Advent of Computing - Episode 24 - Making Disks Flexible, Part 1

Episode Date: February 24, 2020

The floppy disk was a ubiquitous technology for nearly 40 years. From mainframes to home computers, the plastic disk was everywhere. And in the decades it was around there were very few changes made to how it fundamentally worked. So how did it get so popular? What made the floppy disk so flexible? And how did it finally fall out of favor? In this episode we will look at the technology's early days. Like the show? Then why not head over and support me on Patreon. Perks include early access to future episodes, and stickers: https://www.patreon.com/adventofcomputing Important dates in this episode: 1971: 8 Inch Floppy Disk(Minnow) Created at IBM 1976: Shugart Invents 5 1/4 Inch Floppy Disk

Most of you probably have some exposure to the floppy disk. If you weren't old enough to use one personally, then you've at least seen it as the save icon in most modern programs. They've now almost completely gone out of use, except for some legacy users. But there was a stretch of about 40 or so years when they were synonymous with computer storage, and really synonymous with computers in general. But how exactly did this happen? When you stop and think about it, the floppy disk is a really strange and very limited storage medium. The disks themselves are pretty fragile, and they don't hold all that much data. The drives aren't that much better, a floppy drive can easily get gunked up and stop working

completely. But despite their problems, the humble floppy disk really became a mainstay. At least for a time. Over its lifetime, the floppy disk would find its way into generations of computers. But by the end of the 20th century, it would start to disappear from the scene. Today, you'd be hard-pressed to find one even if you looked. And that's part of what makes the story of the floppy disk so interesting. Believe it or not, the floppy predates personal computers. They were originally developed specifically for mainframes. Then, with very few changes, the floppy made the jump to personal computers while the market was still in its infancy. But why did the floppy made the jump to personal computers while the market was still in its infancy. But why did the floppy disk, this strange and limited format, become the centerpiece of computing for so long?

Was it a case of just being in the right place at the right time? Or did the floppy disk enable the personal computer revolution itself. Welcome back to Advent of Computing. I'm your host, Sean Haas, and this is episode 24, Making Discs Flexible, part one. Today, we're going to be looking at one of yesteryear's most well-known devices. That's the floppy disk. It's probably best known as a portable storage medium used with almost every personal computer. In the 80s and the 90s, the floppy disk was as common as the keyboard, and just about as necessary. But the history of this device goes back further than personal computing. As it turns out, it would start all the way back at IBM in the late 1960s.

Before we get into the history, I think it's useful to address what a floppy disk actually is and roughly how they work. At its core, a floppy is just a magnetic storage medium. The disk part of the floppy disk is a thin, flexible mylar disk, and it's embedded with particles that react to magnetic fields, so you can lock in a certain polarity. Think of it as something very similar to a cassette tape, but flattened out into a disk. By using a device very similar, as it turns out, to a tape read right head, it's possible to store binary data on the disk and then read it back. The magnetic disk is then encased in a plastic shell to keep it safe. There are, of course, various other features, changes over the years, and quirks, but we'll get into that. For right now, just keep this in mind as the broad

strokes of what makes a floppy a floppy. What makes things more interesting is that the basic design of the floppy disk actually changes very little over the decades that it was in use. It would start out in a very niche application at IBM, but once it left Big Blue, the technology would be stretched and stretched to fit more general use cases. From living in mainframes to helping add fuel to the early days of the personal computer revolution, floppy disks seemed to always find a place to fit in. Or at least they would for quite a while. In that sense, I think it falls into a pretty interesting category. Some devices still in use today, like the computer mouse or the keyboard,

have changed very little since their creation. Floppies had a similar trajectory. Once invented, their evolution was slow and their adoption was very widespread. But ultimately, the technology fell out of favor as better options appeared. So today, we are going to start part one of a two-part exploration of the floppy disk. Starting with its early roots, rise to favor, and then eventually leading up to its zenith and fall. To talk about the creation of the floppy disk, we need to first talk about the machine that it was designed for use with. That's the IBM System 370.

First sold in 1970, the 370 was the follow-up to IBM's earlier System 360, which was one of the company's most popular line of mainframes. Actual development on what would become the 370 was a pretty long process. It had to be more powerful than its predecessor while maintaining backwards compatibility and making some incremental improvements. In that sense, it was a true successor. And of all the tweaks and upgrades being developed, there was one in particular that interests us today. That update has to do with how the System 370 dealt with a little something called the initial computer program load. One of the things that makes talking about mainframes, especially older mainframes, difficult is the language barrier involved. For the most part,

you can think of a mainframe as just a really big computer. It has everything a personal computer has, just bigger and by a different name. Disk drives are sometimes called disk files. Memory is called storage. I.O. devices are called channels. And beyond just that initial level of strange language barrier around Big Iron, just that initial level of strange language barrier around Big Iron, all of these names usually change by manufacturer. So once you get used to reading about IBM systems, if you look at a DeX system, forget about it. Probably won't even know what they're talking about.

Initial Computer Program Load, or just ICPL, is another one of those strange mismatched terms. On any other hardware besides IBM mainframes of this era, it's known as microcode. And at least in my head, it's one of the more complicated features that a computer can have. To try to simplify it down, microcode, or ICPL, or whatever you want to call it,

is a program loaded up by the computer as soon as it turns on, and it describes how the processor should work. For systems that use microcode, the actual initial program can set up everything from how registers and addresses are structured to even defining the instruction set for the processor. Primarily, microcode helps with overall flexibility of a processor. This is relevant to us today because one of the System 360's big flagship features and

big areas for improvement was its implementation of ICPL. Microcode was one of the factors that made the System 360 so successful, at least behind the scenes. One of the big deals about the System 360 mainframes was that they formed a family of computers. Each model of 360, of which there were actually quite a few, was compatible with every other System 360. Some had more memory, others could support different peripherals, some were faster, but by and large, they were all compatible. For the time, this wasn't the default. Most computers were wildly different. Even systems from the same manufacturer were often

totally incompatible, even down to things like the character sets. The System360 was the first to really build a family of computers like this, and microcode made it a lot easier for the engineers at IBM to do that. The underlying hardware could differ,, and microcode made it a lot easier for the engineers at IBM to do that. The underlying hardware could differ, but the microcode program could implement the needed layer of compatibility. But the issue here, or rather how the system 370 had a chance to set itself apart, was that you couldn't really upgrade the microcode on a 360. As a quick note here, in this section, I'm going to be drawing heavily on Emerson Pugh's IBM's 360 and early 370 systems. Pugh worked at IBM from 57 to 93. And while he wasn't directly involved with the creation of the floppy disk, he was very close to it. And luckily

for us, he's written extensively on the topic of IBM's history. It's not technically a primary source, but it's really darn close, and just the amount of compiled research in his writings is fantastic for my purposes. So, moving on. Upgradable microcode was one of the planned features for the 370. Of course, an upgrade path wouldn't be that simple. It rarely is. There were some design constraints right off the bat. The plan was to have some kind of storage medium that the ICPL could be stored on.

Then it would be read into the 370 as the computer booted up. Once on the computer, the code would reside in a fast cache of memory. But the matter of what it would actually be loaded from had its own considerations. From Pew, Objectives suggested for this ICPL device were a manufacturing cost under $200 and a storage medium costing under $5 that could hold up to 256,000 to 1,240,000 bits, be replaceable, and easy to ship. Price wasn't a median factor. It had to be. The new machine couldn't have such a core feature break the bank.

That could make it a harder sell to clients on down the line. But whatever storage medium ended up being used, it had to be portable. You had to be able to get upgraded microcode out to customers in some kind of sane and easy way. Otherwise, it kind of defeats the whole purpose of the project. That was the state of things in 1967 when an engineer by the name of Dave Noble was assigned the project. And despite the importance of finding a suitable ICPL solution, the project would start as a solo endeavor. By the time this new project came along, Noble already had a pretty long career. He had taught engineering at Rensselaer Polytechnic, worked for Remington Rand,

and then found his way into IBM where he helped develop early hard drives. He was well-versed in the possible options for an ICPL device. So, what did those options look like? Well, as it turns out, there were a whole lot of candidates. It was really an open field. Hard drives did exist, and believe it or not, there were actually portable hard drives all the way back in this era. At least, roughly speaking.

These devices, called disk packs, were large plastic cartridges that held a set of metal hard disk platters. A pack was used by inserting it into a corresponding disk pack drive, which was connected up to a computer and had all the actual electronics and read-write heads needed to access data on that disk pack. This was a proven technology, but there would be some logistical issues. A disk pack cost way more than $5 to produce. That and each pack was relatively heavy and fragile. Hard disk platters can easily shatter if jolted or hit, so mailing one would be a pretty risky procedure.

Another option in play was a magnetic tape. Half-inch open-reel tape was a pretty standard storage medium for computers for most of their early history. It wasn't too delicate to ship, and tape drives were pretty ubiquitous at this point. But like disk packs, tape was bizarrely expensive, prohibitively so. The other issue with both diskpacks and tape was a matter of convenience. ICPL would need to be loaded at every boot, and not all mainframe installations would have the luxury of an extra tape drive or hard drive. Having to swap out a tape or a disk pack every time you turn on a computer, that would add a lot of unneeded complication. With existing options all having

issue, Noble would spend most of the early days of this project looking at novel solutions. I've seen reports that he looked at such far-flung options as 8-track tape cartridges and even 45 RPM records. This may seem ridiculous on the surface, but either of these could have been a good choice. It's possible and actually pretty easy given a little tinkering to store binary data in the form of sound. This is how modems work to transfer packets over phone lines. Devices that can modulate binary data into analog sound and then demodulate that sound back into binary data, well, they're pretty cheap and easy to come by. You can do it all with

just an audio jack and some programming know-how if you have to. But ultimately, no existing device would totally fit the criteria. Noble would end up creating something totally new. Probably partly inspired from his earlier work with hard disks, Noble started looking into producing a magnetic disk, but this one would be made primarily out of a plastic film clad in magnetic oxide. The material was pretty similar to that being used by open reel to reel tapes, so it could be read by and modified by a tape head. The very early designs are already pretty close to the floppy discs that we know today, but there were some intervening steps.

As 1967 drug on, the project, now known as Minnow, was looking promising. Noble was even able to expand his team of one as more technical problems were being tackled. Warren Dalziel is one of those engineers. He described the early prototypes like this. I remember when it was first shown to us. The approach they were taking then was, I think it was a mil and a half of mylar bonded to about an eighth inch of foam rubber, and there was no jacket or anything, and they took a head we bought from a tape drive and loaded that into the foam, and the actuator mechanism was a lead screw, but driving the lead

screw was a fairly Rube Goldberg mechanism. End quote. So right off the bat, there are some pretty strange choices going on, so let me explain. The magnetic medium was a pretty good choice for storage overall. It was well tested in tape drives used at the time, so making it into a disc was a reasonable step. But, well, the disc on its own was just too floppy. And they were big to start with. At 8 inches in diameter, it's easy to see why they would want to, well, flop around. The thing wouldn't stay still enough to read, so it was glued onto a rubber disk so it wouldn't

warp or flex. And that was it for the disk. The reading mechanism, though, was a little more Rube Goldberg. When you get down to it, a floppy drive works in a pretty clever way. Data is read by a head. In this case, it's a tape recorder head, but later models would have specialized floppy disk heads. Data on a floppy disk is laid out over the entire disc's surface, and it's broken up into sectors and tracks. I like to think of it kind of like a table. Tracks are columns and sectors are rows, you just have to think of it as being stretched into a disk instead of left as a rectangle. Anyway, the head has to be able to get to every possible location on the disk.

The trick that Noble came up with makes this simple, or at least it makes the physical device pretty simple. The head was mounted on a screw gear connected to a servo motor. As the motor turned, the head could move across the disc, from the outside to the inside of the disc. That was controlled by the computer, and it allowed for track selection. The disc, on the other hand, moved at a constant rotational velocity. That wasn't controllable, well, except for on or off. To select a particular sector, you just had to wait for it to spin around under the head. Of course, that meant tracking which sector you were on, which is another complication. These early discs were sometimes called hard sector discs because the start of each sector

was marked on the perimeter of the disk with a small punched hole. A sensor in the drive would then detect when a hole went past, so it was possible to keep track of which sector was currently under the drive head. Complicated enough? Well, yeah, I think Dalziel was right about the mechanism. So why would Noble go to the trouble of such a convoluted device? Well, let me count up the components. You have a servo motor, a normal motor, a belt to spin everything, a magnetic tape head,

something for the sector detector, and then just some assorted plastic and metal for the housing. Outside of whatever electronics this thing is actually hooked up to for controlling it, the Minnow floppy drive is relatively cheap and simple to produce. Compared to hard drives of the time that had state-of-the-art everything, even down to voice coil controlled motors, the Minnow really fits its diminutive sounding name. Up to this point, the project had been coming along pretty well. But it was starting to become obvious that an exposed foam and plastic disc wasn't going to last very long. The problem was that it was easy to damage these early prototypes. Dust could fall on it and ruin it. Or you could just accidentally smudge it when swapping out discs. That made it became

very apparent that it wouldn't be very safe to ship a bare magnetic disk. The last change that made the floppy disk recognizable was to ditch the foam backing and put the disk into a plastic sleeve. This served a dual purpose. It protected the disk while also adding rigidity. The sleeve has two holes, one in the center that exposes the very inner edge of the disc. This allows a clamp to grab and then spin the inner disc. The other opening is an oblong window that lets the drive head come into contact with the disc itself. The finishing touch was to add a small layer of fabric between the disc and the sleeve so that as the disc spun, accumulated dust would be wiped off.

By 1971, the Minnow disc, better known as the first 8-inch floppy disk, was completed and started to ship alongside the brand new IBM System 370. The initial version was read-only and could only store data on one side. In total, you could fit just under 82 kilobytes worth of information on the thing. But that was well within the bounds of the ICPL project's goals. Initial disks were read-only, and a special and more accurate drive was needed to write data to them, but that wouldn't last for long. While Minna was prepping to ship, a new team within IBM was already working on a new floppy disk and matching drive. Codenamed

IGAR, this new disk would be readable and writable in the field. The new IGAR disks and drives would ship a few years after Minnow. But in the intervening years, something interesting would happen. IBM had picked up on the general usability of the floppy disk, so it wouldn't be reserved for the ICPL project for long. iGuard was part of that plan. Making a more general-purpose drive would go a long way. But IBM wasn't able to keep a tight grip on this new technology for very long. In the modern day, IBM's probably best known for their PC. It was a huge smash hit for the company, and we're still reeling from its effects. Part of the reason the PC was such a big deal was that it was quickly copied by other manufacturers.

That led to the market flooding with cheap clones. That wouldn't happen until the 1980s, but it turns out that IBM had been dealing with clones of their hardware way back before the PC was ever a dream. And one of their adversaries in this arena was a company called Memorex. Founded in 1961, Memorex started out manufacturing magnetic tapes for mainframes. That alone would serve them well. It gave them a pretty good slice of the computer market. It would only go so far, of course, but it gave Memorex enough income to start work on something bigger. Over the course of the next few years after the company's founding, they would release their own disk packs and other sundry storage medium. But that was all a lead up to their big ticket item.

In 1968, Memorex announced the 630 hard disk drive. It was a plug-compatible clone of one of IBM's own disk drives. It was cheaper than IBM's offering, but it also had more high-tech pieces of equipment inside of it. The 630 used more advanced tracking motors and drivers for controlling the hard drive heads, so it could read and write data faster and more reliably. And on top of everything, you could plug it right into your existing IBM mainframe. At the time, no other company was shipping third-party disk drives like this. You had to buy IBM's hardware, so Memorex was able to make a big splash in the market. But there was another effect, possibly a more long-term effect. Memorex started to poach IBM employees. Over the course of about 1969-71,

Memorex was able to pull over around a dozen ex-IBMers. Most of the new recruits came directly off the Minnow and Igar projects. There were a host of reasons as to why Memorex was able to pull this off, but one that I've seen repeated a lot is that these defectors felt stagnant at IBM. At least in the field of storage, Memorex was starting to push the envelope a little more than Big Blue. There was more of a chance to work with totally new technology, and that was something enticing to engineers. Of all the defectors picked up, the most consequential had to be Alan Shugart. This was immediately important because Shugart was Noble's old boss. In fact, it was Shugart that assigned Dave Noble to the ICPL project that would lead to the first floppy disk. So in a big way, Memorex was buying into the floppy disk.

While IBM was still working on Igar, Memorex shipped their first read-write floppy drives. Designed by a team working directly under Shugart, the new drives, called the 650 Flexible Disk File, were able to store and read data from 175 kilobyte disks. In every way, it was an upgrade from IBM's Minnow drives. But besides just a technical improvement, these new Memorex drives opened up the floppy disk to a whole new market. And while ultimately Memorex would only be a small part of our overall story, they served as a sort of bridge to bring the technology outside of IBM.

Part of the shackles that IBM had was the fact that they were, and they still are, so large. They made mainframes, and they made peripherals for those mainframes, and they made storage medium for those mainframes. Memrex was much smaller, so they didn't have this problem. They didn't have to worry about selling their products alongside, well, anything. Memrex could create whole products for consumers, or they could just sell components to other manufacturers. In the case of their first floppy drive, they went with the latter. The 650 flexible disk file was sold as a bare component, a path that IBM wasn't able to go down. There were some big impacts from having a non-IBM disk drive

hit the scene. It made it possible for mainframe manufacturers, besides IBM, to offer floppy drives on their systems. Just like IBM was finding out, the floppy drive was a really useful device. It filled a very particular niche. The flexible plastic disk landed somewhere between tape drives and hard drives. Magnetic tape could hold a lot of data, and before the floppy, it was relatively cheap as a medium. However, since it's all read on a reel, it had to be accessed sequentially. On the high end, you have hard drives, which were very, very expensive. But in exchange, you got a pretty fast read-write time, since data could be accessed randomly.

The floppy disk really mixed the best of both devices. You get a small and portable disk with a fast read-write time, and it doesn't cost all that much. In 1971, it was an exciting new technology, and in 72, with Memorex entering the arena, it became an accessible new technology. Memorex was the first to make floppy disks outside of IBM, but they weren't the only to adapt the disk. As with any time a new technology proliferates, the details get a little confusing. There's a lot of players in the field, and a lot of them don't last very long. To just start looking at all the new companies making floppy disks in this early period and all the small changes between their products, well, that wouldn't be all that useful. But there are a few manufacturers that should be

mentioned. Memorex is one. But another early player was Sugart Associates. That's right, the Sugart that oversaw the IBM development of Minnow and Memorex's own floppy drive. In 1973, he'd defect again, this time leaving Memorex to start his own company. But here's the thing. Sugar Associates wasn't founded to make floppy drives. The original business plan was to make an integrated office data entry system, essentially a large-scale input-output setup for what would connect to a central mainframe inside an office. That encompassed data entry terminals, printers, tape drives, and yes, even floppy drives.

So armed with a handful of employees peeled off from Memorex, Sugar Art started gathering investors and building up his fledgling enterprise. However, the imagined office system would never come to be. Some of the first products that SugarArt would produce were, of course, 8-inch floppy disks and their accompanying drives. Like I touched on before, the new technology was in high demand at this point, and SugarArt had a lot of experience in the industry. And, as it turns out, the timing here couldn't have been better. In the early 70s, mainframes and minicomputers were still the only viable options, but home computing was just around the corner. Some very early proto-personal systems

were already being developed, and as it happened, the floppy disk would become a desirable choice for these early systems. The SA800, Sugarheart's own 8-inch drive offering, became so popular and it was cloned so much by other manufacturers that their internal drive connector started to become a de facto standard for other drives. Now, when I say personal computers here, I'm not talking about something that we would recognize. In a lot of ways, the very first home computers were attempts to miniaturize mainframes. Something like the Altair 8800 is really in this vein. It has a front panel with switches, like a mainframe, you need to connect a terminal for text input and output, just like a mainframe, and it has storage options very similar to

a mainframe, just downsized. One popular choice was to use audio cassette tapes as a makeshift type of storage, essentially just adapting the tape drive mechanisms used on mainframes for much smaller scale applications. Some systems could even connect up to a standard tape recorder to read and write data. But despite its popularity and the fact that tape drives were just a dirt cheap option, cassette storage was flaky at best. Mainframe magnetic tapes were highly specialized. Tape was formulated specifically for data storage. Lifespan of tapes was well known and accounted for, and the read-write mechanisms were complex and highly engineered. You can't really get any of that with an audio cassette.

Hard drives were another story on the home computer market. That story's too expensive, too large, and too hard to get a hold of. This is where the floppy drive comes in. While more expensive than the cassette option, they were actually obtainable. Some early systems ended up offering 8-inch drives as either a core component or an expansion option. That included the Altair 8800. And most of these disk drives would come straight out of Sugar Art Associates. But this very early period wouldn't last long.

In the grand scheme of things, not that many personal computers would offer 8-inch floppy drives. The change from 8-inch down to the smaller 5.25-inch floppy would be a quick one, and that shift would start during a meeting in 1976. At the time, SugarArt's biggest customer was Wang Laboratories. Wang produced word processors, essentially single-purpose computers designed for text editing. In 1976, the head of the lab would come to their supplier with a pretty big request. Years later, Don Massero, an engineer at Sugar Art Associates, recounted it like this. I need to come out, and this was just as people started to talk about PCs. They didn't quite know what it was.

I want to come out with a much lower-end word processor. It has to be much lower cost, and I can't afford to pay you $200 for an 8-inch floppy. I need a $100 floppy." Wang had been trying to find a way to make a more personal computer. That was an equation that a lot of people were trying to solve at once, and there were a lot of different approaches. As far as Wang saw it, the new system would need to be smaller, it would need to be cheaper, and it would need to be reliable. To that end, a floppy drive would be a must, but current drives were too large and too expensive to fit into Wang Labs' master plan. And there was enough money behind getting the new design out that Sjogart Associates decided

to take up the task themselves. Over the course of 1976, with a whole lot of back and forth between Wang and Sugar, the new so-called mini floppy drive was designed. What's interesting is that no remarkably new technology was actually needed. The final drive was just a scaled down, somewhat modified version of the earlier 8-inch drive. But the small changes were all for very good reasons. One that caught me off guard when researching was the power requirements for this new drive. As it turns out, in older 8-inch drives, it was pretty common to use alternating current motors to spin the floppy disk. I couldn't find a definitive reason why, so I guess it was just

a matter of convenience at the time they were designed. It's especially odd because almost all computer components use direct current. Anyway, an AC motor on a mainframe wouldn't have been a big problem, and there wasn't any concern about circuit complexity or power usage in the original IBM Project specs. But once you get down to smaller machines, you start running into some strange issues. Here's the thing. If you want to have a computer that can fit on a desk, But once you get down to smaller machines, you start running into some strange issues. Here's the thing. If you want to have a computer that can fit on a desk, that means you have to have a lot of different components packed into a relatively small, dense space. One of those had to be a CRT monitor.

And it turns out that old-style CRTs don't play well with AC motors. Specifically, the alternating magnetic field in an AC motor can cause interference on CRT tubes. That meant that any small system like Wang was designing had to do away with the AC spindle motor of older floppy drives. Besides solving the interference problem, the change to a DC motor also simplified the power requirement for a floppy drive. A computer already had DC voltage from its power supply, so you could just tap into that to power the drive. The other big change was, well, the size of the disk. When all was said and done, the new mini floppy was cut down to just

5.25 inches across. Seems like a really specific size, so there has to be a very specific reason for it, right? Well, it may come as a surprise, but the answer to that question is pretty hotly contested. There isn't a consensus, even among the engineers that worked on the project themselves. One commonly told, and most likely incorrect story, is that a few Sugart engineers went out for drinks with the head of Wang Labs after a meeting. Sjogart engineers went out for drinks with the head of Wang Labs after a meeting. While at a bar talking tech, Dr. Wang held up a cocktail napkin and said he wanted the new floppy to conform to those dimensions. When measured, the napkin turned out to be 5.25 inches wide. Another story is that two engineers were driving from Sjogart's office to Wang for a meeting. The drive was pretty far, and it was snowing, so traffic was

delayed. So, on the way, the two commuters started to talk about the upcoming floppy drive project. Over the course of the ride, all the details for the new drive were hashed out, including the size of the disc. Using a piece of cardboard as a mock-up, the two determined that a 5.25-inch disc could fit in a shirt pocket without being bent. The logic behind that was that it would make the new disc convenient to use and carry. However, this theory is also pretty dubious. The other common theory, and the one that's most likely believed, comes down to logistics. In this telling, the drive was designed before the disc, since it would be easy to just cut a sheet of plastic down to any necessary size. In order to make the new drive as easy as possible for manufacturers to use, it was decided to make it fit into the footprint of popular cassette tape drives already in use with computers.

That way, a computer company could just buy the new floppy drive and slot it right in where the old tape drives fit. As it turned out, the largest disc they could fit into that drive was just about 5 and a quarter inches across. By late 1976, Wang would get their hands on the new $100 floppy drive. But the new drive, called the SugarArt SA400 Mini Floppy, would be more than just a Wang exclusive. The original drives could only handle around 86 kilobytes per disk, but that didn't stop the mini floppy from becoming a smash hit. A consumer couldn't really go out and buy the new floppy drive, or at least that's not how most people ran into the sugar drive. With the personal computer explosion just starting, the SA400 and its later incarnations would find themselves

packed away in almost every computer on the market. And it's plain to see why. From inception, it was designed to be used with small-scale systems. The drives were cheap, they were fast, and they stored a reasonable amount of data. Besides being great for computer manufacturers, the introduction of a new, cheap floppy format was also a boon for consumers. In 76, a case of 10 5.25-inch discs cost $45. That's about $4.50 a disc. Once other manufacturers started turning out discs, that price fell as low as $1.50 each.

While maybe not as cheap as a pack of blank cassette tapes, it's not out of reach. And compared to a cassette, a floppy disk was much easier to use. All you had to do is flip open the drive door, put in the new disk, and close the hatch, and you're off. In a lot of ways, it was a really good medium for the new accessible computers that were just hitting the market. The other advantage of the Sugart drive is something that came more by circumstance than design. That's the introduction of a standard floppy drive connector. I briefly touched on this earlier, but I think it's an important factor to consider. Standards are a big deal. It makes designing and building computers cheaper and easier. And in the early days of the 8-inch floppy disk, there wasn't really a standard

to go on for how those drives should be wired up to a computer. IBM had one way of doing things. Memrex followed their lead for the cloned drives, and then there was an assortment of other methods used by other manufacturers. Eventually, Sugaert's SA800 drive became popular, so other manufacturers chose to copy how Sugaert's floppy drive connected. But there was still an amount of variability. That would change with the rise of the 5.25-inch disk. The SA400 hit the scene with a splash. The new floppy drive was a hot commodity, and Sugaert was already established as a leading force in the market. So when other companies started to make their own smaller disk drives,

there was really only one way to go. The 34-pin connector that Sugart used showed up on almost every new mini floppy drive. This level of ad hoc standardization had a big impact for personal computers. PC manufacturers just had to design their floppy bus to have a 34-pin spec. Then, any drive could just be connected up. It doesn't matter if it's from Sugart, Memorex, or some other company. Standards like the 34-pin connector may seem like a small, nitpicky detail, but these small changes gave computer manufacturers a lot more flexibility to make cheaper systems. Standardization is one of the keys here, but more than that, the new mini floppy had just the right combination

of features to fit the new personal computer paradigm. Smaller machines, cheaper machines, and more accessible machines were becoming possible. Beyond just the creation of totally new hardware, the late 70s and into the 80s was when we saw a dramatic shift in how and where computers were used. The fact of the matter was that personal computers had different use cases than mainframes. To me, that's part of what makes the floppy disk so interesting. It was designed for large computers, but with a few minuscule tweaks, it became perfect for the home computers. There are a lot of machines that use the 5.25-inch disk, so I won't ramble on about each of them.

But I think it bears looking at one in particular in order to give some more context around the new floppy drive. And once again, this brings us back to IBM and the IBM personal computer. I think this machine exemplifies the personal computer revolution for a few reasons. Well, outside of the name for that coming from this system. One of the big ones is that the PC was a nearly open architecture. From the beginning of the project, it was decided to use all off-the-shelf parts to design and build the machine. Besides a small chunk of boot-up code, nothing was custom on the device. And later down the line, that would

open up the market to clone manufacturers, increased competition, and a spread and very rapid rise of the machine. In 1981, when the PC was released, the mini floppy had only been on the market for about 5 years. That's enough time for floppies to have become a standard. So it was a no-brainer for IBM to use the floppy drive as the primary storage medium for the computer. A stock PC would come with two 5.25 drives front and center. The fact that SugarArt's original mini floppy had been copied and used as a standard meant that IBM and PC clone manufacturers were able to use increasingly cheaper and more available drives to power computers. It fit into the larger strategy of off-the-shelf construction the PC became

known for. And it became possible to open up the PC market thanks to that. Outside of the drive towards standardization, there's one more side piece that I think made the Mini Floppy fit in so well in the era. Remember that one of Wang Lab's requirements for the Mini Floppy was a DC motor. That simple change, from AC to DC, made a big difference moving forward. The idea of a computer that can fit on a desk was a driving factor in a lot of these early system designs, and I think it's something that kind of fades into the background after a while. If you have a vision of a computer from the 80s or 90s, or even later, it probably includes a bulky monitor placed atop the actual box containing the computer's circuits and drives. Or maybe it's a monitor next to a

tower case. Now, if disk drives were very different, if they didn't miniaturize in the way they did, and if some smart choices weren't made along the line, that form factor wouldn't have been possible. The floppy drive being exactly how it was when it was quite literally set the face of computing. Alright, that does it for part one of our exploration of the floppy disk. From a very niche use deep inside IBM mainframes to more general use as a storage medium and eventually making the jump to early home computers, the floppy disk was able to adapt to play a series of very important roles. And with the rise of the five and a quarter inch

mini floppy, the personal computer market got a very powerful tool, and just at the perfect time. Would we have PCs if the floppy disk had stayed in the realm of the mainframe? Probably. Something would have come along to fill the gap between tape and hard disks, but, at least in form, and probably in function, computers would have been quite a bit different. And the story won't end here. I think it's a logical place to take a breath because when we come back for part two, we'll be looking at a totally new cast of players. Sugar Associates wouldn't be the one to crack the code on the 3.5-inch disk. The mini-floppy, if we can still call the 5.25-inch iteration that,

was a golden age of standardization. Floppy disks and drives really never changed that much in the 40-some years the technology was in use, but especially so within the era of the 5.25-inch disk. Early standardization, thanks largely to SugarArt's dominance in the market, ensured that. But with the push for further miniaturization, a wild west of strange floppy disks starts to open up. However, that would all change thanks to two key players. So next time, we're going to be looking at Sony and Apple and the creation of the final floppy drive, the three and a quarter inch disk. Thanks so much for listening to Advent of Computing. I'll be back in two weeks time with the conclusion of

the story of the floppy disk. And hey, if you like the show, there are now a few ways that you can help support it. If you know someone else who would like the show, then why not take a minute to share it with them? You can rate and review me on Apple Podcasts. And if you want to be a super fan, you can also support the show through Advent of Computing merch or signing up as a patron on Patreon. Patrons get access to early episodes, polls for the direction of the show, and other assorted perks. You can find links to everything on my website, adventofcomputing.com. If you have any comments or suggestions for a future episode, go ahead and shoot me a tweet. I'm at Advent of Comp on Twitter. And as always, have a great rest of your day.