One thing I've always wondered is why no-one tried to double the track density for read-only floppies.
The reason floppies have so few tracks (100000 bits per track, only 80 tracks per side!) is that the read-write mechanism didn't track the track. It was just indexed to a specific position by a stepper in the drive mechanism. Because there were a lot of sources of radial position error, and the writing and reading process could be misaligned in opposite directions, the tracks needed to be wide and far apart for reliability, and the erasehead needed to erase a much wider area than what the rwhead could read/write. The various later superfloppy standards that failed to get market traction usually only got slight increases in bit density along the track, and most of their capacity increase from having some way for the rwhead to follow the track, enabling dramatically higher track density. But all those mechanisms would be much more expensive and complicated to do than a simple floppy drive.
But if you use a more precise and expensive machine to write the disk, the normal amount of precision in the drive that is used for reading ought to be sufficient for about twice the track density. The only hw change needed would have been allowing half-step positions for the rwhead, for no cost increase. And I can imagine 80's/90's software devs being interested in a format that is harder to copy at home. Yet no-one ever did it, does anyone know why it was never tried?
> Perpendicular recording was later used by Toshiba in 3.5" floppy disks in 1989 to permit 2.88 MB of capacity (ED or extra-high density), but they failed to succeed in the marketplace.
I think that is something else entirely, unless you consider that to be the expensive writing version.
Adding a little capability (in this case permitting finer stepping) and letting software guys figure it out was often the way to great new abilities.
I often think about tiny little changes to old hardware that would have been an insignificant cost but added great functionality. Like a 4 bit latch xored with the rgbi output of CGA would have enabled a huge gain in what colour options were available. It would only be a few gates. You would still only have 320x200 in four colours but the colours available wouldn't melt your eyes.
Your inquiry reminds me of Game Boy. Some 4000 early Game Boy consoles had a specific hardware quirk that limited the refresh rate of the screen under a specific method. For the whole decade of existence of Game Boy, every single game was tested for this specific bug by Nintendo, and a few developers got burned because they wanted to use the refresh method but couldn't.
I bet it's a similar situation. Something like 99% of all drives could have handled what you're describing, but the 1% stopped anybody from trying your method.
On a similar note, a couple of days ago I bought two factory sealed magneto-optical disks, one 230 MB and the other 640 MB, for 300 yens each. I had never seen one in person, and I think that they look really cool in a retrofuturistic sense.
When I was a kid I remember reading and dreaming about all these alternative storage formats: MO, flopticals, Zip drive, LS-120, etc. The idea of a little disk with the capacity of a hundred floppies was amazing, it's a pity that none could replace them.
As far as I remember the Zip drives came closest... but were held back by being proprietary and therefore relatively expensive. And of course all of these were eventually replaced by USB sticks.
I put the death of Zip more on CD-Rs getting popular. Even if you didn’t have a writer you could read them. And then UsB thumb drives came in to clean house.
I think it was a mix of reasons: complexity (you needed scsi or parallel port available, CDR, internet, hard drive capacity increasing and evolution of work policies.
Zip drives were fairly expensive so weren't really a solution for archival. CDR and their immutability were more interesting for archival. Zip drives weren't useful for sharing data as they weren't OEM deivces so you wouldn't expect anyone to have the drive and it soon became easier to email files. So basically Zip drives were only useful as an extension of your own hard drive or to share files between 2 computers you owned that were equipped with a zip drive (say home and office). This has been gradually frowned upon by companies so this was only really a thing for people working independently. Additionally in the later years while the size of the usb flash drive could increase constantly, the zip drives had to evolve in parallel with the medias. An early adopter who had to upgrade from zip100 to zip250 then zip750 would have had 3 drives to purchase, even more so because of the click of death while a newer usb2.0 flash drive would work on an old computer with usb1.0 or 1.1 port.
I think this is probably right. Zip never quite got the penetration needed to take up the mantle of the floppy as far as I remember. And by the time it gained traction the CD burner was coming onto the scene in a big way. The market badly needed a higher capacity floppy like media, 1.44MB was a complete joke by the time floppy drives were killed. But if you had to choose between buying a zip drive or a CD burner the burner was the way to go. A burner was terrible for shuffling a few files around and early burning with janky disks and slow processors was a gamble, but copying games and albums or making mix CDs as well as the superior capacity just left it with broader utility than a zip drive with expensive disks had.
But CD-R and USB drives weren't mutually exclusive in my experience; a lot of PCs couldn't boot off of USB yet and early USB sticks didn't have the same capacity (I believe my first USB stick was 64 MB?). But, burning a CD was more involved, and re-recordable ones were expensive. Some time later we found you can buy spindles of burnable CDs for pennies per disk.
Going in the other direction, prior to the 5 1/4" disk drive (a true "floppy" disk) there was the (already rare by 1995) 8 inch floppy disk (and truly floppy) basically an enlarged 5.25" disk. Supposedly they only stored 1.2MB but I recall seeing 2.X MB models for sale at boeing surplus in the mid 1990s and feeling like that was a much more economical option than 1.44mb 3.5" disks. The whole world ran on 1.44mb as a hard limit on how large files could be. Any larger, and you could not transfer them -- email wasn't an option yet, and serial transfer between two computers was damn near impossible. Having an 8 inch disk with 2mb would allow much better images and longer multimedia documents to be transfered.
Alas, I could not convince my dad to shell out $45 for the 2.X MB 8" floppy drive, so I leaned on the 3x sony 1.44mb HDD I was given sometime in early elementary school all the way through middle school.
2.88mb would have been an absolute luxury. Sometime around 1998 we got a sony viao with the 100MB ZIP drive which felt palatial by comparison.
Oof, I felt that. I recall reading, many years ago, some statistics on how CD-ROM discs filled to about 600-650 megs made them much more reliable than the ones pushed to theoretical maximums. This was about regular, retail data CDs, not (re)writables.
I have always felt the floppy was a great demonstrator of the concept: a 3.5 inch might be perfectly fine for everyday usage and storage, but the moment you're trying to carry some data in 1.44 meg [0] zip portions, you're bound to find out just how many defective discs you had the entire time.
[0]okay, it was 1.37 meg of data actually IIRC but still
It's been a long time, so maybe I'm misremembering things, but I seem to remember ARJ being the "cool" replacement for ZIP, but then when RAR came along, ARJ quickly died out.
Also I remember ARJ and PKZIP being used legitimately while I can't think of a single time I encountered a rar file in a non piracy context. Actually when I got my first job fixing up desktop computer in a large public institution the presence of winrar was a sure sign that I was better off airgaping the computer immediately, scanning with an offline antivirus and backing up noninfected user data and then reinstall, hoping I would not encounter child porn.
Eeeeeh I spent two decades pirating stuff ending with being a co-admin for one of the largest warez sites in Central Europe and I can tell you in all this we were clean. (Note this ended twenty years ago.) There was enough trouble with the law for just stealing software we didn't need the forces hunting for CP on us too.
Those 8 inch disks were not enlarged 5¼-inch disks. The 5¼-inch ones were shrunk 8 inch disks.
They were a fairly inconvenient form factor, the main benefit was the much reduced density, which made them super resistant to wear-and-tear. A lot of 8 inch floppies written in the 1980s still work just fine: last time I checked, BART used them.
I'm not sure they were ever commercially available in high density configurations.
Not me either, I used some utility whose name I've forgotten (might be fdformat as mentioned on that page) to get something like 1.74MB pretty reliably. I think I needed a driver loaded for it to work, but seem to remember either Windows 95 or 98 supporting it natively.
That was totally different. DoubleSpace made a virtual drive in a file and did software compression/decompression with a driver during writes and reads. It was a clone of Stacker, and both were slow as heck (but hard drives were only 40MB, so sometimes it was worth it).
There are some old drives that store 2 Mb per disk, but that's megabits, not megabytes. It seems the very largest 8" disks ever achieved was 1.2MB, exactly the same as the 5.25".
> Going in the other direction, prior to the 5 1/4" disk drive (a true "floppy" disk)
The “floppy disk” is named for the round mylar storage medium inside the square casing, which is both floppy and a disk in all the sizes they came in, not the square casing, which, while somewhat flexible in 8” and 5¼” sizes, is not really floppy, and absolutely not a disk, for any of them.
the 8" floppies predate the 5.25 ones (i.e. the original size when invented by IBM). 5.25 and then 3.5 came later.
When I was growing up, we had an old Z-80 (I think?) cpm machine that had a tiny crt and 8" floppy drives. We then had a TI-99/4a that used 5.25" and then a panasonic 8086 pc compat also with 5.25" From there we moved to a 386 with both a 5.25 and 3.5 drive, but that was it for the 5.25s
I still have some of the games from the 5.25 era on their original disks, though last time I glanced at both ebay and archive sites, there wasn't much value to keeping them as readily available, but hard to toss that element of my childhood.
Interesting I was not aware of that. I was told the reason was that in addition to a different format, the drive was more flexible so that it could adjust more dynamically. But it seems that you could store it more efficiently and get similar results on PC.
That said, I found a forum post that claimed an Amiga could store 1MB (1.1 in the comments) on a DD disk (standard size is 880k on Amiga, 720k on PC). So perhaps an HD disk could have been formatted to store over 2MB? I wouldn't be surprised if there were some drawbacks to this method that makes data corruption more probable.
HD 3.5" disks have a 2MB unformatted capacity, so I don't think anything could go over that. The Amiga drive was certainly more flexible (and some games used weird custom formats as a form of copy protection).
>small pockets of Amiga users that could store 1.76mb
realy reaaally tiny fraction of Amigas userbase had access to HD floppy drive only ever shipping in ONE top of the line most expensive model 4000. It was a custom modified Chinon FZ357A spinning at half rpm, needed because Amiga could only handle 250Kbit/s data rate while standard HD floppy drive uses 500Kbit/s. Commodore being hugely mismanaged didnt invest anything in engineering, they lacked will and resources to update technology in its subsequent models. 1994 A4000T, the very last Amiga ever manufactured shipped with the very same Chip responsible for sound and Double Density floppy interfacing as the very first model 1000 from 1985.
> drive was more flexible
Stock Amiga FDD was standard and could be replaced with PC after little pinout modification. Controller on the other hand could be called software defined as it lacked any smarts. It could only decode/encode raw datastream to/from ram buffer whole track at a time. In contrast PC controllers target specific sectors.
In the late nineties I had a laptop that supported SuperDisk¹ floppies that could store up to 120 MB. The disks were pretty expensive and the same laptop even had a CD-RW drive so I never got around to try it.
Those drives were nice. Windows 98 could fit on a ls-120, but not a zip-100. And, they were really good at reading regular floppy disks too. Handy as media seemed to be getting worse and worse towards the end.
> Much less known is ED 2.88MB. Only ever shipped in some 1990 IBM PS/2 and NEXT computers. 80 tracks ~100KB/s speed.
It shipped in some 1990s IBM RS/6000s. I know for sure it was available on the Model 250, which came out in 1993 (and was one of the first PowerPC systems).
I remember encountering one of these PS/2s in my school's storage in the late 90s and was kind of thrown for a loop to see 2.88 printed on the blue eject button. Until then I was unaware that standard existed and didn't understand why it hadn't take off, I assumed it was just another proprietary IBM standard that the market rejected.
Thanks for Adafruit link. I didnt know they picked up my Vogons thread and actually found picture of dedicated floppies. Sadly 2TD drives seem to be unobtanium :(
Technically its extended ED and should be as reliable, even the media should be interchangeable. I would love to get my hands on one too :)
I recall seeing 2.88mb disk drives in Tiger Electronics catalogues circa 1997. Someone with more time than me will find PDF scans of them and can verify. 1.44mb HD 3.5" disks were totally the norm as our elementary and middle schools were flush with Apple LCII and LCIII (LC meaning "low cost") computers of the era. If you were going to play Bolo, you needed your own 1.44mb disk to keep a copy of it to use at the computer lab.
It's not so much that storage sucked, as the interfaces for storage sucked.
We didn't have much you could plug a better drive into. The floppy cable's protocol was too restrictive to support anything much fancier, exposing too many internal details. And the parallel port was awfully slow. And SCSI was unfortunately absent from the consumer hardware.
That made it really hard for anyone to make a better storage device, because it'd find itself immediately constrained.
It's only in modern times where we finally have technology-agnostic interfaces that make it possible to use any storage tech you like, bandwidth to spare, and a generic USB storage driver meaning that even if you do own some oddball device you can still plug it into any random computer and have it work out of the box.
ISA/PCI bus? (And IDE, which is just the ISA bus over a differently shaped cable)
In that era, addon cards were quite normal - not like today when the addon card slot is de-facto just a graphics card slot and motherboards are being designed around that assumption.
Nowadays a 2TB nvme USB drive that is barely bigger old usb pen drives. I wouldn't call that "officially shit".
Also, SD and MicroSD cards are fairly reliable in my experience as long as you don't stretch too much their usable life. But that has always been the case of flash technology and it was even worse in the floppy and CDRW days.
That is true and that is what I use when I know I'll need one. But they are overkill, expensive and thus as a consequence - not something you have on you when you need it.
And for times you need to create a boot disk and similar use-cases there is a lot of value in being able to completely nuke and reformat a drive.
Which you might not want to do if you have data on the 2TB drive you'd rather keep. Even if it is only for convenience - which you typically do after a while.
And usb-drives you can have a handful of, give one away without care etc. They do have a place and it is a shame we can't trust them (for no good reason).
You could format a CD-RW to use it pretty much like a (slow) USB flash drive. I don't remember the details but it definitely was an option in Windows.
But then, before CD burners became popular and blanks cheap, most files I had to carry between computers (e.g. to school or to/from a friend) fit on a floppy. In the rare case they did not, I used WinRAR to make one of those split archives.
>It's amazing (in a bad way) how storage sucked in the 90s, early 2000s
>Floppy technology sucked even back then.
No, not "even": floppy technology sucked in the 90s, but it absolutely did NOT suck in the 80s. Floppies were extremely, even ridiculously reliable back in the 80s, and into the early 90s. They totally went to shit in the mid-to-late 90s because the quality of both the media and the drives went down the toilet.
Absolutely agree. Source: I've read (in the last decade) tons of old floppies from (mostly) the late eighties. The recovery rate for 5.25" DD and HD floppies were nearly 100%. I had only a couple of floppies with issues (though it must be said that I know people who have had more problems, but that is always about Really Bad Storage Conditions).
But, as I've mentioned before in other threads, this does not translate to 3.5" floppies. Or at least not to HD 3.5" floppies. Even though mine were stored in the same conditions as the 5.25" and 8" floppies, the success rate was really low, and mostly non-existent for "HD" 3.5" floppies (the so-called 1.44MB floppies). The 720KB ones fared much better, but never as good as the older 5.25" ones. Of course that also when the nineties arrived, but more than that - the HD 3.5" floppies stretched the density too high for the medium, according to Chuck(G) and others with more understanding of the physical medium than myself. In any case my experience supports that claim.
So, the title of this thread is "Triple Density Floppy, Anyone?"
Well, obviously I have no belief in the feasibility of trying even higher density than what existed, with the problems already apparent with the existent densities (mind, the 5.25" ones didn't exceed what was possible).
As for other media.. CD-R used to fail reading for me after a year in storage. I quickly stopped using that as any kind of backup. And we all know about flash-based storage.. or at least I hope we all know. They're just like very slowly leaking capacitors. Spinning rust or spinning floppy media or even tape retains data for much longer (yes I have tons of CCTs which are also still readable, from back to the beginning of the eighties)
I used lots of 1.44MB floppies in the early 90s and generally had few, if any problems, as I recall. But later in the decade, I had more and more trouble, and by the late 90s or early 2000s they could absolutely not be relied upon at all.
I wonder if you went through all your HD 3.5" floppies, and sorted them by manufacturing date, if you'd find a trend showing the early ones being more reliable than the later ones.
It's more like how long it was since they were written - I couldn't find any other pattern (I started reading through old floppies a long time ago). They simply seemed to fade with age.
One thing I've always wondered is why no-one tried to double the track density for read-only floppies.
The reason floppies have so few tracks (100000 bits per track, only 80 tracks per side!) is that the read-write mechanism didn't track the track. It was just indexed to a specific position by a stepper in the drive mechanism. Because there were a lot of sources of radial position error, and the writing and reading process could be misaligned in opposite directions, the tracks needed to be wide and far apart for reliability, and the erasehead needed to erase a much wider area than what the rwhead could read/write. The various later superfloppy standards that failed to get market traction usually only got slight increases in bit density along the track, and most of their capacity increase from having some way for the rwhead to follow the track, enabling dramatically higher track density. But all those mechanisms would be much more expensive and complicated to do than a simple floppy drive.
But if you use a more precise and expensive machine to write the disk, the normal amount of precision in the drive that is used for reading ought to be sufficient for about twice the track density. The only hw change needed would have been allowing half-step positions for the rwhead, for no cost increase. And I can imagine 80's/90's software devs being interested in a format that is harder to copy at home. Yet no-one ever did it, does anyone know why it was never tried?
https://en.wikipedia.org/wiki/Perpendicular_recording - Apparently what you're describing
> Perpendicular recording was later used by Toshiba in 3.5" floppy disks in 1989 to permit 2.88 MB of capacity (ED or extra-high density), but they failed to succeed in the marketplace.
I think that is something else entirely, unless you consider that to be the expensive writing version.
Adding a little capability (in this case permitting finer stepping) and letting software guys figure it out was often the way to great new abilities.
I often think about tiny little changes to old hardware that would have been an insignificant cost but added great functionality. Like a 4 bit latch xored with the rgbi output of CGA would have enabled a huge gain in what colour options were available. It would only be a few gates. You would still only have 320x200 in four colours but the colours available wouldn't melt your eyes.
No, this is about increasing bit density inside a track by placing the magnetic domains in a different orientation.
The floppy standard failed, but it became the norm in hard disk drives as soon as the patent on the special rwhead that was needed for it expired.
Your inquiry reminds me of Game Boy. Some 4000 early Game Boy consoles had a specific hardware quirk that limited the refresh rate of the screen under a specific method. For the whole decade of existence of Game Boy, every single game was tested for this specific bug by Nintendo, and a few developers got burned because they wanted to use the refresh method but couldn't.
I bet it's a similar situation. Something like 99% of all drives could have handled what you're describing, but the 1% stopped anybody from trying your method.
On a similar note, a couple of days ago I bought two factory sealed magneto-optical disks, one 230 MB and the other 640 MB, for 300 yens each. I had never seen one in person, and I think that they look really cool in a retrofuturistic sense.
When I was a kid I remember reading and dreaming about all these alternative storage formats: MO, flopticals, Zip drive, LS-120, etc. The idea of a little disk with the capacity of a hundred floppies was amazing, it's a pity that none could replace them.
As far as I remember the Zip drives came closest... but were held back by being proprietary and therefore relatively expensive. And of course all of these were eventually replaced by USB sticks.
Zip drives got killed by the click of death scandal
I put the death of Zip more on CD-Rs getting popular. Even if you didn’t have a writer you could read them. And then UsB thumb drives came in to clean house.
I think it was a mix of reasons: complexity (you needed scsi or parallel port available, CDR, internet, hard drive capacity increasing and evolution of work policies.
Zip drives were fairly expensive so weren't really a solution for archival. CDR and their immutability were more interesting for archival. Zip drives weren't useful for sharing data as they weren't OEM deivces so you wouldn't expect anyone to have the drive and it soon became easier to email files. So basically Zip drives were only useful as an extension of your own hard drive or to share files between 2 computers you owned that were equipped with a zip drive (say home and office). This has been gradually frowned upon by companies so this was only really a thing for people working independently. Additionally in the later years while the size of the usb flash drive could increase constantly, the zip drives had to evolve in parallel with the medias. An early adopter who had to upgrade from zip100 to zip250 then zip750 would have had 3 drives to purchase, even more so because of the click of death while a newer usb2.0 flash drive would work on an old computer with usb1.0 or 1.1 port.
I think this is probably right. Zip never quite got the penetration needed to take up the mantle of the floppy as far as I remember. And by the time it gained traction the CD burner was coming onto the scene in a big way. The market badly needed a higher capacity floppy like media, 1.44MB was a complete joke by the time floppy drives were killed. But if you had to choose between buying a zip drive or a CD burner the burner was the way to go. A burner was terrible for shuffling a few files around and early burning with janky disks and slow processors was a gamble, but copying games and albums or making mix CDs as well as the superior capacity just left it with broader utility than a zip drive with expensive disks had.
But CD-R and USB drives weren't mutually exclusive in my experience; a lot of PCs couldn't boot off of USB yet and early USB sticks didn't have the same capacity (I believe my first USB stick was 64 MB?). But, burning a CD was more involved, and re-recordable ones were expensive. Some time later we found you can buy spindles of burnable CDs for pennies per disk.
Most people didn't care about booting a CD, even more so a custom CD.
Oooh ... I had totally forgotten about that! Thanks for the reminder - now I can hear it in my head.
In my experience, writeable and rewriteable CDs killed Zip drives well before USB drives could hold as much.
Going in the other direction, prior to the 5 1/4" disk drive (a true "floppy" disk) there was the (already rare by 1995) 8 inch floppy disk (and truly floppy) basically an enlarged 5.25" disk. Supposedly they only stored 1.2MB but I recall seeing 2.X MB models for sale at boeing surplus in the mid 1990s and feeling like that was a much more economical option than 1.44mb 3.5" disks. The whole world ran on 1.44mb as a hard limit on how large files could be. Any larger, and you could not transfer them -- email wasn't an option yet, and serial transfer between two computers was damn near impossible. Having an 8 inch disk with 2mb would allow much better images and longer multimedia documents to be transfered.
Alas, I could not convince my dad to shell out $45 for the 2.X MB 8" floppy drive, so I leaned on the 3x sony 1.44mb HDD I was given sometime in early elementary school all the way through middle school.
2.88mb would have been an absolute luxury. Sometime around 1998 we got a sony viao with the 100MB ZIP drive which felt palatial by comparison.
>The whole world ran on 1.44mb as a hard limit on how large files could be.
pkzip spanning disks was a normal thing back then. I'm totally not showing my age by knowing that (and having used it a lot).
Oh yes, and in a set of 14 disks, it was always the last one which was corrupt.
Oof, I felt that. I recall reading, many years ago, some statistics on how CD-ROM discs filled to about 600-650 megs made them much more reliable than the ones pushed to theoretical maximums. This was about regular, retail data CDs, not (re)writables.
I have always felt the floppy was a great demonstrator of the concept: a 3.5 inch might be perfectly fine for everyday usage and storage, but the moment you're trying to carry some data in 1.44 meg [0] zip portions, you're bound to find out just how many defective discs you had the entire time.
[0]okay, it was 1.37 meg of data actually IIRC but still
RAR was even better for multi-volume archives.
Among me and my peers we used arj.exe
It's been a long time, so maybe I'm misremembering things, but I seem to remember ARJ being the "cool" replacement for ZIP, but then when RAR came along, ARJ quickly died out.
LHA(lzh/lharc) was also a cool replacement at some point, especially since it was just one binary for both compressing and decompressing.
The switch didn't happen until 1995 when WinRAR came along, if memory serves. RAR was 1993 and ARJ was 1991.
Also I remember ARJ and PKZIP being used legitimately while I can't think of a single time I encountered a rar file in a non piracy context. Actually when I got my first job fixing up desktop computer in a large public institution the presence of winrar was a sure sign that I was better off airgaping the computer immediately, scanning with an offline antivirus and backing up noninfected user data and then reinstall, hoping I would not encounter child porn.
Eeeeeh I spent two decades pirating stuff ending with being a co-admin for one of the largest warez sites in Central Europe and I can tell you in all this we were clean. (Note this ended twenty years ago.) There was enough trouble with the law for just stealing software we didn't need the forces hunting for CP on us too.
Especially with par files for those pesky corrupt disks.
> basically an enlarged 5.25" disk
Those 8 inch disks were not enlarged 5¼-inch disks. The 5¼-inch ones were shrunk 8 inch disks.
They were a fairly inconvenient form factor, the main benefit was the much reduced density, which made them super resistant to wear-and-tear. A lot of 8 inch floppies written in the 1980s still work just fine: last time I checked, BART used them.
I'm not sure they were ever commercially available in high density configurations.
> The whole world ran on 1.44mb as a hard limit on how large files could be.
Not Microsoft, they used 1.68MB: https://en.wikipedia.org/wiki/Distribution_Media_Format
Not me either, I used some utility whose name I've forgotten (might be fdformat as mentioned on that page) to get something like 1.74MB pretty reliably. I think I needed a driver loaded for it to work, but seem to remember either Windows 95 or 98 supporting it natively.
Sounds like DoubleSpace (later renamed to DriveSpace), which was built into MSDOS and early Windows versions.
https://en.m.wikipedia.org/wiki/DriveSpace
That was totally different. DoubleSpace made a virtual drive in a file and did software compression/decompression with a driver during writes and reads. It was a clone of Stacker, and both were slow as heck (but hard drives were only 40MB, so sometimes it was worth it).
FDFORMAT is probably the one you are thinking of
https://github.com/christoh/fdformat
> The whole world ran on 1.44mb as a hard limit on how large files could be.
Not the whole world. In the late 80's through early 90's the publishing industry used 44mb SyQuest disks to transport files.
[dead]
>prior to the 5 1/4" disk drive (a true "floppy" disk)
The 3.5" disks were "floppy" too: you just had to take the actual disc out out of its rigid protective shell.
>Supposedly they only stored 1.2MB but I recall seeing 2.X MB models for sale at boeing surplus in the mid 1990s
I can't find anything about this on this Wikipedia page: https://en.wikipedia.org/wiki/List_of_floppy_disk_formats
There are some old drives that store 2 Mb per disk, but that's megabits, not megabytes. It seems the very largest 8" disks ever achieved was 1.2MB, exactly the same as the 5.25".
> Going in the other direction, prior to the 5 1/4" disk drive (a true "floppy" disk)
The “floppy disk” is named for the round mylar storage medium inside the square casing, which is both floppy and a disk in all the sizes they came in, not the square casing, which, while somewhat flexible in 8” and 5¼” sizes, is not really floppy, and absolutely not a disk, for any of them.
>not the square casing, which, while somewhat flexible in 8” and 5¼” sizes, is not really floppy, and absolutely not a disk, for any of them.
Not true. The square casing plus the circular thing inside was absolutely a "disk". But it was not a "disc". The thing inside was a "disc".
Depends where you're from. In American English, both are disks: https://en.wiktionary.org/wiki/disk#Usage_notes
the 8" floppies predate the 5.25 ones (i.e. the original size when invented by IBM). 5.25 and then 3.5 came later.
When I was growing up, we had an old Z-80 (I think?) cpm machine that had a tiny crt and 8" floppy drives. We then had a TI-99/4a that used 5.25" and then a panasonic 8086 pc compat also with 5.25" From there we moved to a 386 with both a 5.25 and 3.5 drive, but that was it for the 5.25s
I still have some of the games from the 5.25 era on their original disks, though last time I glanced at both ebay and archive sites, there wasn't much value to keeping them as readily available, but hard to toss that element of my childhood.
> The whole world ran on 1.44mb
There were small pockets of Amiga users that could store 1.76mb on the same floppy disk due to a different type of hardware
That was possible on DOS with the standard hardware, just different formatting: https://en.wikipedia.org/wiki/Distribution_Media_Format, https://en.wikipedia.org/wiki/Fdformat#DOS_tool
I relied on it for years until CD writers became common and affordable.
Interesting I was not aware of that. I was told the reason was that in addition to a different format, the drive was more flexible so that it could adjust more dynamically. But it seems that you could store it more efficiently and get similar results on PC.
That said, I found a forum post that claimed an Amiga could store 1MB (1.1 in the comments) on a DD disk (standard size is 880k on Amiga, 720k on PC). So perhaps an HD disk could have been formatted to store over 2MB? I wouldn't be surprised if there were some drawbacks to this method that makes data corruption more probable.
Ref: https://eab.abime.net/showthread.php?t=102630
HD 3.5" disks have a 2MB unformatted capacity, so I don't think anything could go over that. The Amiga drive was certainly more flexible (and some games used weird custom formats as a form of copy protection).
https://en.wikipedia.org/wiki/Floppy_disk_variants#Amiga
Some info about different sectors, tracks, etc.: https://retrocomputing.stackexchange.com/questions/27412/how..., https://retrocomputing.stackexchange.com/questions/12768/wha...
>small pockets of Amiga users that could store 1.76mb
realy reaaally tiny fraction of Amigas userbase had access to HD floppy drive only ever shipping in ONE top of the line most expensive model 4000. It was a custom modified Chinon FZ357A spinning at half rpm, needed because Amiga could only handle 250Kbit/s data rate while standard HD floppy drive uses 500Kbit/s. Commodore being hugely mismanaged didnt invest anything in engineering, they lacked will and resources to update technology in its subsequent models. 1994 A4000T, the very last Amiga ever manufactured shipped with the very same Chip responsible for sound and Double Density floppy interfacing as the very first model 1000 from 1985.
> drive was more flexible
Stock Amiga FDD was standard and could be replaced with PC after little pinout modification. Controller on the other hand could be called software defined as it lacked any smarts. It could only decode/encode raw datastream to/from ram buffer whole track at a time. In contrast PC controllers target specific sectors.
Article about playing with stuffing more data on PC floppies https://www.os2museum.com/wp/floppy-capacity-math/
In the late nineties I had a laptop that supported SuperDisk¹ floppies that could store up to 120 MB. The disks were pretty expensive and the same laptop even had a CD-RW drive so I never got around to try it.
[1] https://en.wikipedia.org/wiki/SuperDisk
Those drives were nice. Windows 98 could fit on a ls-120, but not a zip-100. And, they were really good at reading regular floppy disks too. Handy as media seemed to be getting worse and worse towards the end.
SuperDisk was far faster for random access than CD-RWs. Even with normal floppy disks. I liked them a lot.
> Much less known is ED 2.88MB. Only ever shipped in some 1990 IBM PS/2 and NEXT computers. 80 tracks ~100KB/s speed.
It shipped in some 1990s IBM RS/6000s. I know for sure it was available on the Model 250, which came out in 1993 (and was one of the first PowerPC systems).
Documentation: https://www.ardent-tool.com/RS6000/docs/pdf/7011_Operator_Gu...
I remember encountering one of these PS/2s in my school's storage in the late 90s and was kind of thrown for a loop to see 2.88 printed on the blue eject button. Until then I was unaware that standard existed and didn't understand why it hadn't take off, I assumed it was just another proprietary IBM standard that the market rejected.
Always wanted one of these, not sure if it would work with a greaseweazle otherwise I'd probably add one to the collection.
With HD 1.44 MB SCSI and ED 2.88 MB drives, my collection lacking a 2TD drive shall not stand.
Other reference from Adafruit containing an image of the media this drive took: https://blog.adafruit.com/2024/10/31/yes-there-was-a-13mb-tr...
I would bet my left pinky finger that it was even more fragile and prone to errors than HD much less ED media.
Thanks for Adafruit link. I didnt know they picked up my Vogons thread and actually found picture of dedicated floppies. Sadly 2TD drives seem to be unobtanium :(
Technically its extended ED and should be as reliable, even the media should be interchangeable. I would love to get my hands on one too :)
I recall seeing 2.88mb disk drives in Tiger Electronics catalogues circa 1997. Someone with more time than me will find PDF scans of them and can verify. 1.44mb HD 3.5" disks were totally the norm as our elementary and middle schools were flush with Apple LCII and LCIII (LC meaning "low cost") computers of the era. If you were going to play Bolo, you needed your own 1.44mb disk to keep a copy of it to use at the computer lab.
I once played with a DEC desktop that had 2.88 drives, but I never saw any media for them. This must have been the late 80s or early 90s.
It's amazing (in a bad way) how storage sucked in the 90s, early 2000s
Floppy technology sucked even back then. Then we got CDs, which were fine, but read-only
Iomega came up with Zip drives only to promptly shoot themselves in the foot instead of standardizing the format and fixing their issues
There were also a couple of also-rans that never got much traction
USB drives got there but they had to wait for USB and good enough/big enough flash chips to exist
It's not so much that storage sucked, as the interfaces for storage sucked.
We didn't have much you could plug a better drive into. The floppy cable's protocol was too restrictive to support anything much fancier, exposing too many internal details. And the parallel port was awfully slow. And SCSI was unfortunately absent from the consumer hardware.
That made it really hard for anyone to make a better storage device, because it'd find itself immediately constrained.
It's only in modern times where we finally have technology-agnostic interfaces that make it possible to use any storage tech you like, bandwidth to spare, and a generic USB storage driver meaning that even if you do own some oddball device you can still plug it into any random computer and have it work out of the box.
ISA/PCI bus? (And IDE, which is just the ISA bus over a differently shaped cable)
In that era, addon cards were quite normal - not like today when the addon card slot is de-facto just a graphics card slot and motherboards are being designed around that assumption.
No, floppies are not IDE
They're this weird interface that sends data byte by byte IIRC (or even worse)
Not IDE, it has none of its smartness
No, IDE is not floppies. It's IDE.
The floppy drive in the Itanium I owned used the IDE interface. Go figure.
I'm sad to say that usb-drives are now officially shit.
They are so unreliable and there is no brand that you can trust.
Which truly sucks for the rare cases when you actually do need some form of physical media.
Nowadays a 2TB nvme USB drive that is barely bigger old usb pen drives. I wouldn't call that "officially shit".
Also, SD and MicroSD cards are fairly reliable in my experience as long as you don't stretch too much their usable life. But that has always been the case of flash technology and it was even worse in the floppy and CDRW days.
That is true and that is what I use when I know I'll need one. But they are overkill, expensive and thus as a consequence - not something you have on you when you need it.
And for times you need to create a boot disk and similar use-cases there is a lot of value in being able to completely nuke and reformat a drive.
Which you might not want to do if you have data on the 2TB drive you'd rather keep. Even if it is only for convenience - which you typically do after a while.
And usb-drives you can have a handful of, give one away without care etc. They do have a place and it is a shame we can't trust them (for no good reason).
> Then we got CDs, which were fine, but read-only
You could format a CD-RW to use it pretty much like a (slow) USB flash drive. I don't remember the details but it definitely was an option in Windows.
But then, before CD burners became popular and blanks cheap, most files I had to carry between computers (e.g. to school or to/from a friend) fit on a floppy. In the rare case they did not, I used WinRAR to make one of those split archives.
>It's amazing (in a bad way) how storage sucked in the 90s, early 2000s >Floppy technology sucked even back then.
No, not "even": floppy technology sucked in the 90s, but it absolutely did NOT suck in the 80s. Floppies were extremely, even ridiculously reliable back in the 80s, and into the early 90s. They totally went to shit in the mid-to-late 90s because the quality of both the media and the drives went down the toilet.
Absolutely agree. Source: I've read (in the last decade) tons of old floppies from (mostly) the late eighties. The recovery rate for 5.25" DD and HD floppies were nearly 100%. I had only a couple of floppies with issues (though it must be said that I know people who have had more problems, but that is always about Really Bad Storage Conditions).
But, as I've mentioned before in other threads, this does not translate to 3.5" floppies. Or at least not to HD 3.5" floppies. Even though mine were stored in the same conditions as the 5.25" and 8" floppies, the success rate was really low, and mostly non-existent for "HD" 3.5" floppies (the so-called 1.44MB floppies). The 720KB ones fared much better, but never as good as the older 5.25" ones. Of course that also when the nineties arrived, but more than that - the HD 3.5" floppies stretched the density too high for the medium, according to Chuck(G) and others with more understanding of the physical medium than myself. In any case my experience supports that claim.
So, the title of this thread is "Triple Density Floppy, Anyone?" Well, obviously I have no belief in the feasibility of trying even higher density than what existed, with the problems already apparent with the existent densities (mind, the 5.25" ones didn't exceed what was possible).
As for other media.. CD-R used to fail reading for me after a year in storage. I quickly stopped using that as any kind of backup. And we all know about flash-based storage.. or at least I hope we all know. They're just like very slowly leaking capacitors. Spinning rust or spinning floppy media or even tape retains data for much longer (yes I have tons of CCTs which are also still readable, from back to the beginning of the eighties)
I used lots of 1.44MB floppies in the early 90s and generally had few, if any problems, as I recall. But later in the decade, I had more and more trouble, and by the late 90s or early 2000s they could absolutely not be relied upon at all.
I wonder if you went through all your HD 3.5" floppies, and sorted them by manufacturing date, if you'd find a trend showing the early ones being more reliable than the later ones.
It's more like how long it was since they were written - I couldn't find any other pattern (I started reading through old floppies a long time ago). They simply seemed to fade with age.
Yeah, kinda. I remember running into a lot of read errors with the 5¼ disks (I think being floppy also contributed to their unreliability)
But yes there was a point in the 90s where quality went down even for the 3½ ones (and that elementary pre-IDE floppy interface certainly didn't help)