30

When I began high school in 1987 all the computers in the lab were Apple IIe desktop computers. A question has haunted me since then:

if you inserted a floppy disk "upside down" the Apple IIe would make a hideous noise like a vacuum cleaner. Why? What a shocking design; it scared the heck out of me as a newb. Was it usual to do that to a user? I never really recovered from that; I was put off computers for the rest of high school.

Why design a computer like that for newbs? I'd never seen a computer before. Even worse, the noise would not end. It would go on for minutes. Who on earth decided to make it like that? I'd fire the design team.

hippietrail
  • 6,646
  • 2
  • 21
  • 60
Snack_Food_Termite
  • 551
  • 4
  • 9
  • 1
    I don't recall such a noise - and we had a special notcher-tool for adding a second notch specifically to flip the disk and use the other side for data storage. Were you putting the disks in with the read window in the wrong place, ie, sideways? – Criggie Jun 27 '19 at 20:45
  • 1
    Floppy disks can be used either way up, you just needed to cut another notch to write to the other side. The magnetic material on the back was supposedly sometimes not as good quality as that on the front but I never had an issue with it. So this doesn't really make sense. Are you sure you're not thinking of 3.5" disks? – Jonathan Potter Jun 27 '19 at 21:24
  • 2
    @JonathanPotter unlikely - the apple ][ could use a 3.5" disk but it was massively rare, and only arrived after the macintosh release. So for ~all users it was 5.25" floppies storing 113 kbytes or 140 kbytes/side. – Criggie Jun 27 '19 at 23:04
  • Even if it was a 3.5" disk, these were by design impossible to fully insert upside down. There's a corner cut off on one side that would prevent you from fully inserting it in the wrong orientation. – Darrel Hoffman Jun 28 '19 at 17:14
  • 1
    After inserting the disk, how did you access it? If you were at a DOS prompt and then typed "CATALOG" or similar, the noise was DOS recalibrating the drive after getting repeated read errors on a presumably unformatted disk. In that case, it's not specific to being upside down or not, but readable by DOS or not. – peter ferrie Jun 28 '19 at 17:46
  • @Johathan Potter Was it really a good idea to write magnetically on the other side of a thin material? Doesn't that risk disturbing the magnetic domains on the opposite side? I never heard of the practice at the time; never tried it myself. – Kaz Jun 28 '19 at 18:47
  • 1
    It wasn’t that easy to put a disk in upside down... the top was smooth and had a label on it, the bottom had ridges from the fold over and sonic welding... I mean you'd really have to not be paying attention. – Harper - Reinstate Monica Jun 28 '19 at 22:49
  • @Kaz: The flip side is still further away than adjacent tracks. – Nick Westgate Jul 01 '19 at 02:54
  • @NickWestgate Say what? We are talking about 5.25" disks, formatted with just 35 tracks.I suspect that even the quarter-track increments of the stepper motor are still wider than the thickness of the floppy mylar. – Kaz Jul 01 '19 at 05:35
  • @Kaz: Hmm, yes, you're right. The erase gap between tracks is at least 0.15mm, and the thickness of the mylar is only 0.07mm plus the oxide coating. Maybe it's the head geometry and the fact it touches the media. – Nick Westgate Jul 02 '19 at 11:17

5 Answers5

115

While some floppy drives have a sensor to determine when the head is at the outermost position, the Apple uses four approaches:

  1. On startup, it blindly tries to move the head outward about 40 tracks. When the motor hits the end stop it will be unable to move further and will consequently stop. Once the Apple has done this, it will assume the head is at the outermost track.

  2. When formatting a disk, the Apple uses the same procedure to force the head to the outermost track.

  3. The Apple will keep track of how it has moved the head in and out since it forced it to the end stop, and will thus generally know where it expects the head to be. Because each track of a formatted disk includes markers that identify it, the Apple will generally assume that if it thinks it's on e.g. track 12, and it sees a marker for track 12, then it is in fact on track 12 and it can proceed on that basis.

  4. If the Apple is unable to see markers that indicate that it's where it thinks it should be, it will force the head to the outermost track, then step it by the proper number of steps to reach the proper position, and look again.

The noise you heard was from step 4. When trying to access a valid formatted disk, the Apple will generally not need to run the head against the end stop, but the back side of most disks won't typically be validly formatted.

supercat
  • 35,993
  • 3
  • 63
  • 159
  • 7
    Note that this isn't unique to Apple. I have yet to encounter a drive that didn't behave this way, it's just more modern drives are quieter about it and the stepper motors move the heads faster so it doesn't last as long. It even applies to hard disks but you won't encounter it unless you have a dying drive. – Loren Pechtel Jun 28 '19 at 03:25
  • 5
    @LorenPechtel: The Commodore 1571 drive has an optical sensor, as does the IBM PC; I think 3.5" drives have also had sensors from the beginning. – supercat Jun 28 '19 at 04:35
  • 2
    Yes, this is the correct reason for the noise. I owned an Apple //e as a child (and still have it today!) and encounter this noise whenever I try to use a corrupted disk - the issue is that the drive is trying to recalibrate itself and gets really noisy while doing so. – TheHans255 Jun 28 '19 at 16:16
  • @Mattman944 ...or never hear again – A C Jun 30 '19 at 07:05
11

If you inserted the disk "upside down" the Apple //e would make a hideous noise like a vacuum cleaner.

Not sure what the 'vacuumer sound' should be. could it be that you're refering to the way the drive returns - and calibrates - to track. I would rather describe it as a fast clicking noise over some scratching.

Also, there was no different operation between "upside down" or not. Simply as the drive had no way to detect which was the'right' side (*1). In either case the II tried to read the disk. The difference may eventually have been that there was nothing to read on the back side (of the disks you used), resulting in an endless attempt to read.

There was effectively no room in the 256 byte ROM of the disk controller for a more sophisticated error recovery than to try again and again and again...

Why? To me what a shocking design;

Because it's an awesome simple design. The head ist moved across the racks by a spiraled disk. Turned in one direction the head moves outward toward track 35, when reversed it turns toward track 0. After power up/reset the head may be positioned at any of the tracks. So it needs to move in a way to get to a defined point - with track 0 being a desirable goal.

There is no way to tell on what track it is, so the computer moves the head back toward track 0. Now, they could have installed a sensor telling when track 0 is reached. Except, Woz tried to minimize parts and logic, so he devised a way to do it without a sensor and electronics. So the spiral disk was made in a way that the head simply hits a stop were track 0 is supposed to be. Now the controller can always turn the head back for full 35 tracks. It'll move until it hits the stop. after that the disk continues to move (until all 25 steps are done), but the slider moving the head just jumps the spiral track, as the head can't move anymore.

Splendid solution to solve an enginering problem without any logic, sensors and whatsoever, just a groove in a disk.

it scared the heck out of me as a newb. Was that usual to do that to a user?

Jup.

I never really recovered from that; I was put off computers for the rest of high school.

If it got you off that easy, I guess there where other things for you to enjoy.

Why design a computer like that for newbs?

The computer wasn't designed for 'newbs' at all, it was designed as a computer for general use. A quite cost efficient design, which BTW saved Apple several times from otherwise disastrous failures.

*1 - A fact that many owners used to double their disk capacity - either by adding a second write notch at the opposite side or by adding a Protected/Standard/Unprotected switch to their floppies.

Raffzahn
  • 222,541
  • 22
  • 631
  • 918
  • Well it was thrown at newbs! High school kids who had never seen a computer! That was NOT a clicking sound. It was exactly as I described: a prolonged sound like a vacuum cleaner. That occurred with every //e in the lab if you put the disk in upside down. So it wasnt one bad or rogue Apple //e; it was design. I mean it when I say that I never recovered from that; I was left with the Pavlovian reaction of computer = a monster and I get punished for putting a disk in wrongly. I went on to fail senior maths; I was interviewed for technophobia. I had never touched a computer for a stats course. – Snack_Food_Termite Jun 26 '19 at 23:06
  • 4
    @athornton My childhood computer would scream at me very loudly for several minutes whenever I turned it on (something wrong with the Del key, I assume) and would then proceed to try to erase all of my files. It terrified me. This wasn't enough to stop me messing around with it. YMMV. – wizzwizz4 Jun 26 '19 at 23:12
  • Not so much splendid as a disgusting hack. $595 retail for the drive and controller, and they couldn't afford a $2 microswitch? Those were the days! – Bruce Abbott Jun 27 '19 at 00:19
  • 5
    @BruceAbbott, a $2 microswitch, and a pin to read it, and circuit-board traces (or wires) connecting the two, and code for reading and interpreting it, and ROM space to store the code. A limit switch would have cost significantly more than the bare price of the switch. – Mark Jun 27 '19 at 00:52
  • @athornton: It's still not completely clear if you're talking about normal boot and track reading sounds, which are in this video or the recalibration sound which occurs 12 seconds into this video. – Nick Westgate Jun 27 '19 at 00:58
  • 3
    I dunno... To me those Apple II's sounded like the Greek Sirens lulling me astray. I never came back. – Geo... Jun 27 '19 at 01:02
  • It was neither sound. I never heard that beeping noise ever. It was just like I said; the noise was that loud that you had to shout. It was about as loud as the dot printers of the time. It only happened when you inserted a floppy disk the wrong way up. – Snack_Food_Termite Jun 27 '19 at 01:03
  • The youtube comments below the second video speak of I/O errors and a loud "machine gun" noise thats is loud! See! It is NOT just me! There is other testimony that these drives did operate like this! – Snack_Food_Termite Jun 27 '19 at 01:13
  • 2
    This answer gives a fine explanation of why the drive makes loud noises, but it doesn't seem to explain why it makes a different noise when the disk is inserted upside-down. – Kaz Jun 27 '19 at 07:08
  • 4
    @athornton "Well it was thrown at newbs! High school kids who had never seen a computer!" Then that's the school's fault for not telling you to not put disks in upside down. "That occurred with every //e in the lab if you put the disk in upside down" That's your fault for not thinking "Hey, let's not put the disk in upside-down next time". – TripeHound Jun 27 '19 at 11:10
  • 2
    @Kaz There is no different noise when inserting upside down, it just never stops, as the backside is usually not formated. – Raffzahn Jun 27 '19 at 11:53
  • 1
    @BruceAbbott Sure, one can make a complicated by the books solution to solve a problem with active components - or just make a self contained without any active controlled parts, simplified and without many follow up problems of a more complex solution. It's like with plastic wast: Not using a second layer of packaging is way better than recycling. isn't it? – Raffzahn Jun 27 '19 at 11:57
  • There's proper engineering, and then there's "lets just hammer it into the backstop 35 times and hope nothing breaks!". It was all about cutting costs to the bone to make obscene profits. Too bad if the drive goes out of calibration after a while, or accidentally plugging the non-keyed connector in backward erases all your disks (been there, done that!) https://en.wikipedia.org/wiki/Disk_II#Disk_II – Bruce Abbott Jun 27 '19 at 23:54
  • 1
    @BruceAbbott Awww come on, it is propper engeneering, as it wasn't hammered into the stop, but made to skip at that point - exactly the same way records worked (remember them?) - in fact, my guess is that's where Woz got the idea from - neither unreliable nor damaging - just unpleasant to listen to. And for the damage and deajustment part, Never had that issues with a Disk II in more than 7 years of daily usage - and many more years since then as part of the hobby. In all the years there was only real fail ever - the offending disk is still nailed to a shelf :) Ofc, YMMV. – Raffzahn Jun 28 '19 at 00:18
  • 2
    @BruceAbbott: Replace "hope nothing breaks" with "ensure that things are strong enough to survive that" and I see no problem. Using a slightly-elastic backstop might have been a good idea since the critical calibration is the attachment between the stepper and the drive head. The backstop shouldn't be precisely at the first-track position, but a bit below it. If software's written properly, it should have no trouble tolerating almost a track worth of slop on the back-stop position provided it's fully below the first "real" track. – supercat Jun 28 '19 at 16:59
11

The obvious answer is that it wasn't designed 'for newbs' but as a general purpose personal computer that would find its way into many homes and businesses as well as educational establishments. Even in the latter, they also had to be suitable for running some fairly advanced software as well as for introductory courses.

By the standards of the 70s, when serious computing was generally the realm of the scientifically literate, the Apple II was pretty user friendly though. It actually had a monitor and keyboard, you didn't need to flip switches and look at lights to boot it up, and programming errors gave fairly comprehensible error messages.

It also had to be built to a budget though, so they couldn't design it to prevent every instance of user error from causing problems without the cost ballooning. Apple computers were already relatively expensive for their time and they'd just have priced themselves out of the market if they'd tried. So yes, they probably knew what would happen if you put the disk in upside down, but since it wouldn't break anything permanently it wasn't seen as worth the cost of fixing.

For what it's worth, you could swap in pretty much any other computer from the 70s and 80s and there'd be plenty of other issues where the clumsy interactions of novice users could cause bad things to happen with them.

Matthew Barber
  • 1,519
  • 8
  • 9
11

The short answer is they did it to save money. They could have made the drive behave better when a disk is inserted upside down, but that would have required larger ROM chips to store the firmware or more expensive drives and controllers to detect the problem.

As long as it didn't damage the computer such behaviours were considered acceptable to get the price down.

user
  • 15,213
  • 3
  • 35
  • 69
  • 17
    That's simply wrong, as the drive didn't behave different according to which side was 'up' – Raffzahn Jun 27 '19 at 11:59
  • 5
    @Raffzahn what does that have to do with anything? I'm saying that any less-than-optimal behaviour was mostly just due to it costing more money to implement better behaviour, like detecting when the disk is in upside down or a better abort procedure so it doesn't keep grinding away for minutes on end. – user Jun 27 '19 at 14:08
  • 7
    @Raffzahn That's the point. – StackOverthrow Jun 27 '19 at 19:19
  • 8
    This answer is still wrong. The drive made the same noise for an unformatted disk. The backside of a disk is unformatted (unless you punch a hole to disable write protection and format it, which basically all Apple II users did to save money). The drive has no way to know if the disk is upside down or not. The drive has no way to know if the disk is unformatted or not. They couldn't have made the drive "behave better", no matter how more expensive it was. The software that made the noise was actually in RAM, and not in ROM (unless you booted, when it always would make that noise). – dirkt Jun 28 '19 at 05:50
  • 3
    They did what to save money? They added a noise? Surely that would have been more expensive? – EML Jun 28 '19 at 07:37
  • @EML they used the cheapest drive they could find, and the minimum amount of code to make it work to save on ROM space. They omitted orientation sensors. – user Jun 28 '19 at 08:06
  • 5
    Repeat after me: There are no orientation sensors for disks. Because 5.25" floppy disks are symmetrical, there's no way to know which side is up when you put them into a drive. They could have added a sensor for track 0, which they omitted, so the way to recalibrate is to move the read-write head to lower tracks for like 80 tracks, and when it bumps into a mechanical stopper, it makes that noise. – dirkt Jun 28 '19 at 11:02
  • 1
    And please do read up on where DOS 3.3, ProDOS, or any other OS in the Apple II resides (hint: It's not in the ROM). – dirkt Jun 28 '19 at 11:04
  • @dirkt the index hole can be used to determine orientation for single sided disks. Using the index marker is optional though. – user Jun 28 '19 at 11:35
  • And in fact all floppy disks for the Apple (and many other micros) are soft-sectored, so not only is the index marker optional, it's completely useless. – dirkt Jun 28 '19 at 11:44
  • 1
    @dirkt except for determining which way up the disk has been inserted – user Jun 28 '19 at 11:45
  • 1
    While in theory you probably could have added an index sensor and waited one whole disk revolution and then declare "I can't see an index hole, so the disk must be upside-down", now that would have been totally ridiculous. (And IIRC there were also floppies without an index hole in the first place; those wouldn't have worked at all in that drive, which would have been bad.). And it wouldn't have been hard to include a "real" upside-down sensor if they wanted one: Just use a notch like the write protect notch in an asymmetrical position. But such a thing doesn't exist. – dirkt Jun 28 '19 at 11:54
  • @dirkt okay, okay, it's just one small aspect of the answer, don't get too hung up on it. I imagine ROM space was the primary reason. – user Jun 28 '19 at 12:37
  • @user But the code that did this was stored in RAM, loaded off a floppy. DOS directly handled floppy reads and writes, iirc. – wizzwizz4 Jun 28 '19 at 19:18
  • @wizzwizz4: I pointed this out to "user" about 3 times, but he still insists it is in ROM. I decided to give up. It's like talking to a wall. – dirkt Jun 29 '19 at 06:19
  • 1
    It's not wrong. – user Jun 29 '19 at 19:18
3

This was the sound it would make anytime it tried to access an unformatted or unreadable disk, as it continuously tried and retried reading it.

Because the Apple Disk II only used one side of a floppy disk, the unused side would usually be unformatted.

Kaz
  • 8,086
  • 2
  • 38
  • 81
Tim
  • 37
  • 2