”?!

That’s right. We now have two very, very different definitions for the word “Byte”. Both creations of the word happened independently… and at almost the exact same moment in time.

• “The Buchholz Byte” - A grouping of 6 Bits.
• “The Dooley Byte” - A grouping of an undefined number of bits, less than a full word size. Often used to describe 4 Bits.

You’ll note that neither of these definitions — from the men who created the term — have the number “8” in them.

The shift towards 8 Bits per Byte started to happen in the 1970s… with the development and gaining popularity of 8-Bit processors, such as the legendary Intel 8008.

Interestingly, some of those early 8-Bit CPU’s had specific functions for handling 4-Bit chunks of data. Because, up until that point, 4 and 6-Bit “Bytes” were incredibly common (including in the predecessor to the Intel 8008… the 4-Bit Intel 4004).

Fun Factoid: Nowadays a 4-Bit group is called a “Nibble”. Which is adorable.

For quite some time the term “octet” or “octad” was used to denote 8 Bit groups. At some point along the way, most people phased that out as well… simply referring to all “groups of bits” as a “Byte”. Though you will still find “octet” used here and there, especially when talking about various network protocols.

All of which means…

Dooley invented the modern “Byte”… not Buchholz

While many writers, enthusiasts, and computer historians are quick to say that Werner Buchholz coined the term “Byte”… they are obviously mistaken.

Besides the fact that it’s hard to discern who (Dooley or Buchholz) actually used the term first… the Buchholz definition is no longer used at all in modern computing.

The Buchholz definition is specific. 6 Bits. Which modern computing has determined is not the amount of Bits in a modern Byte.

The Dooley definition, on the other hand, allows for wiggle room. Which means that an 8 Bit “Byte” would fit the Dooley definition. But not the Buchholz.

The facts are clear: Louis G. Dooley created the word “Byte”. At least as it has been used for the last 40+ years.

But Buchholz — an absolute legend in the computing world — gets one heck of an Honorable Mention trophy.