Advent of Computing - Episode 151 - The Friden Flexowriter

Episode Date: February 3, 2025

Have you ever looked at an old computer and seen a weird typewriter thing tacked on? In most cases that's a device called a Flexowriter. It's half electric typewriter, half teleprinter, half tape reader, and all business! This episode we are chronicling the rise, fall, and weird business dealings of the Flexowriter.  

Something I often think about is how we view computers. Now, I mean that very literally, as in, how do we view what a computer is doing? What's our window into that digital world? It's one of those fundamentals. A computer on its own, unadorned, sure, it's a thing of wonder. It's full of clacking relays or glowing tubes or little magical crystals. That's neat and all, but you need some way to communicate with it. You can't just scream at the wall of tubes and expect an answer.

You have to have some way to see what it's thinking, give it orders from on high, and maybe get a little feedback. This window dictates so much. Maybe get a little feedback. This window dictates so much. It dictates how fast you can access information, or set up new instructions for the computer. A monitor and keyboard is faster than a paper-fed terminal, which in turn is faster than punch cards. It dictates how easy a computer is to understand. Blinking lights and switches may be simple to to build and pretty cheap, but they can't

hold a candle to just a normal keyboard. On a more subtle level, your interface dictates how you can express ideas, and sometimes even what kind of ideas you can express. This is especially the case when we're talking about programming. Input and output devices always have some set of characters they understand. For more modern devices, that character set is truly massive. Your window can show off all kinds of complex ideas.

But for older devices, it's more hit or miss. What happens if you only have parentheses but no square brackets? What if you have no quotation marks? What if a line of text can only be 10 characters long or what if you don't even have spaces? That's all bound to impact how you see the computer and what you can tell the computer to do It's easy to assume. This is one of those linear progression things, right? Our window starts out small and cloudy,

but it gets larger and clearer every year. That's not really the whole story though. Rather, things bounce around for decades. That's especially true in the earlier eras of the computer. And it gets all the more complicated once we look at the specifics. Welcome back to Advent of Computing. I'm your host, Sean Haas, and this is episode 151, The Freedon Flexor Rider.

That name may not sound familiar to you, but I can promise it's at the center of a wild number of stories. The FlexoRider was this very specific kind of teletype machine, or teleprinter. It was used by many early computers. The Harvard Mark I, one of the first computers, used a FlexoRider. Whirlwind, the first interactive computer, used a flexo. Dex PDP-1 shipped with a flexo rider bolted to a desk. The LGP-30, one of the first steps towards personal computing, also had a flexo rider bolted on top of it.

We've touched on the device before, but never really taken the plunge. This episode, we'll be doing just that. I just want to flesh in some parts of the larger story that we've ignored a little bit in the past. The FlexoRider develops from a much older lineage. In general, the hard copy terminals used with early computers, that is, terminals that type onto paper instead of to a screen,

they all evolved from telegraph equipment. I want to trace that lineage through the development of the flexo. The other big thing to investigate is the staying power of Frieden's device. The flexo is used for quite a long time. We're talking over a decade on the conservative side. So what's with that staying power? What made the flexo so much

better than say some other device? Was there even a difference? And why exactly is it flexible? You might notice this is a bit of a short introduction that's because I'm kind of going in blind. I honestly don't know where to start or even where this is going, so we're going to have a bit of a fun and wacky ride. But first, I have some announcements I have to make. The first one is on February 8th, I'm going to be speaking at the Intelligent Speech Conference. This is an online conference where some of your favorite podcasters get together to talk history. This year the theme is deception, so all the talks are centered around deception of one

kind or another. I'm going to be doing a talk and also sitting on a roundtable discussion. If you want to attend that, well, you can do it from anywhere. It's all remote. You can go to IntelligentSpeechOnline.com and pick up tickets. And if you use the code AOC, that's AOC for Advent of Computing, you get a bit of a discount. Now, announcement two is on the 15th and 16th of February.

I'm going to be at VCF SoCal in Orange, California. It's actually just down the road from Disneyland. There, I'm going to be giving a presentation on some of my research. I'm going to be giving a presentation on some of my research. I'm gonna be doing an interview on stage. I'm hosting a panel and I'm sitting on at least one panel. So it's gonna be a busy weekend for Sean, to say the least. If you wanna see any of that or come stop by and say hi

or just go to VCF, they're great events, then you can find information about that conference at VCFsocal.com. Now, announcement three is... I'm taking a little break. Well, I'm taking a break from publishing the show. I'm gonna skip next episode because I'm gonna be very busy for the first half of February. That means after this episode there'll be about

a month delay before the next episode drops in the public feed, but I will be back. I should have a lot of stuff in the tank to keep going. And I got some plans lined up for when I get back, so never despair, the show always goes on. And with that, I think it's time we actually start talking about the Flexo Rider. Allow me to open the discussion with the first twist. The story actually includes IBM. It includes a lot of IBM. Now, I sometimes joke that IBM is behind every episode, but this time I did not suspect it at all. As such, there's a bit of a mysterious core to this story. That is antitrust, or at least the allegation of antitrust. I think I've been able to piece things together,

but just be aware, the timeline is somewhat shaky because of Big Blue. We begin in 1933 when IBM buys a small company called Electromatic and just to point it out this is one of those fabulous early 20th century names we don't really name companies nearly as well anymore. Now this was done for one simple reason. IBM wanted to get in on the electric typewriter business, and Electromatic had a particularly good typewriter. Electric typewriters are strange devices because they eventually bundle up two different technologies, one for labor saving and another for data transfer. Now of course that gets all the more tangled because technically the lineage of electric

typewriters go back to, well, telegraph equipment but let's digress a little bit. Let's start with the first of these technologies and then get to the second. Mechanical typewriters, if you've ever used them, take a little bit of manual force. When you type a letter, you're providing all the energy needed to hammer that letter into the page and then advance forward. That, according to Electromatic's own brochures, is a big problem. Quote, Electromatic eliminates typists fatigue, pains in the shoulder blades, tired wrists, afternoon exhaustion, and nervous headaches are relieved when the change is made to Electromatic. The ease of operation of the Electromatic is a revelation to the typist.

The response of the Electromatic to every touch of the key, even the lightest, is so swift, so smooth, so rhythmic, and so efficient, that the machine seems actually alive. 99% of the usual manual labor and effort of typing is transferred from the fingers of the typist to the small electric motor. Fatigue is reduced to a minimum. Once again, we just don't make companies like we used to. The core idea here is to use a motor instead of the operator's hands as the source of energy. The keys are no longer physical linkages to some type head. Rather, they're buttons. You hit a button, some circuit figures out which letter that is, then the motor provides the power necessary to stamp the letter into your paper.

In that way, an electric typewriter is more like a modern keyboard. You use the keys to issue commands, then something else acts on those commands. You're sending signals to some other machine, just so happens to be bundled together. That same motor, through a series of clutches, can also handle the carriage. That's the part that actually holds your paper and moves as you type. Usually a carriage is manually operated. When you reach the end of the carriage, there's a ding, and then you slam it back to the beginning. If you hit a new line key, then you have to manually return the carriage to the start

of the next line with, well, a bit of a slam. The Electromatic automated that part of the process for you. Now that's one aspect of the electrification of the typewriter, but it's also just the beginning. As we all well know, when you introduce automation, you start down this slippery slope. The keyboard now issues commands, the direct linkage is gone,

so what's stopping you from issuing commands from some other device? Well, there was already precedent for that, the Telegraph. During the 19th and 20th century, telegraphy had rapidly developed. By the 30s, we had somewhat practical teleprinters. These were, for all intents and purposes, primitive computer terminals.

You had a keyboard which, at a press, would send out an encoded letter over a telegraph line. They also worked in reverse, receiving encoded data over a line, decoding it and printing it onto a sheet of paper or a spool of tape. The key difference between that and the computer terminal is we didn't really have computers yet. These were used for sending messages or in some cases with some very specialized machinery.

Electromatic made use of part of this technology, namely the paper tape. The Electromatic wasn't exactly a teleprinter, but it could be rigged up to play a similar trick. Using a tape reader punch combination, it could read in encoded data and hammer the characters onto a page. It could also go the other way around,

allowing you to encode data on a tape at the press of a key. We're talking about paper tape here, specifically using five hole encoding that was common in telegraphy at the time. IBM buys Electromatic when their typewriter is up to this level of complexity.

It gets reworked, repackaged and relaunched buys Electromatic when their typewriter is up to this level of complexity. It gets reworked, repackaged, and relaunched as the IBM Model 01 in 1937. The acquisition opened up new doors for Big Blue. The Electromatic wasn't just a new product. It could also work together with many existing IBM machines if... you did a little bit of fiddling. You can turn a Model 1 into a printer, for instance. Remember, the key presses are just electrical signals.

It's all electronic. So you can take the bundle of wires that the keys use to send commands, put it on a big plug in the back of the machine, and then hook that plug up to, say, a punch card tabulator. Now you can print cards from punch cards onto a sheet of paper, or go the other way around. Send those wires out into a card machine,

and boom, you can now type onto punch cards. All you really need are a few circuits to glue everything together. And then, boom, you have an interface. This is where we get to our next big player, the National Postal Meter Company. This is also where I have to delve into the world of old newspapers to try and scrape together the story. The Postal Meter Company is founded in 1935 in Los Angeles and moves to Rochester, New

York a few years later. They produce machines for metered mail. You know, metered mail, what we use every day. Jokes aside, it's those letters that don't have stamps on them. It's the ones that businesses send out. So the postal meter company is quite literally making business machines for big businesses. That puts them dangerously close to another larger fish. What other purveyor of business machines lives in Rochester, you may ask? Well, it is IBM, my friends. It's hard to be that close to the orbit of Big Blue without some kind of strange corporate

machinations occurring. In 1941, one of IBM's vice presidents is elected president of National Postal Meter. This is one of those mega-cursed business things. IBM has technically always had multiple vice presidents. This is so more IBM employees sound powerful. They could dispatch a vice president in a blue suit at literally a moment's notice. Maybe a tough customer doesn't want to deal with a regional head of accounting equipment and machinery, but a vice president of IBM. Well, that could break a negotiation wide open.

In this case, IBM sends out one of their best to head another company. This brought NPMC into the IBM fold. But to what end? What is the purpose of this corporate machination? Well, that's where things get confusing and quick. It's possible that this was part of some bigger plan, some deep trick that IBM was trying to pull.

I'm going to press on with that theory because, well, there's not a whole lot of information to fill in the gaps. In 1944, NPMC is renamed CCC, the Commercial Controls Corporation. In 1946, IBM spins off a company called, and get this, this is a real name, Gistowrider. This new company takes over some manufacturing in Rochester. They are physically transferred control of a plant in that town. They also take over development of electric typewriters and are given a trove of key patents for that technology.

The board and management of Gisto Rider is constituted of IBMers. This is, as near as I can tell, some kind of shell game. There's something shifty going on. Gisto Rider kind of just exists for less than a year. They go to trade shows, they actually have a softball team that does somewhat well in some regional games, and they have stock. In December, they are, in turn, bought by CCC, which, I remind you, is controlled by an IBM man. Further, at this point, IBM owns 40% of CCC stock.

Now to be on the board, or to be a president of a company, you usually have to vest into shares of that company. That means that between the 40% stock of CCC that IBM as a corporate entity owns and the stock that we can assume the president of CCC, an IBMer, owns, the company is just an IBM cutout. It's wholly owned by people in blue suits. So when this deal was signed for CCC to buy just a Rider, well, it must have been IBM on both sides. Good ol' TJ Watson just walked into an empty room and shook his own hand, I guess. We get the most information about Justo Rider from a few articles about its sale.

It's like only learning about a relative through their obituary. This is also where we get the first name drop. Gisto Rider, the very real company that totally existed for normal reasons, and don't worry that we had stock, we were actually planning to exist for decades, well, they manufacture this thing called the Flexo Rider. It's described as an automated letter writer. But what exactly does that mean? Well, this is another trick that can be pulled off by modifying an electro writer, or when I be a model or whatever you call it, the names

are going to continue piling up. Don't worry. You can use a flexo writer to print directly to paper tape, recording every keystroke. You can also add in a special character called Stop. You set up the reader so that when the Stop character is encountered, a bell rings and the Flexo waits for user input. You've just created an automatic letter writer. Dear, stop. We regret to inform you that stop has stop. Kind regards, stop. This was supposedly used during World War II to write death notices to families of soldiers. Now, I've seen that said a couple of places, but I've seen no sourced information about that.

I believe the story, except that those letters were sent out as telegrams. Now, the Flexo is very flexible, but without sourcing that explains exactly how this was set up, I don't know if I 100% believe it. Technology aside, what exactly is going on here? Why did IBM just split off JustoRyder and then use CCC to buy it up a few months later? In summary, to make a long story short, I have no idea. I've seen two guesses as to why this happened. One is that TJ Watson did not like typewriters. Now, I have no way to prove that, that's just random speculation I've seen in some interviews. The other guess I've seen is that this

is somehow antitrust related, and that's the more commonly stated line. Although to be fair, this isn't exactly a commonly discussed story. And sure, we all know that IBM has had issues with antitrust, but I don't know if Jisto Rider is an example of that. The whole Jisto Rider debacle happens smack in between two very high-profile antitrust cases. One started in 1932 and finishes up in 1936, and the other starts in 1952 and, funnily enough, resolves in 1956. The first ends in a consent decree which forces IBM to allow third parties to make and sell punch cards and other consumables.

The second has to do with leasing and service practices around accounting machines. The flexo doesn't really fit into the timeline here. I think one of the reasons for this speculation is that in some cases and actually very many cases, antitrust decrees will say that a company has to stay in one specified market. That's what happened to Bell. That's the reason Bell couldn't sell Unix directly. They had an antitrust ruling that says, hey, you got to pick market and stay in it, and they chose telephone equipment. Unix wasn't telephone equipment, so they couldn't sell it. I think some people online might be convoluting these two different types of antitrust actions.

IBM's never had a consent decree that says, hey, you guys can only make accounting machines. You have to split off everything that's not accounting related. You guys can only make accounting machines. You have to split off everything that's not accounting related. But maybe I can spin a yarn that could make some sense. Now, we do see in 1969 that IBM is willing to be proactive about the Sherman Antitrust Act. In 69, ahead of yet another antitrust trial, IBM preemptively unbundles services, hardware,

and software. The practice of bundling would be at the heart of the upcoming case. IBM knew it because they had been told by the Department of Justice they were being sued in the very near future about their bundling issues. In this instance, IBM took a step. They adjusted to prepare for the upcoming trial so they'd have a better position going into proceedings. Perhaps if we go out on a very long limb here, IBM had a spidey sense.

They felt an upcoming antitrust case in the air, so they tried to zig or zag. Either the maneuver worked and it prevented some theoretical USVIBM 1947, or someone read the wrong tea leaves. Or maybe they were just trying to separate concerns and keep a separate product or arm's length from the main company. This kind of stuff would have been buried in memos or meeting minutes that we just don't have access to. Or maybe TJ Watson just didn't like typewriters. The point is we don't know exactly why IBM does this maneuver, but I don't think that's

likely to change unless something serious happens. Once we reach 1947, the Flexo writer name starts showing up more publicly. One weird wrinkle here is that CCC is still selling machines under the Gisto writer name for a number of years. Ads and documents about the early Flexo writers are on letterhead that can come in a few possible forms. The flexo is initially billed as a labor-saving device. From a 1947 issue of the magazine Office Appliances, quote, of individually typewritten matter is made possible by its operation, the Gistowriter Corporation, one Layton Avenue, Rochester, New York, has introduced the new Flexowriter automatic typewriter. The manufacturer

asserts that this all-electric, completely automatic machine enables one stenographer to individually type as many as 600 letters in one eight-hour day, depending on the number of units employed." And just as an aside, I really love all these old niche trade journals that existed. There's something about the idea of subscribing to administrative management quarterly, or office appliances that just tickles me pink. Anyway, early on, Gisto is pushing the form letter aspect of the Flexo writer. It works just as I described.

A typist prepares a tape, adding in special stop commands where customization is needed. Then you run the tape through your trusty Flexo and fill in the blanks. One magazine calls it a quote, stenographer's assembly line, which sounds pretty accurate. Then how do you make the jump to the computer? How did the Flexo go from a stenographer's assembly line to a window into the digital realm? assembly line to a window into the digital realm.

The history of human computer interfaces is complicated. It's also pretty messy and confusing. When I went snooping for the first interactive text interface about a year ago, I was quickly, let's say, stymied. It all comes down to definitions. In the modern day, when you say computer terminal, that has a very specific definition. It's an interactive text interface, where a user types in text, and the computer sends back text to display instantly. Or, at least pretty fast. There are a lot of moving parts here. You have

input, the keyboard, you have output, the screen these days, but that was once a roll of paper. You have communication from keyboard to computer, from the computer to the output device, and then you have some kind of processing that happens in between. Because of that middle step, the processing part, many early interactive interfaces show up tacked onto other projects. Early research into Lisp led to an interactive interface because that was convenient and the hardware to do so was sitting around in a couple pieces in the right lab at the right time.

The interactivity, that's actually a combination of a handful of different technologies. Call them tools or pieces to a larger puzzle. Input, output, and communication. The FlexaWriter shows up at the dawn of computing, but it's not always used in its entirety. Its flexibility made it useful in almost a hacky sort of way. This all starts in 1944 with the Harvard Mark I, also known as, get this, the IBM automatic sequence control calculator. This, once again, is a surprise IBM episode, so I guess this is only fitting. The Mark I used an Electromatic,

one of those primitive proto-Flexo-riders, as a printer. In fact, it had two of the things. The reason for this is a little funny. So, all these early machines developed during World War II, at least in America, were meant to calculate artillery firing tables. The Mark I had printers so it could directly print out those tables. It ends up being used for much more, but that's the official reasoning. The setup here was primitive, as to be expected from the era. The two typewriters were wired into a bus via a plugboard,

which could be used to configure how the typewriters worked. The Mark I had a number of instructions for turning the typewriters on and off, and printing numbers on each. We have good evidence of these electronics being used as output devices because we have the original printouts. We actually have books of firing tables that were printed on these very typewriters. From what I've read, this was all the typewriters were used for on the Mark 1. They weren't wired up as inputs and they didn't have the tape punch extension. The Mark 1 read programs off tape, but it used a special tape format that wasn't compatible with the Flexo's 5 or 6-bit tape. The typewriter is just

being used since computer printers didn't really exist. It was a convenient tool that IBM already had access to, so they just slapped a couple onto the Mark 1. In 1950, just a few years after the advent of the electronic digital computer The flexo rider was already entrenched for instance the National Bureau of Standards used a flexo with their CAC and SWAC computers an MIT used one with whirlwind in fact the fact, the flexo on Whirlwind may have been the first interactive text interface, either that or some other flexo writer software written

at MIT. The big change we see in the early 50s is the use of the flexo as a two-way interface with a computer, as in, the key press signals are now being sent back into the computer. This is a substantial shift. At first, the Flexo is really just used as a tool of convenience, a machine on hand that can be used to cobble things together. But increasingly, the Flexo becomes the glue that holds computer rooms together entirely. Let's use the LGP30 as an example, since that's the case I know best.

The LGP30 is a desk-sized machine, nearly personal in its aspect. The machine starts selling in 1954. It has a flexo rider physically built into it. The flexo is hardwired into the LGP 30 and bolted on top of the computer's case. Here it's used to full effect. The printer is used as the computer's primary output, the keyboard as an input. It also makes complete use of the tape reader rider. That's the component that was initially intended

for form letters. Well, since you're already hooked into a flexo, you can reroute signals to and from the tape. You could use it to store and replay commands for some lgp30 programs, but that's boring. The real hotness here is that the tape can be used for more than text. You can use it for raw data. The LGP-30 loaded programs from this tape, and it could also output data to the tape. How's that work in practice? Let's say you want to compile a program on your LGP-30. You would first prepare a tape on the Flexo Rider. This was done in offline mode,

meaning that the Flexo wasn't connected to the computer. The computer didn't even need to be on. This was set up by flipping a switch. You could just type on the keyboard and your source code would stream out to the paper tape. When you were done, the LGP-30 would be switched on and the Flexo would be moved into online mode. That would actually hook it up

to a bus inside the computer. Once booted, you'd pull out your compiler. This, of course, was stored on another reel of paper tape, which had to be fed through the tape reader on the flexo writer. Once that was loaded, the computer would prompt you to load your source program. Same deal, you grab the tape you just prepared, run it through. Then to finish things off, you grab some blank tape, and the LGP-30 would punch out your binary program. That's a bit of a process, but hey, programming back in the day was a lot more tedious. You can see from this process that the Flexo is really the heart of the operation.

It's how you write software, it's how you get output from the computer, it's how you feed input into the machine, and crucially, it's how all your prep work is done. It's not just a peripheral that only works with computers. The Flexo is its own independent machine. In fact, there's accounts of LGP30 labs having an extra flexo writer so someone can program while someone else is using the computer. While that may sound a little strange nowadays, this was very much in line with older practices. On punch card machines like old IBM mainframes, you'd have complementary independent devices that were only loosely connected. You'd use a

key punch to prepare punch cards which wouldn't even be connected to the computer. You had a machine for transferring cards onto tape or data from tape onto cards. Sorters and tabulators also survived into the computing era and those had no physical connection to the mainframe, not directly. In this context the flexor rider fits into a larger picture of supporting machines. A key difference though is that the flexor rider was connected directly to the computer, at least often. That made it especially powerful as a tool.

It was, quite literally, the window into the digital realm. We're still missing something, though. We're still missing a name. Where does Freedon come into the picture? What is a Freedon, anyway? Whenever you hear the name Flexorider, it's almost always written as the Frieden Flex-O-Rider. Well, I think this is the right spot in the timeline to discuss that.

Frieden is short for the Frieden Calculating Machine Company, which in turn is named after one Carl Frieden, who established his company in 1933. Now, Carl's story is a little wild. It's one of those tales that only fits in the early 20th century. It gets stuck down there with names like Electromatic. A hagiography of Carl is preserved at his fansite, Friedenites.com. I love that name. Freedon was born and raised in Sweden. An engineer by trade, he got his first big job working for the Swedish Match Trust. There he found himself traveling around Europe, setting up matchmaking factories which were full of intricate machines. On

June 27th, 1914, he landed in Australia on assignment to get another factory up and running. While he was on a ship traveling south, something concerning happened in a small town named Sarajevo. Namely, a man named Franz Ferdinand was shot. The day after Karl arrived in Australia, war broke out in Europe. He was effectively stranded. He pretty quickly gave up on the whole factory idea at that point. Karl effectively lost his job the day war was declared, because the Swedish Match Trust was blacklisted in Australia. It had something to do with embargoes and political connections with Germany. Frieden would go to ground, take a job in a factory to get

by, and began the long process of seeking transit through a war zone. As Friedenites describes, this is when Karl set to work on something grand. During his college days, he had taken an active interest in mechanical calculators. He now had plenty of time on his hands, so he started scheming, planning, and building. In 1916, he was permitted passage back to Europe via America.

He would set sail with a prototype mechanical calculator. His ship would first stop in California before he would, ostensibly, cross the country by rail and pick up a boat to sail the Atlantic. But he decided to just stop and stay in California. The reason why? Well, according to what I've read, it was an offhand remark by a fellow traveler. Someone on the boat told Carl that, you know, you could make a lot more money selling this new-fangled calculating machine in America than you ever could in Europe. Carl took that advice, apparently. Once in California, he got a job as an engineer at the Marchant Calculating Company. There, through a series of wacky early 20th century happenstances, he would design many

calculating machines. Carl made a wild amount of money selling and licensing his patents to Marchant and down the line to third parties. He was doing very well until the Great Depression struck. Carl left Marchant during the Depression, and after a little bit of work, he founded Frieden Calculating Machine Company. Basically, he took his trove of patents and struck out on his own. Many of the successful machines sold by Marchant had come from the mind of Frieden himself. So, when the company was set up, he was on a route for success. In 1956, after 20-some years of selling desktop calculators, the Freedon Corporation bought

CCC for $7 million. From what I can tell, Carl was never employed by IBM. Now for that fee, the Flexo becomes the Freedon Flexo Rider. But why buyout CCC? Well, we don't know. It's a business thing and was only really reported in business articles. Whatever the reason,

it seems to have been a very good financial move. In the next years, Freedon stock skyrocketed. Business was so good that the Rochester plant, previously CCC itself and previously IBM, had to be expanded. We can infer this was due, in large part, to the FlexoRider. Initially, Frieden just spits out FlexoRiders in addition to their existing calculator lineup. By this time, the Flexo is commonplace in computer rooms and in offices around the country. If you aren't

using batch processing then you're probably using a flexo. One early freedom ad sings the praises of this device as the glue to any operation. Quote, flexo writer automatic writing machines, auxiliary tape punching, reading, duplicating, and conversion units are used for input and output in a wide variety of data storage and processing machines in business, industry, government, and research. Detailed program tapes for actuating and instructing various types of data processing machines can be punched by the FlexoRider as the data is typed in proof form. These tapes are used to accurately direct or program the tape-operated equipment." The flexo was really hitting its stride. It was being used

aggressively in many applications, not just to print form letters. In 1959, the PDP-1 hits, with an integrated flexo as one of its main interfaces. It's also used to control CNC machines and a lot of industrial applications. But it's with this computer lineage that we start to see recognizable text interfaces. That's the ultimate mark that the flexo leaves on the world. In many ways, it's the original computer terminal. But that's not the only trick Frieden had up their sleeve. This is the ultimate flexible device after all.

As such, it's used for more than just computers. Allow me to introduce the CompuTyper. Remember, the Flexo is just a component. It's the glue that makes a lot of things possible. George Comstock, a Frieden employee, describes the CompuTyper like this in an oral history interview. Quote, "...one of the children of that marriage was a line of products called CompuTypers,

which were a combination of Frieden merchant calculators as an arithmetic unit and the FlexoRider as an I.O. and printing device. And the CompuTypers were used for simple billing accounting applications in small businesses. In a sense, they were competing with the bottom end of the Burroughs line of accounting machines. But it was a successful product line. They were selling several thousand CompuTypers a year. It was was a successful product line. They were selling several thousand CompuTypers a year. It was really a pludge. I mean, the calculator portion of it was rigged up with solenoids to punch the buttons, and with rotary contacts making pick-offs to read

the contents of the accumulator out of relays." Early on, the CompuTyper was essentially just a reading calculator wired up to a flexo. Like Comstock says, a clutch, but a smart one. This could be used to fully automate accounting using a single machine. IBM could sell you a whole setup to do the same thing, but that requires a lot of different moving pieces and a lot of money. This is, once installed, a single unit. A pretty cheap one, supposedly.

Brochures describe the CompuTyper automating away invoices, inventory, and the usual form letters. It could tabulate data and even keep running variables. This is where I have to say that we know very little about early copytypers, as in, we don't even have manuals for their early machines. Really just ads, some articles and newspapers, and some later remembrances. What we do know is that the Model A, the first in the series, was at least somewhat electromechanical. Brochures describe a plug board, and if that was similar to the boards IBM was using at the time, then the first copytypers weren't so much programmable as configurable. You would have been able to link up registers and columns to inputs and outputs. It would have been

primitive, but perhaps pretty effective. As the copytyper line continued, it led to fully fledged computers. Freedon was one of the first calculator companies to fully electrify and the copytyper followed suit. By the 60s, they were shipping actual silicon. It sounds to me like we're dealing with that elusive idea of an almost personal computer. Apparently there were even some multi-user copy-typer systems. Perhaps this is a tangent to flesh out another time. There is one more aside that I want to mention

just because I don't think I'll ever get another chance to. The advertising around the Flexo writer is a little strange in all of its eras. The Flexo was sold, on label, as a clerical machine. It's meant to help with office and secretarial work. Most of the advertising I've seen for the Flexo is selling it specifically for female office workers. Ads always show women operating Flexos, or they show a female office worker next to a

pile of flexos. There's this one old ad that I keep kinda chuckling at from the CCC days that shows a woman sitting surrounded by an entire ring of flexo riders. This carries forward into the CompuTyper. One brochure from the early 60s opens with, quote, For the average business, one girl and one CompuTyper can be the entire billing department. What's more, the CompuTyper can do many other basic paperwork jobs. Ask your local Freedon Systems man to show you how your own applications can be

handled at this automated work center." This is kind of funny to me in context because the 60s is when programming starts to shift towards being a male-dominated field. That's also true for computing at large in this time period. There's this larger story of the perception of programming and programmers shifting over the middle of the century. Initially, programming as a profession is dominated by female programmers. It swings towards being almost completely male-dominated, and in the modern day, it started to shift towards being more egalitarian. But in the 60s, this is where the pendulum is very much in the modern day, it started to shift towards being more egalitarian. But in the 60s, this is where the pendulum is very much in the other direction. That makes it really interesting to me to see Frieden marketing the coffee typer,

something that is very much a programmable computer, explicitly as a machine for women in the office to use. There's some in-depth cultural analysis that could be done here, but I'm not quite the right person for the job. I just want to point this out because it's something that I clocked as interesting and maybe a hook for someone else to look into. So let's get back to the main thread. By the 1960s, the Frieden-Flexo writer is fully formed and it's everywhere you find

computers. But there's another player. The teletype ends up becoming much more popular. I mean, in the modern day, we still sometimes call computer terminals teletypes. I've never heard someone call a terminal device a flexo before. So what happened to the flexo between the 60s and now? To answer all these questions, I should introduce the teletype proper. A teletype is a device that has a keyboard and printer.

It can send and receive digital signals. Any signal it receives is printed, and any press of the keyboard can be sent out. That should sound very familiar at this point, so let me untangle something for you. Both the teletype and the flexo rider are different brand names for a teleprinter. Teleprinter is just the generic term for this whole class of devices. We already know where the Flexo name comes into play. The name Teletype comes from the Mörkrum Kleinschmidt company, which started producing teleprinters by that name in the 1900s.

The company would pretty quickly change its name to match their best-selling product. They're better known as the Teletype Corporation. So the story of how we all end up talking about teletypes is a classic tale of competition. There are a few possible reasons that teletype beat out the flexo, so let me run down the list. According to Nicholas Bodie, who once worked for Frieden, the differentiator may have been cost. This is one of those times where someone says it was a matter of cost, and then doesn't provide concrete numbers tied to years. Honestly, whatever. The flexo rider had a reputation for being rugged, powerful, and I hear a little expensive. Bodley isn't the only

person pointing this out. I'm also gonna be one of those dudes that just says it was expensive without any numbers because I can't find well-attested flexo prices in these few years specifically, but we'll get back to numbers later. In 1963, the Teletype Corporation released the Teletype Model 33, and it cost $1,000. That's a little over $10k today, so not exactly a cheap machine. It was a teleprinter. It had a keyboard and a platen for paper printouts. It could punch and read tape. It could wire up to a serial bus. It could do everything the Flexo could do. And it, presumably, cost less.

There was another reason to buy Teletype over Freedon. In 1963, the first ASCII standard was released. It specified 7-bit character encoding which, with a few revisions and expansions, is still used in the modern day. The Model 33 supported ASCII. Specifically, it did 7 bits plus an extra control bit, giving it an 8-bit data stream. All very nice and digital. The flexos of that time used either the older 5-bit Baudo code or a 5 or 6-bit custom data encoding. Now, Baudo was a standard developed for telegraphy. That was fine for many applications, but it was very limited. You can only have 32 different characters in the Baudot system. ASCII hits the scene supporting 128 possible characters. That made ASCII a much

better choice for computing in general and for programming specifically. Baudot code only has a handful of special characters, only parentheses for grouping and only the most basic math symbols. Programming in Baudo would be challenging. It restricted what could be expressed. There were ways around this. Many computer-bound flexos were customized, giving them a more useful character set. But still, there were some limitations caused by the older, smaller

encoding scheme. That's not to mention the lack of the one big word we all love, standardization. The Flexo on a PDP-1 had a different character set than the Flexo on an LGP-30, and was wired differently than the Flexo on a Comp copy typer. You couldn't just grab a stock flexo writer and hook it up to a computer. At least not in 63. That would only work if your machine spoke in strict Bado or if you got just the right custom flexo. In practice I've only seen setups with those customized machines. The model 33 however would happily speak to anything that could talk ASCII, and ASCII was a standard with some space to it. Plus, the whole 8-bit thing

can't be ignored. 5 is not a power of 2. 8 is. Computers by this period now prefer powers of 2. It's just easier for software and digital circuits to work with. Although, due to electromechanical nature of teleprinters, that may not have been the biggest deal in the world, it's still worth mentioning. There is another complicating factor here. At least, I'd like to propose one. That's the rise of time sharing. As the 60s began, we reach a new era of computing. Time sharing operating systems, these systems that can split a single computer between multiple interactive users, come into common use.

For these systems to be viable, you need some kind of remote terminals, and ideally, you need a lot of them. The more terminals you have, the more useful a timesharing system is. Instead of having one computer hooked up to a single teleprinter as a novelty, timesharing shifts us to this world where the terminal is your main and usually only view into the computer. That represents many changes including cost. Suddenly you need 10, 20, 30 or more terminals. The unit price starts to matter a lot at that point. Teleprinters were perfect for this application. They were designed for

remote communications after all. If you had a teleprinter set up on a phone line, it could talk to a computer just as easily as another teleprinter. But for the trick to work, you needed commodity hardware. And you needed standardized hardware. The Model 33 was just better positioned for that. Case in point, the story of BASIC. That language's direct predecessor, DOPE, was developed on an LGP30, a flexo machine. BASIC was developed to run under DTSS, the Dartmouth timesharing system. Its interface was a pile of Model 33s, tastefully organized around the Dartmouth campus, of course. Then we get to the matter of Singer. The Flexo's strained corporate history

didn't stop with Frieden. In 1965, Frieden was acquired by Singer. And that's the sewing machine, Singer. The buyout itself was a little messy, to say the least. It started in the early 60s, possibly around 62, when Singer approached Frieden looking for a computer system. They wanted a multi-user system that could automate an entire store, something like a mini-computer with multiple point-of-sale terminals, almost time-sharing. Frieden took that contract in

part because they were already making similar machines. They had every piece of the puzzle they needed. It was during this contract that the buyout happened. This, according to some, spelled doom for the Flexo. Traditionally, there are two factors cited for this. The first is that Singer wasn't a computer company. They didn't know exactly what to do here. For the FlexoRyder to survive and thrive, it needed to make a shift towards the computer side of the market. Clerical work was well and good, but increasingly

that meant computer work. The teletype was very well positioned to take advantage of this. Cheap, standardized, commodity hardware. So was the flexo rider if it was given a bit of a push? If a commodity version of the flexo shipped, one that was designed around a standard like ASCII, then it could have become a real contender. Instead, Singer takes some weird steps. In 1965, they refresh the Flexo line. This leads to a machine that cost around $3000, but it did have a nice sleek coat of paint for the new decade. This was possibly the wrong direction to go. At least it meant the Flexo was missing out on the market that the Model 33 was flourishing in.

It just didn't make a lot of sense to get 20 or 30 $3,000 terminals when you could do the same for a third of the price. And I'll remind you, $3,000 in 1965 is about $30,000 in modern-day currency. That said, Singer was trying to become a computer company. Sewing machine sales had been declining for years. This is usually.2 given as why the Flexo died, because Singer itself was dying. The Freedon buyout was part of this larger strategy of survival. The same year that Singer bought Frieden, they also bought

general precision equipment. Frieden of course came with the portfolio of technology that we know. GPE came with their own, including a number of computers. And actually, in a funny twist, GPE owned Libroscope, which had manufactured the LGP-30. How's that for a bit of a messy corporate history? With these buyouts, Singer was trying to shift. They would launch the Singer System 10 computer, the result of that old Frieden contract in 1970. That's exciting and all, but there's a quick route to trouble here. The System 10 was marketed as a business computer, a general purpose business here. The System 10 was marketed as a business computer,

a general purpose business computer. The company had even changed their name to Singer Business Machines. This is happening in 1970. We're talking about an era where IBM is in their full awful power. When people talk about the dominance of IBM, they mean the 50s, 60s, and 70s. This is one of the eras when antitrust cases are brought against IBM because of their market

share. By some calculations, it was hovering around 70%. Men in blue suits were doing very technically legal things to secure and retain contracts. It's tempting to see that market and imagine dollar signs. You can just muscle your way in, but you'd have to do something drastic to break up that space. It was very difficult, if not legally speaking impossible, to compete with IBM in this period.

The exceptions were the so-called seven dwarves and companies that could slip in between lines in IBM in this period. The exceptions were the so-called Seven Dwarves, and companies that could slip in between lines in IBM's business plan. I'm talking things like mini computers, some special purpose applications, and micros. Singer didn't make much of a dent. This also meant that while Singer was dreaming of blue suits and selling computers, they weren't moving into the commodity terminal market. They weren't exploiting the flexo for all it could be. For Teletype, this was huge. An ancestral competitor was receding, and the Model 33 and later iterations could

dominate the market. I think that's why we call it a Teletype today and not a flexo. Alright, that's the story of the rise and fall of FlexoRider, and I will reiterate, I came into this whole thing pretty blind. I mean, I knew what the flexo was, but I had no idea where it came from. The fact that it all traces back to IBM, and to some strange corporate dealings, well, that came as a shock even to me. The show's unofficial motto is that IBM will always show up, but I don't know, maybe I'm stupid, I never actually expect that to be true until it happens. The Flexo writer marks a very specific period in computing. It's the gap between the

first digital computers and the rise of time sharing. It's when computers, even the mass-produced ones, are still pretty custom. The computers that were produced in bulk in this era sold in the hundreds, or low thousands in some special cases. The Flexo was only really used on machines that had a single real-time terminal. Cases where multiple flexos were installed were very rare. It wouldn't survive the jump to time sharing. The shift in use patterns from batch mode with an ancillary terminal to multiple real-time terminals, that perhaps proved to be too much for the venerable flexo. Maybe it was the change in the medium. Maybe it was the fact that the teletype corporation

could undercut the market, or maybe it was bad timing. Whatever the root cause, we can find a pile of options. That's usually indicative that a technology's time is up. The flexo, like so many other things from a bygone era simply faded away Thanks for listening to addin of computing and as a reminder, I'm skipping next episode I'm taking a little break to do a bunch of other podcast stuff So I'll see you in one month on the main feed

Until then if you like the show there are a few ways you can support it. You can rate and review on Apple Podcasts and Spotify. If you would, please take a minute to share the show with your friends. You can also buy merch or support the show directly through Patreon. Patrons get early access to episodes, polls for the direction of the show, and bonus episodes. For links to everything, go to my website, adminofcomputing.com, and until next time, have a great rest of your day.