17

According to https://en.wikipedia.org/wiki/Floppy_disk#8-inch_and_5%C2%BC-inch_disks

A small notch on the side of the disk identifies whether it is writable, as detected by a mechanical switch or photoelectric sensor. In the 8-inch disk, the notch being covered or not present enables writing, while in the 5¼-inch disk, the notch being present and uncovered enables writing.

As a Commodore 1541 disk drive user in the 80s, I would have preferred the retention of the 8-inch mode. Then I wouldn't have had to keep a nail scissors on my computer desk.

Why did 5.25″ disks invert the mode? Did Shugart et al. leave any record of the rationale behind this decision? Or did anyone who used them, find the inverted mode helpful in some way?

9
  • 1
    forum.vcfed.org/index.php?threads/… has one supposition, but no citation for it.
    – Jon Custer
    Commented May 30 at 14:08
  • 5
    My recollection doesn't quite go back that far, but if 5.25 disks generally had photo sensors rather than mechs, it would make sense that notch-open = write; that way, if the photo fails, it defaults to "don't write", which is the "safe" way to fail. Commented May 30 at 14:09
  • 13
    It's obvious isn't it? It makes it easy to turn a read/write diskette into a read only diskette simply by covering the notch with a bit of opaque sticky tape. That's probably a more common use-case than wanting to turn a read only diskette into a read/write diskette.
    – JeremyP
    Commented May 30 at 14:32
  • 2
    For the record, I could make neither head nor tail of the post Jon Custer links to, i.e. "You can manufacture write-protected 5.25" floppies that cannot be write-enabled easily, yet it's easy to manufacture write enabled floppies just as easily." So it's easy to make one sort and, by the way, just as easy to make the other? I'm going to assume it's meant to say the same thing as Dirkt but can't seem to find a construction compatible with that, no matter to what I attach the various 'easy/easily's. Maybe I'm having a bad day.
    – Tommy
    Commented May 30 at 17:47
  • 11
    Why did you need scissors? Floppies were sold with the notch (so read-write), and you could write-protect them by covering the notch with the small sticker (provided). And back to read-write by removing the tape. The only case you needed scissors was when you wanted to take a single-sided floppy and use both sides: you would cut a notch on the other edge to make the second side writable by flipping the floppy over. But that's an abuse of the intended use of the floppy.
    – jcaron
    Commented May 31 at 0:07

6 Answers 6

16

TL;DR

If all disks are produced with a notch there is no fundamental difference between "cover it to make it read-only" vs. "uncover it to make it read-only". But essentially:

  • 8" initially read-only, so no reason for a notch at all.
  • First read-write 8" used the same media design and end-user write-protecting didn't matter much.
  • 5-1/4" a total redesign considering increased usage with microcomputers, bulk software distribution and other factors, so reversing the notch status made sense.

Full Explanation

I have a logical (to me at least) sequence of how this all came about:

According to Wikipedia, the first 8" floppy drives were developed by IBM primarily as a read only device. Obviously the 8" disks had to be written at some point, but IBM did that internally with other equipment and apparently the field units were read-only. Since the field units were read only, a notch was not needed. The disk's mere existence in an end-user's hand (a mainframe admin user) made it read-only. Cutting a notch wouldn't make a difference - if you needed an updated disk you got it from IBM.

That effectively meant that disks used by others, as they were developed over the next few years, used the same basic media design. No notch at all. Then somebody (I don't know who/when, I suspect between 1972 and 1976) came up with the idea that since end-users were now able to use these disks in a rewritable fashion, that write protection was important. That just didn't matter in the initial, primarily mainframe, usage as in that context floppy disks were for booting and similar uses and not for actual end-user data storage. Since there were already drives/disks without the notch, and at the time archival use was handled primarily with tape drives, the read-only (beyond mainframe read-only boot disks as originally used, which was read-only by virtue of the drive design) usage was relatively limited, so making the default no-notch read-write made sense. A software manufacturer could easily produce disks with a notch, and an end-user, if they felt the need to make a disk read-only for whatever reason, could carefully cut a notch of their own. But the typical user (e.g., early CP/M days) didn't need to make disks read-only all that often.

By 1976 with the development of 5-1/4" floppy drives, there was a clear need for easy conversion between read-only (for end-user archives and for mass-production of software). Affordable single-user desktop computers had arrived, but affordable hard drives had not. That meant both more user archives (write a book and then make an extra copy and write-protect it; make a copy of your yearly accounting records and write-protect it; etc.) and more mass-produced software (produce it without a notch, so users won't accidentally reuse it for something else after the tape falls off - they have to deliberately cut a notch to reuse because your software is so bad that it isn't worth the value of a blank floppy disk).

But wait, you say, after hard drives became affordable, there was even more use as archives that people would want to make read-only after they were initially created. That is true, except that once hard drives became affordable, the typical power user would do all work on the hard drive. Little bits of write-protect tape are trivial to remove but they are useful because they protect against accidental writing on an important disk. When your disks are only for transfer between users and for backups, there is very little chance of overwriting your critical operating system (on the hard drive), application software (on the hard drive - limited exception for some copy-protected software that required inserting a particular disk at startup) or documents (on the hard drive). Half the backup copies weren't write-protected as they were reused, hopefully on a cycle so that you were not reliant on a single backup disk if you had a hard drive failure. The only disks that still needed serious write-protection were things like your master copy of the Great American Novel or one copy of year-end financials for each year. But since you weren't writing disks all the time - all regular work being done on a much faster hard drive - you would box them up and stick them on a shelf and being out of easy reach was more write protection than the little piece of tape.

5
  • 1
    I'm not convinced by the part about the era of hard drives. The 3.5 inch floppy developed write-protection further, with a plastic slider, so it was clearly still seen as a useful feature. Interestingly, it also switched back to "open hole means read-only", allowing disks to be "permanently" placed in that mode by removing the slider completely, rather than relying on tape staying in place.
    – IMSoP
    Commented May 31 at 18:49
  • 1
    @IMSoP Hard drives didn't eliminate the need for floppies. But (at least for myself and my customers at the time and anyone who asked for my advice), it moved floppies from "working media" to "transfer and backup media". Same thing I later told everyone for CD-RW and then for USB flash drives. Commented May 31 at 20:31
  • That's not what I was saying. I was responding to the logic of your last paragraph, that write protection became less important - if that was the case, why would the designers of the 3.5 inch format go to the effort of building in the manufacturing complexity of a sliding tab?
    – IMSoP
    Commented May 31 at 20:52
  • 2
    @IMSoP Because even though it was less important, in terms of the percentage of situations where it mattered, it was still important to some degree. As far as the complexity, with a hard shell case, which is a huge benefit of the 3-1/2" media over 5-1/4" and 8", it really isn't that hard to add in the sliding piece. Commented May 31 at 20:58
  • 2
    1973 - IBM 3740 key to disk systems used single density, single sided 8 inch floppy disks, with no notch (so write enabled) en.wikipedia.org/wiki/IBM_3740 . Later CP/M systems also used 8 inch floppies, including double density, double sided 8 inch floppy disks, still no notch. The ATR8000 could use 9 x 1024 byte sectors per track resulting in 1,419,264 of disk space on a double sided, double density 8 inch floppy.
    – rcgldr
    Commented Jun 2 at 13:39
11

I always feel creepy answering questions that are obvious, but doesn't it seem that the users would always want blank floppy disks to be write-enabled by default? That is also much easier to manufacture because it only involves punching a notch along with the center hole, not punching one and then covering it with a removable tab.

Assuming that system and commercial software disks are always protected, the only time you wouldn't want user disks to be write-enabled is if you got an archived copy of something you never want changed. I don't think I ever made a 5¼ inch floppy disk read-only. And if I did, it would just be a matter of putting a piece of tape over the slot -- not cutting into the disk and if you cut too far, you destroy it permanently.

7
  • 3
    Because it takes the same effort as punching the center hole and the user can make it read only without having to cut into the disk. Commented May 31 at 11:04
  • 1
    Doesn't that just reverse the question, though? If it's obvious that write-enabled is the best default, why did 8-inch disks use the reverse? Either there is some counter-argument you haven't considered, or it isn't as obvious to everyone as it is to you, since it wasn't obvious to the people designing the 8-inch disk.
    – IMSoP
    Commented May 31 at 13:49
  • 1
    @IMSoP: Early 8" drives had no means of sensing the presence or absence of a notch, and early disks had no reason to include a notch. Though based on a recent answer, perhaps having notches be present only on media that were sold for customer rewriting might not have been a bad idea.
    – supercat
    Commented May 31 at 17:25
  • 2
    "I never made a floppy read only." After one or two accidents I adopted a policy for recovering a version from floppy: always write protect it first. Worse, a client's tape streamer was controlled by 4 buttons. 'Write' was one button, 'Read' was this one and another, pressed together. How bad is that? I posted a note on the machine saying "please write-protect a tape before restoring" and guess what? A colleague ignored it and promptly destroyed another backup. Commented May 31 at 17:30
  • 1
    Three different times, a tape backup failed to read. Usually it was the incremental backups that did it. I stopped trusting them Instead, I started backing users up to the local network and on and for important users, on a second hard drive I bought for the user's machine. Commented May 31 at 17:46
10

Assumption: It's easier to manufacture a disk without a notch.

Guess: In the era of 8" drives, the main use case for disks was to store data (after all, the OS was on tapes, or copied to fixed disks from distribution tapes). So the used the default (no notch) for read-write disks.

In the era of 5.25" drives, many computers had only floppy drives, so system software or application software was distributed on disk. And the estimation was that there would be a lot more of these kind of disks then data disks for users. So to mass produce these kind of disks, the default (no notch) was switched to read-only.

Also, it stopped the backside from single-sided disks to be used.

Of course, with piracy becoming rampant, it turned out that read-write disks were in much higher demand than read-only disks, and it also turned out that even single-sided disks had a usable backside, so both assumptions backfired...

6
  • 4
    "Also, it stopped the backside from single-sided disks to be used." For assorted values of "stopped". We used a simple hole punch to notch the back of the diskettes they had chained to the Apple II at school to put games on the back. Commented May 30 at 17:46
  • 2
    @WillHartung And I used the aforementioned nail scissors, cheaper than a hole punch :)
    – rwallace
    Commented May 30 at 17:52
  • 1
    @WillHartung I just used simple paper scissors, easier to create square notches, and easier to position than hole punchers (use second disk to mark, cut two straight lines along notch of second disk, bend, cut off). Nail scissor don't make straight lines, so I never liked them. And yes, the intention was to not be able to use the second side, as I described in the last sentence.
    – dirkt
    Commented May 30 at 19:19
  • 6
    I don't see what piracy has to do with. Programmers wanted to write frequent versions of their project to diskette. Why would the industry say "it is because pirates need it?" And it is no easier to manufacture a disk without a notch, if there is also a window in the sleeve for the r/w head. The engineering cost is the same. Commented May 30 at 19:43
  • 7
    Why would it be easier to manufacture a disc without a notch?  The plastic cover already has a large circle cut out of the centre (to expose the disk's hub) and one or more smaller circles (index holes); and the whole cover will have been cut from a large sheet of plastic (and maybe fibre inner).  Those will probably have all been done in one operation using e.g. a die to stamp out the cover.  Having an extra notch (or two, one for each side of the cover) in the die shouldn't have any noticeable impact on the manufacturing process (except in slightly increasing the waste plastic).
    – gidds
    Commented May 31 at 10:28
5

The write protect or write-enable notch on a disk doesn't physically make the media any more or less immune to being written. Instead, drives treat the presence or absence of a notch is interpreted as an indication of whether they should honor host-computer requests to write the media. If a drive doesn't pay attention to the notch, its presence or absence won't affect anything.

The first 8" drives didn't use a notch sensor, and thus manufacturers of disks would have had no reason to notch them. If drives had started requiring that notches be present in order to write disks, purchasers of blank disks would have had to modify them prior to use with such drives.

By the time 5.25" drives were invented, the concept of write protection was recognized as useful. Having a write-enable notch rather than write-protect handled three situations well: (1) disks that would be mass-produced with identical content and should never be written after that; (2) disks that should be changed sometimes, but only rarely and deliberately; (3) disks that should be easily changeable. Using a write-protect notch would tend to erase the distinction between #1 and #2, unless the sensor were located near enough to a corner that disks could be made "temporarily" read-only by removing a small piece of material to leave a notch that could be easily bridged with tape, or "permanently" read-only by removing the entire corner. I'm unaware of any system using the latter approach, though I've seen it used at a college Macintosh lab for the lab's software disks.

2
  • 2
    3.5 inch floppies handled the three-way distinction fairly well: general use disks had a plastic slider to switch write protection on or off, but were read-only when the hole was uncovered, so disks could be made "permanently" read-only by snapping out the slider, or manufacturing a disk without it. (It wasn't really permanent, because tape could be used to cover the hole, but it was explicitly distinct from moving the slider.)
    – IMSoP
    Commented May 31 at 18:46
  • 1
    @IMSoP: Using a band-saw to remove the corner of 3.5" disks that people weren't supposed to monkey with is an approach I've seen used, even though the disks weren't designed for it.
    – supercat
    Commented May 31 at 18:55
2

The 5.25" floppy disk was essentially a smaller, cost-optimized version of the older 8" floppy design. Inverting the use of the write-protection notch would also allow the disk to be automatically seen as "write-protected" if inserted upside down. Making one feature serve two purposes would have been a nice optimization of the manufacturing process.

Commodore 1541 was a single-sided disk drive. But the optimization of the 5.25" disk manufacturing process made double-sided disks the norm, as having one side of the disk not have the magnetic material would have been more complicated than simply coating both sides the same way.

So, as a result, users of Commodore 1541 and other single-sided disk drives wanted to minimize storage media costs by using the other side of their disks too, by the simple expedient of cutting a second write-enable notch to the opposite side of the disk and then inserting the disk into the drive upside down.

This was possible because the drive mechanism used in the 1541 (and other single-sided drives of that era?) did not actually use the sector-0-indicator/speed-sensor hole, and apart from that and the write-enable notch, the design of a 5.25" floppy was entirely symmetrical, so it was perfectly possible to insert a disk upside down. But without the second write-enable notch, the drive would have treated the flipped floppy as write-protected.

Users of real double-sided disk drives would not have needed to cut the extra notches, as the standard single notch would apply to the whole disk.

Since turning the disk upside down effectively reverses the direction the disk rotates in, a double-sided disk written in a double-sided drive would have been unreadable if inserted upside down. Likewise, a disk that had a second write-enable notch cut into it and then had its second side written in a single-sided drive would still need to be turned over manually in a double-sided drive, as the data of the second side was written with the disk rotating the opposite way from what a double-sided drive would expect.

By the time 3.5" disks became the norm, 3.5" disk drives were essentially always double-sided, and so there was no widespread need to defeat the protection against upside-down insertion inherent in the 3.5" disk design (the notched corner + the slider protecting the read/write hole + the centre hub of the disk accessible from one side of the disk only).

3
  • This answer implies that Commodore's 1541 implementation drove the industry standard.
    – dave
    Commented Jun 1 at 14:34
  • 1
    Rather the opposite: first the industry developed the cost-reduced 5.25" version of the earlier 8" floppies, with the notch optimized to serve two purposes: write protection and protection against data loss when inserted upside down. Then Commodore developed a widely-sold single-sided drive that was capable of using the flip side too, causing people to start notching their disks for flipped use.
    – telcoM
    Commented Jun 1 at 14:50
  • Edited to clarify.
    – telcoM
    Commented Jun 1 at 22:56
1

The first 8 inch floppy disks were used to boot IBM mainframes. In 1973, the IBM 3740 data entry system used single sided, single density 8 inch floppy disks, no notch, write enabled.

https://en.wikipedia.org/wiki/IBM_3740

Early CP/M systems also used 8 inch floppy disks, single or double density, single or double sided, still no notch, write enabled. I had CP/M systems that used 8 inch floppy disks, the Pertec PCC-2000, which had two half height 8 inch floppy drives, and could store 1.2 MB on a double sided double density floppy disk. I wrote utilties for the PCC-2000 that could read | write IBM 3740 disks (this involved translating beween EBCDIC and ASCII, and adding | removing ASCII return and line feed characters).

https://en.wikipedia.org/wiki/Pertec#Pertec_and_MITS

Missing from this description is an optional 10 MB 8 inch hard drive which later became standard for it. I worked at Pertec so I got one of these cheap.

http://www.1000bit.it/ad/bro/pertec/pertecpcc200.pdf

I also had an ATR8000 with the 8088 add on card, so it could run either CP/M 2.2 or MSDOS 2.11. It could use 9 x 1024 byte sectors per track, so a double sided, 77 track floppy would hold 1.4 MB (1,419,264) bytes of data. The ATR 8000 could work with 8 inch and 5.25 inch floppy drives.

https://www.atarimagazines.com/v3n4/ATR8000.html

Images of 8 inch floppy disks, none of them have a notch.

https://rcgldr.net/misc/flop8ss.jpg

https://rcgldr.net/misc/flop8sd.jpg

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .