The bigger 8" floppies were remarkably robust. The boot floppy for an 11/780 fitted into an LSI-11 mounted on the front door, with a vertical disk drive slot so you opened up the cab, ignored all the CPU cards, put it into a slot, closed it up and booted.
The operator I knew frequently had this folded up in his pocket: You just had to smooth it out a bit, make sure it rotated inside its kevlar sleeve, and off you went.
(DECTAPE was famously robust, everything was written out about 5 times in the tape, it was designed for light industrial deployments. It was wide like old school 16 track audio tape, not like classic 1/2" 800/1200/9600 bpi computer tapes, before DAT/DLT cartridges and the like. But that said, the old tapes were pretty robust too)
I remember reading that there was a lot of data storage "left on the table" with floppy drive tech before it was finally abandoned. They apparently stored bits uniformly across the disk instead of more densely at the edges where distance between the points was greater. Considering how there was a real missing niche for a denser floppy media for a number of years (that allowed things like zip drives to become somewhat common) I found this idea pretty surprising.
Although single-speed motors were most common, some drives had variable speed motors. In combination with an appropriate disc controller, the rotation could be sped up or slowed down so that bit locations arrived at the read/write head at a uniform rate. Thus you could pack bits as closely on the outer tracks as on the innermost.
Similar story with C64 5.25” floppies. At the computer shop I worked at in 1989 we had a Microprose Soccer floopy disc without its shell, just the magnetic core. We used that to load the game and it always worked flawlessly. Sometimes we missed the floopy drive’s door, and the disc would crumple, but still work.
One piece missing from the description is the "write protect" notch in the 8" & 5 1/4" floppies. if the notch was present, the biscuit was read/write on that side. If it was not present, or was covered with a piece of tape, that side became read-only. If looking the diskette from the top, the notch would be in the upper right edge.
Early drives only wrote on one side of the diskettes. Someone discovered that diskettes had magnetic material on both side, so the diskette could be flipped and written on, doubling the storage. A notch would have to be cut to make it read/write, but it worked relatively reliably.
Later drives got second read/write heads and the flipping went away.
On the 3 1/2" diskettes this became a sliding tab. But, in their wisdom they flipped the concept. They reversed it. "missing" or open notch meant write-protect, and closed notch meant read/write.
The magnetic material on both sides had a very practical reason to be workable in all kind of floppy drives.
For example the write/read-head in a C64 floppy was placed on the bottom, so the down side of the floppy got magnetized. But the head of the floppy drive from a Atari 800 was placed at the top, so the upper side got magnetized.
So basically, when a floppy was sold as 1S (single sided) this was pure marketing. :) At least for the most floppy drives.
There's also a second thing, the "index hole". On real 2S floppy there are 2, one for each side. But most of the drives didn't care about this index hole. So, this wasn't relevant for C64 users and such.
Interesting article. I immediately thought of one (very obscure) thing I know about floppy disks which the author appears not to – "deleted sectors".
Before the data of each floppy sector, is a sector header. Among other information, it contains a flag byte, which can have two different values, indicating two different types of sectors – normal and deleted. There are different commands to read/write normal sectors vs deleted sectors. Historically, this came from the IBM 3740 data entry system, where each (128 byte) floppy disk sector stored one database record, there was not much in the way of a filesystem, and you would delete database records by changing a normal sector into a deleted one. Subsequent systems (such as IBM PC floppy controllers) retained this feature despite it being useless for their purposes. Almost no microcomputer/PC software ever used it, the principal exception being that it saw repeated (ab)use to implement an (easily circumventable) form of copy protection. Apparently, some late model PC floppy controllers (1990s / early 2000s, just before floppy drives disappeared entirely) dropped support for deleted sectors completely, viewing it as a disused legacy feature which was a waste of circuitry to support.
Well, that's true for any floppy based on the traditional IBM minicomputer/mainframe floppy format, which includes IBM PCs and many CP/M systems. It isn't true for floppy disk formats which lack that IBM heritage, for example Apple II floppies.
There were no sectors, just data fields. Also each data field had a text key field. I think IBM only gave this up (really virtualized it) when commodity hardware became better than their own.
So I'm wondering if CKD was ever a thing on floppies?
That's not really true – an (E)CKD record is fundamentally the same thing as a sector. The difference is that on standard hard disks (what in IBM nomenclature is called FBA DASD), the norm is every sector on the disk has the same size; CKD permits each track to have a different sector size, and even different sized sectors on the same track. Also, CKD optionally has separate key and data fields, with a physical gap between them; in floppies and standard hard disks, there is only count and data fields (count = sector header), and any key-data separation has to be marked logically (by a byte offset) not physically.
> So I'm wondering if CKD was ever a thing on floppies?
No, CKD was never used for floppies. In the standard IBM floppy format, the sector size is in the header of each sector, so in principle you could have a mix of different sized sectors on the same track, like CKD permits; in practice, I've never heard of anyone doing that, I question whether the floppy disk controller could cope with it; even if the FDC somehow could, almost all software would be confused by it. (I suppose if the FDC can handle it, someone may have adopted it as part of a floppy disk copy protection scheme.)
Something which did actually happen, albeit rarely, was formatting the first track on the disk as 128 byte sectors, and the remaining tracks with bigger sectors (such as 256 or 512 byte), to retain some limited backward compatibility with older systems that only supported 128 byte sectors; at least being able to read the first track might make clear it was a valid disk that just couldn't be read on this system, as opposed to thinking it was an invalid disk, and then formatting it in response, causing the loss of that valid data. Systems that did this sometimes also physically recorded that first track as FM, while using MFM for the remaining tracks. This was largely a 1970s concern; by the time the 1980s came around, systems which only supported 128 byte sectors had become practically irrelevant.
The classic IBM 3740 floppy filesystem supports record-oriented files. Some vendors other than IBM adopted it (e.g. Olivetti), but most non-IBM vendors ignored it, and developed their own filesystems instead (e.g. the CP/M filesystem, Microsoft's FAT). ECMA standardised an ASCII version of it as ECMA-91 in 1984 [0]. That standard didn't support keyed/indexed files, but could easily have been extended to support it–if anybody had cared to do so, as I said, relatively few vendors ever paid attention to it. The 3740 did support a key field in each database record (sector), but the key-data separation was only marked logically (by a byte offset into the data), not physically by a recording gap.
>in principle you could have a mix of different sized sectors on the same track, like CKD permits; in practice, I've never heard of anyone doing that, I question whether the floppy disk controller could cope with it
That was actually done by IBM to distribute some versions of OS/2 2.x (or was it a later one?). It was called XDF and was originally invented by a company making backup-to-floppy software. And of course it needed a special driver to access: IIRC the one built into OS/2 was also read-only, and the first few installation floppies were in standard format so that you could boot from them.
The PC floppy controller itself has no problem reading or writing different-sized sectors, provided you didn't try to do multiple ones in a single command.
But creating the low-level format with PC hardware required a crazy hack: fill the track with 128-byte sectors without any gap between, but setting up the "C,H,R,N" header fields for larger sectors with the correct spacing. The remaining dummy sectors would then become part of the larger sectors' data and gap field, so their headers didn't matter (and would be overwritten when writing the actual data).
Although there is still a big difference between floppies and CKD hard disks – on IBM standard floppies, the sector size has to be a power of 2, minimum 128; on CKD hard disks, the sector size can be anything – 80 byte sectors, 300 byte sectors, whatever one wishes. On z/OS, several file formats require a 3120 byte block (=sector) size. Under CMS, the standard sector size is 800 bytes.
Hard disks (as opposed to floppies) commonly support non-power of 2 sector sizes, of 520 bytes or 528 bytes – for RAID systems which need to store additional metadata with each 512 byte sector. And with newer 4096 byte sector support, one sometimes finds support for 4104, 4160, and 4224 byte sectors. But, that's still not like CKD, in that most hard disks only support a fixed set of sector sizes (512 or 4096 base size plus some bytes of per-sector metadata), not arbitrary sector sizes like CKD does.
One thing that isn't covered is that 3.5" discs (and presumably 3" as well) actually do still have an index pulse, even though there's no index hole. The metal disc attached to the centre of the disc is keyed (the rectangular hole offset from the centre), so the drive generates the index pulse from the position of the spindle, rather than using an optical sensor to read the disc.
It is, but also because it's grabbing something with fixed orientation to the disk, it also means that the disc is exactly aligned with the spindle every time, and so the index pulse can be generated from a cheap electromagnetic sensor on the spindle instead of requiring an optical sensor.
However, it seems you might actually be right in practice. I think early 3.5" definitely did this, but while googling to determine what orientation of the hole triggers the index pulse came up with nothing, except a few patents about how you can fake an index pulse by using a timer and a speed sensor instead.
So maybe because almost nothing was actually using the index pulse anymore when 3.5" drives were introduced because almost every modern system (for the time) had settled on soft-sectoring, it's entirely possible that I've always been mistaken and there's never been a standardised index pulse for 3.5" drives between different drives relative to the "start of the track".
I guess this might explain why even though the Amiga allowed you to read the INDEX signal, it actually didn't use it and instead read a whole track at once starting when it saw a particular sync word.
It just hit me that when I was doing IT support for some departments in college I laughed at how antiquated tape storage backups were as I saved my papers and homework on to magnetic storage 3.5" floppy drives.
A friend in high-school used to do the old trick of copying a DLL onto a floppy disk, renaming it "essay.doc", and handing it in to the teacher. Next week he'd feign concern that the disk must have got corrupted somehow, but by then he'd bought himself an extra week to write the paper.
I had a QuickBASIC program that'd spit out a generic disk error screen, and later a Visual BASIC program that'd put up a Windows dialog saying that it needed to be open on NT 4 or later (school system was Win9x at the time). Good to hear we all slacked off the same :P
That is funny of course (in hindsight), at least if you tried to save on 1.44MB 3.5" floppies. When I was restoring all of my old tapes and floppies I could recover nearly all my 9-track CCT tapes, nearly 100% of my 5.25" floppies (various densities, including 1.2MB), many 720KB 3.5" floppies (8" ones are also good) and zero 1.44MB 3.5" floppies. Discussed this with many other people who had basically the same experience with HD 3.5". All of these were stored in the same physical environment. Some other tape media weren't particularly good (exabyte, 4mm tape etc, but then again those used to fail at the very beginning).
Might be because 3.5 disks outnumbered in production quantities any other kind of floppy disk, thus the average quality was lower? Massive production lowers costs, so to keep sane margins they just decreased quality?
Or is there any theory about the technology itself being less reliable?
From what I've gathered the issue was simply that for the HD 3.5" floppies the density was actually more than the medium could take. 720KB had lower density and didn't have that problem.
There are tools that record the actual magnetic flux from the disc, over multiple passes. In skilled hands, this has resulted in some "irreparable" discs being recovered. The most prominent one I know is KryoFlux.
If all the disks failed it could be the recording drive, assuming they weren't separately verified at the time or writing or since.
Our school's collection of BBCs included one drive that could read everything, including data it had created, but most of the other drives would reject anything it had written. Once identified that drive was taken out of general circulation and only used to read data for writing to a new disk in a second drive, and eventually skipped once it was unlikely any data someone cared about had been written by it and not copied elsewhere.
This often comes down to regularly testing old data. Floppies shouldn't have been used for archive storage but often were, with a problem only being noticed years down the line when the data needed to be read. Other media similarly: recorded CDs and DVDs from a decade or more ago are sometimes not readable now. An archive is only an archive if you can actually read it, much like an untested backup is just a hope not really a backup.
> The Victor 9000 was perhaps best known for how it was able to achieve such high density on it's floppy disks. It used variable speed disk drives; there were 9 different speeds used. As the drive head moved outward the speed would increase. It was really neat to hear the speed change as the drive head moved.
I miss portable physical storage in general. USB drives aren’t the same because they are both a media and interface. I guess the spiritual successor is SD cards but I would love a cheap, not necessarily huge or super fast storage, but with excellent resilience and archival properties for 50+ year storage. That’s a niche I feel like was never filled. The debate around CD/DVD/Blu-ray and various dyes never really died so I never dared using them for that. Best recommendation for home use (because tape is more enterprise oriented) seems to be multiple hard drives and verifying them every now and then, but that always felt like such a crutch, haha.
SD cards are a completely useless size these days: they're too small and easily lost, but they're too big for modern devices. For adding storage to a phone, for instance, a microSD card is much more useful (although easier lost of couse). A regular SD card is just way too big for that, and most of the space inside wasted.
A thumb drive is much better, as long as it not one of those thumb drives that's barely bigger than a USB port. But for offline storage that can be easily filed into a cabinet and paged through, floppy disks really were a great size. You can't file thumb drives; they're too narrow and thick, so a collection of them just ends up being a pile.
It's really too bad we don't have some kind of inexpensive, high-reliability, high-density offline storage format that resembles 3.5" floppies.
Regular SD cards are at least easy to handle. MicroSD are so darn tiny, sometimes pretty hard to pick up.
I know modern devices don't have a lot of space, but regular SD cards and mini-SIMs weren't really that big. (The full credit card sized original SIM card size is dramatically oversized though)
>MicroSD are so darn tiny, sometimes pretty hard to pick up.
This is true, but the idea is that you should almost never need to do this. It should be a rare occasion when you need to install or remove a microSD card. It's just like nanoSIM cards: when was the last time you needed to swap that out?
Floppy disks, and today USB thumb drives, serve a different purpose: they're supposed to be easy to plug and unplug and move data around. I don't see microSD cards as having that purpose at all. I have a microSD card in my phone, and it's great, but I haven't taken it out since I put it in there several years ago; it's just a storage expansion device, not a way of easily moving data between devices. It's not that easy to swap between devices, but it's easy enough to do, with some care, for those rare occasions it might be necessary (like migrating to a new phone).
I hate devices that have spring loaded microSD slots that practically launch the things when you remove them. The springs have way too much tension and shoot your microSD across the room once you push it in with your nail.
Yes, I really liked the memory stick and UMD formats. Too bad they were made by Sony and were thus uselessly crippled for any normal purposes.
You'd really think that Sony would have figured out that every single storage medium they've ever made has been an utter failure, but Sony is gonna Sony.
You'd really think that Sony would have figured out that every single storage medium they've ever made has been an utter failure, but Sony is gonna Sony.
This is a little funny in a thread about floppy disks, considering that the extremely popular and ubiquitous 3.5” Micro Floppy Disk was a Sony design ;)
There are also some other Sony formats that actually stuck, like DAT, Minidisc (yes, not so much in the US from what I gather, but it was A Thing in Europe, Australia and Japan) and Betacam.
Sony: Look at this awesome new storage format! Everyone should use it!
Also Sony: No, not like that.
Sony's positioning in storage formats is just weird. They spend untold money developing media formats intending to recoup the development costs licensing those formats. Then make the licensing fees and conditions so onerous that no one is willing to license them. The only time anyone touches a Sony influenced format is when it's developed and licensed by a consortium.
Sony has (or at least had; I haven't seen them try again for a while now) a long history of trying to push proprietary standards on everyone. They really seemed to think they could come up with some proprietary standard for something where other, more open, standards already existed, and people would somehow adopt theirs en masse.
It reminds me a lot of IBM when they introduced the PS/2 computers. They really, really thought that they could single-handedly kill all the PC "clones" by pushing their clunky, overpriced PC with a bunch of proprietary standards in it (namely the MCA bus). The people running these two companies had an extreme amount of arrogance and hubris.
CFexpress Type B cards (i.e. having the Sony XQD size) would be a perfect floppy disk replacement, except that these cards are overpriced and restricted to the niche of expensive photo cameras, instead of having a slot for them in any laptop and being available at the same price as the USB drives.
These cards are much sturdier than the SD cards, and much faster, due to the PCIe interface. If a laptop had a slot for them, they would have been much more convenient than external SSDs or USB drives, which protrude from the case or are hanging on cables.
https://retrocmp.de/ is also a great reference for particular drives. And if you need floppies, email Tom at https://www.floppydisk.com/ - he’s a top bloke and the disks I’ve got from him have all been in great working order.
The 3.5" Micro Floppy Disk was my favorite. I remember converting over 2000 files from 5.25" to 3.5" to save space at home; it took me months. Additionally, after a year and a half, I bought a 1x CD burner, which rendered my software collection obsolete within a matter of months too. Ah, those days of installing Windows 3.1 or 95 using disks, switching dozens of diskettes in the process!
The HP Itanium dev units we worked with at ActiveState in 2000 shipped with an Imation floppy drive. If I recall correctly, the install media for HPUX shipped on an Imation disk. It was a true beast.
I remember the Amiga let me customize the format with some tools. One could format more tracks than the specification to fill 900+ kB on a DD floppy beyond the initial 880 kB Amiga standard format. Gave important bragging rights when put against 720 kB PC plebs. Some games used it for copy protection i.e. copy protection on track 80+ https://eab.abime.net/showthread.php?t=101343 Of course, X-Copy and others knew and copied extra tracks.
Is there a similar post about LTO tape storage drives? I'm wondering how the newer generations achieve the insane transfer speeds and storage capacities.
There's quite a lot of similarity with how hard disks have dramatically increased their aerial bit density in that improvements in bit encoding and decoding techniques and DSP technology play quite a big part - actually LTO is behind hard disks on areal bit density as tapes have a different set of trade-offs as a removable media and archival format.
I have fond memories of 2MF 3.0 by Ciriaco García de Celis which I used to store an UC2 archive of a maximum of 2 megabyte on a single HD floppy. Booting a minimal DOS, the AUTOEXEC.BAT created a fast RAM disk D: where I extracted the UC2 contents in, a minimal windows 3.1 in d:\win - the last command was starting windows from the ram drive in the 8M machine.
I had to boot with this floppy on the library index computer as all the other text processing wordperfect 5.1 computers were taken by other students.
Windows write.exe was excellent as a wysywig wordprocessor.
https://en.wikipedia.org/wiki/2M_(DOS)
I still have some 5 1/4" floppies and I wanted to read them (if still possible at all). I still have an old drive, even an HD 1.2MB one.
However, it's literally impossible to find USB to 5 1/4" floppy interfaces now in Europe. I only found one outfit in the US that sells them, but they refused to ship to europe :(
If anyone has any tips I'd appreciate it because I'm pretty lost now.. I have much old hardware but nothing that can drive a 5 1/4" floppy drive. It's mostly unixy stuff like HP9000's and Macs.
If you’re willing to buy/be given/dumpster dive some e-waste, there are endless old-but-not-vintage PC motherboards that can be had for minimal money (like, $10-20 with CPU and RAM) that will still accept a 5.25” floppy drive and will still run a modern Windows or Linux for ease of transfer. Think Core 2 era. I have a HP Z400 workstation from ~2010 that will take a floppy drive. Beware though because some machines will be BIOS/chipset limited to only accept 3.5” 1.4MB drives. I think my HP is one of those.
EDIT: actually, no, I just looked it up — the Z400 will take a 1.2MB 5.25” drive. It will not take DD drives of any kind.
If you have the drive, you can probably bit-bang the interface pins with a Raspberry Pi or the like. There might already be something in a dusty corner of GitHub for this, or it would be a fun little project for anyone with hobbyist-level EE skills.
I've had very bad experiences with those. Most vendors don't want to send to those either and then cancel orders that were already paid, and it was a hassle to get my money back.
Just so happens I watched this on YouTube the other day. Someone is building up a board that will hook a floppy drive up to a MiSTer FPGA build so you'll be able to access it from the MiniMig Amiga emulator core. I'm always impressed by what people are doing in the retro scene these days.
This has reminded me of the awe felt by 12 year old me when someone showed me that you could use a hole punch to make a hole in the corner of a single sided 3.5" disk to double the capacity.
Adult me is now going to go look into how or why that worked.
I loved the sound the original Mac disk drives made as they sped up and down. Incompatible with everything else but it was at a time when that didn't matter all that much.
Indeed, in my experience, there's some kind of biofilm that forms on the disk surface. Doesn't look like anything more than a thin stain or discoloration, but disks so afflicted never read correctly. Suspect that whatever it is, it's between the mylar and the oxide layer so the bits literally flake off when the head hits those patches.
Why do I think it's biological -- just the shapes of the patches resemble what you see in laboratory petri dishes, and that the amount of such corruption seems to correlate to humidity. Disks from an attic seldom have this issue, compared to ones from basements.
The insides of the sleeve were lined with felt to cushion the disk and allow it to slide easily. It also unfortunately attracted moisture and could promote mold growth in very humid environments.
It also means if you drop it in water (as I once did) the felt will expand then be dragged along with the disk and literally crash into the head if you were foolish enough to put it in a drive. (Which I fortunately didn't.) Drying it out is not simple. But for the 3.5" types opening the shell is. You can then transfer the disk into a donor shell after giving it a good wipe with lint-free cloth and be on your way.
For dealing with mold on the disk itself, a brief Duck search reminds me that UV light kills most biologics. As does ozone. That will at least stop it from growing.
Jonathan, your deep dive into the intricacies of floppy disks is truly enlightening! It's fascinating to see the evolution and the technicalities behind something that's now a part of tech history. How do you think the transition from floppy disks to USBs impacted data storage and transfer?
The operator I knew frequently had this folded up in his pocket: You just had to smooth it out a bit, make sure it rotated inside its kevlar sleeve, and off you went.
(DECTAPE was famously robust, everything was written out about 5 times in the tape, it was designed for light industrial deployments. It was wide like old school 16 track audio tape, not like classic 1/2" 800/1200/9600 bpi computer tapes, before DAT/DLT cartridges and the like. But that said, the old tapes were pretty robust too)