back to article Bill Gates: Yes, Ctrl-Alt-Del salute was a MISTAKE

Microsoft supremo Bill Gates has claimed that the ctrl-alt-del keystroke - once a way of admitting defeat in the face of crashing software - was a mistake all along. Anyone with a passing knowledge of PCs will remember hammering those three keys to forcefully reboot the computer as code locked up. Some people even called the …

COMMENTS

This topic is closed for new posts.

Page:

  1. HCV

    WAT

    "Microsoft. We're so confident that you'll need to reboot, we want a DEDICATED KEY for it."

    We Unix people laughed and laughed when one of the most vaunted features of Windows 2000 was that it would go *so long* without rebooting -- like *30 days* -- that they built in a special "reboot at a regular interval" feature.

    Which, of course, we would have called "cron". If we thought that it was a good idea in the first place.

    1. asdf

      Re: WAT

      Yes a nearly crash proof OS was 1970s technology (sure someone will point out even earlier probably) but Billy still can buy any yacht he wants.

      1. Hud Dunlap
        Joke

        @ASDF

        But can he win the America's cup?

        http://www.americascup.com/en/news/3/news/18450/ellison-this-regatta-has-changed-sailing-forever

    2. Anonymous Coward
      Anonymous Coward

      Re: WAT

      Yes, just not everybody could afford a Unix machine then, and its very expensive applications, and most Unix users worked on system they didn't pay for, but their universities and companies did. Thus most people were happy with cheaper machines although sometimes you needed a reboot....

      1. ThomH

        Re: WAT

        MS-DOS and early versions of Windows were designed for a processor without protection domains. You couldn't do anything to protect the OS against a crashing process. In Windows you not only couldn't protect the OS but you had to rely on third-party hardware providers writing good drivers.

        I guess to the extent that Microsoft deserve criticism, it's not transitioning their OS fast enough as the x86 architecture matured. With computer sales growing exponentially throughout the 80s and early 90s that made legacy software compatibility much more of a problem than it needed to be. But probably they genuinely believed OS/2 would happen.

        1. Anonymous Coward
          Flame

          Re: to the extent that Microsoft deserve criticism?

          There is no "extent." There is no limit to the criticism that Microsoft deserves, especially for those dreadful "Early" versions (say up to W2K) each of which it expected us to pay to escape.

          1. Tinker Tailor Soldier
            Facepalm

            Re: to the extent that Microsoft deserve criticism?

            Well, at least MS had the excuse that the processors it ran on 8086 and 8088 didn't have mem protection. Apple COULD have wired in protection for the 68000 (Mac onward) but elected not to due to cost. OS/2 used mem protection from the 286 onwards. NT, based on the 386 did too.

            The consumer OS's didn't. Because:

            1. RAM was expensive, and the memory protected OS's (OS/2 and NT) both needed more of it.

            2. Many apps didn't respect process boundaries and crashed anyway, e.g. directly accessed interrupt tables to hook stuff. So, to keep compatible you couldn't protect memory.

            But, rewrite history all you like?

            1. ThomH

              Re: to the extent that Microsoft deserve criticism? (@Tinker Tailor Soldier)

              I'm unclear who you think is rewriting history — essentially both the PC and the Mac came from companies that understood protected memory and MMUs perfectly but choose to omit the hardware for cost purposes. There was an MMU in the Lisa, there wasn't in the Mac. There was one in any number of IBM machines going back decades, there wasn't in the PC.

              The history of Apple's multiple subsequent internal OS development screw-ups is interesting but quite distinct from Microsoft's errors. The stories start similarly but Apple get to the point between 6 and 7 where the resources were such that they could have afforded just to run multiple instances of 6 simultaneously and preemptively to multitasking properly with memory protection on those devices with MMUs, but instead they double down on cooperative multitasking and spend the newly spare resources on rewriting a bunch of things in C. After the PowerPC move they have a full preemptive, protected memory handling nano-kernel which is used to run the existing OS. For quite a while large parts of the stack remained in 68000 code and just ran through an emulator — they spent engineering time getting the emulator down small enough to fit entirely within the processor cache because it was a more effective way to transition than dealing with the OS proper.

            2. Tom 13

              Re: But, rewrite history all you like?

              I wouldn't say rewriting. That's your interpretation.

              As far as your point #2 goes, Apps didn't respect the boundary because they didn't HAVE to. If they'd been forced to respect it, we'd probably all be better off now.

              And yes, I recall when memory was super-expensive. I also recall being sorely disappointed when having finally plopped down a huge wad so I could buy enough to have a proper RAM disk to speed up my system, it turned out the VESA local bus disk controller I had already installed made its addition barely noticeable.

          2. Hans 1

            Re: to the extent that Microsoft deserve criticism?

            @Thad

            Why just the dreadful early versions ? True, w2k was good - by MS standards, that is - but anything before that was crap and anything after it was bloatware, with the notable exceptions of Windows Me, Vista and 8, which are crap (latter two being both bloated crap).

          3. Anonymous Coward
            Anonymous Coward

            Re: to the extent that Microsoft deserve criticism?

            @Thad - I think you're re-writing history there - Windows was the software that allowed people to break the stranglehold of Big Iron Unix on the datacentre. At the time there was simply no need to use Unix for many tasks, it was over engineered, but it took Windows (and to a lesser extend OS/2 and Netware) to break into the datacentre and do those commodity tasks at commodity prices.

        2. tom dial Silver badge

          Re: WAT

          1. Early Unix machines also lacked protected memory, as did early versions of Minix. Both were more stable than MS-DOS before 6.2.

          2. OS/2 did happen, in a technical sense at least, and it ran Windows applications better and more reliably than Windows did.

          1. Neil 44

            Re: WAT

            The Interdata 7/32 was a very early Unix platform (the first non-AT&T according to K&R Ed 1!)

            It certainly had memory protection and used user and system level interrupts to process things. Processes were isolated from each other.

            (admission time: there weren't any 7/32s around when I joined the company, but there were the slightly newer 8/32's!)

        3. Tom 13

          @ThomH: You were doing ok

          until that very last statement. Their dealings with IBM on OS/2 even more than their dealings with Lotus 1-2-3 forever cemented the "Microshaft" nickname.

          1. ThomH
            Unhappy

            Re: @ThomH: You were doing ok (@Tom 13)

            I buy the version where the surprise success of Windows gave Microsoft the idea to renege — they stabbed their partner in the back as soon as an opportunity arose but had not expected or been planning for the opportunity.

            I can't think of another reason why they'd create the multitasking, new executable DOS 4, barely license it and then push all its code off into OS/2, subsequently picking up DOS from the version 3 code base.

        4. Anonymous Coward
          Anonymous Coward

          Re: WAT

          Many computers were like that. They added MMUs and all sorts of things later.

          But adding protection and things to stop the machine crashing almost parallels safety protections on cars. It just resulted in worse coding just like a safer car results in worse driving.

          Of course now we're glad of mem protection because of malware and keeping that in check.

    3. Anonymous Coward
      Anonymous Coward

      Re: WAT

      Windows NT4 was quite capable of going 30 days without reboot, w2k could go far longer. I am not aware of any "reboot at a regular interval feature" unless someone created their own scheduler job. It's certainly not built in functionality.

      At around the time of NT4, I had a pretty good Sun Sparc UNIX workstation on my desk, it cost around £10k, at the company I worked for at the time we could get fully licensed NT4 Servers running on Compaq proliant hardware for half that price. They were good servers, too. Consequently UNIX servers started their inexorable departure from the datacentre. Proprietary RISC Unix doesn't deliver the bang for buck of either Windows or Linux on Intel. What's the point of being able to stay up for years on end, if no-one can afford it and the alternatives are good enough, particularly with modern clustering.

      1. Carl

        Re: WAT

        I've seen many UNIX/Linux machines over my career with uptimes of 1000+ days.

        I even ran a mailserver on an old 386/FreeBSD that had 200+ days of uptime until a power snafu happened.

        But my favourite "uptime" story is the (possibly apocryphal) one about the Novell Netware box that nobody could find. It worked, responded to pings etc, but nobody had any clue where exactly it was. It was finally discovered, amongst other "old" kit behind a wall that contractors had put in about 5 years earlier.

    4. BillG
      Happy

      Re: WAT

      30 days? C'mon, where'd you get that? I had a Windows 2000 laptop that ran for 14 months before the battery died during a winter power outage.

      1. Anonymous Coward
        Anonymous Coward

        Re: WAT

        Options:

        (a) you are full of sh1t

        (b) you never did any kind of update

        (c) you never ran any software or

        (d) you never plugged it into a network

        My money is on all 4

    5. 9Rune5

      Re: WAT

      "Microsoft. We're so confident that you'll need to reboot, we want a DEDICATED KEY for it."

      Uhm, no.

      BillG talked about a dedicated key to produce the _login_ screen. I.e. Windows NT, not 16-bit Windows or DOS.

    6. DrXym

      Re: WAT

      Even systems like Minix, AmigaOS (on the 68000) and MacOS (68000) implemented a 3 fingered salute for those when a rogue process stomped on the kernel and rendered the whole machine unresponsive.

      The issue is not the OS but the CPU's lack of memory protection.

    7. Anonymous Coward
      Anonymous Coward

      Reboot Jockeys

      Where I worked we called the windows "sys admins" reboot jockeys. A name that they fully deserved. It was not uncommon, at the end of the day, to see them drop a windows machine from the load balancer, reboot and then put it back in again and then work their way through all of them like this in the hope that they would last overnight. The idea of investigating why something went wrong never entered their head.

      A dozen years later and things have improved significantly. Most of them now look to perform a reboot as a last resort rather than a first and they will actually investigate the cause when something goes wrong.

  2. HCV

    "Oops. Did hitting that mess something up for you?"

    And I would like to add that asking for a single dedicated "BORK YOUR COMPUTER" key on a computer keyboard may be the single stupidest thing I have ever heard a multizillionaire brag about.

    1. Anonymous Coward
      Anonymous Coward

      Re: "Oops. Did hitting that mess something up for you?"

      There's already such a key and it is the PC power switch! (I know Ctrl+Alt+Del performs a soft reboot...)

      Anyway a single key to reboot a PC would have been really very, very dangerous - and silly. The IBM guy did a good work selecting a combination you can't use by mistake.

      1. Tom 7

        Re: "Oops. Did hitting that mess something up for you?"

        would have been really very, very dangerous ???? You mean like that key that you hit and it ignores everything else until it goes to sleep - ten seconds after you hit the power switch so when you power up again whatever was running needs a re-install.

        A Windows keyboard - smells less than pouring your beer into the fan but not as nice.

        1. M Gale

          Re: "Oops. Did hitting that mess something up for you?"

          A Windows keyboard - smells less than pouring your beer into the fan but not as nice.

          Power buttons on the keyboard is something Apple Macs had first. It's just as fucking annoying on them, too. Think different? Yeah, right.

          Amusingly, the keyboard I'm using thankfully has no such retardedness on it. Model? Microsoft Wireless Keyboard 800.

      2. Chad H.

        Re: "Oops. Did hitting that mess something up for you?"

        Reset button was usually right there on the chasis.

        1. Caesarius
          Facepalm

          Re: "Oops. Did hitting that mess something up for you?"

          Reset button was usually right there on the chassis.

          I remember a MAC with the reset button at "mouse height". The number of people in my lab who accidentally ran their mouse into it. Disbelief, cold sweat, realisation, anger, and finally the search for something like a rubber (1) to sit in front of it.

          (1) No, John Doe: that's a pencil eraser

      3. Anonymous Coward
        Anonymous Coward

        Re: the PC power switch!

        I've seen a server downed by a foolish knee movement, but generally, one's finger wandering doesn't extend off the keyboard to the case itself. Not unless one is really kinky.

        1. Nigel 11

          Re: the PC power switch!

          In the early days of PCs, the power switch sometimes protruded proud of the case. Back then it was a hard mains switch, not a soft switch. I've seen major damage caused by pushing the keyboard back against the idiot-designed case. Later, the power switch (and reset button) were always recessed. Ease-of-use is not always a good thing. Big red switches ought to come with Mollyguards. Needing three fingers for C/A/D was definitely good design not bad. It's pretty much impossible to C/A/D by any kind of mistake.

          1. Daniel B.

            The PC KEYBOARD power switch

            Soooo.... Nobody here remembers the retarded keyboards from the mid-00's with the awful "shutdown" key? Especially the ones that had a row of "shutdown, suspend, sleep" keys right below the "insert, delete, hone/end" keys. So you could be trying to press End, but press the cursed shutdown key, which would send an unstoppable shutdown command to Windows! Weeeeee!!

            I usually had an unsaved Notepad session somewhere, so that I could click Cancel on the save dialog to stop the process. Though if you took too long, all the other apps would have been killed by the OS anyway... :(

            1. Rattus Rattus

              Re: The PC KEYBOARD power switch

              Oh gods, yes. I had one of those fucking things. Most ridiculous idea for a keyboard. I ended up removing those three keys physically after one unintended power-cycle too many.

    2. Mat Child

      Re: "Oops. Did hitting that mess something up for you?"

      The BBC micro's BREAK key basically was what Gates was asking for.

      Although you could intercept the call and stop it from performing the reset.

      1. h3

        Re: "Oops. Did hitting that mess something up for you?"

        It didn't work on its own though only as a key combination. (With either shift or control).

      2. Charles Manning

        Re: "Oops. Did hitting that mess something up for you?"

        IIRC, BREAK was more like CtrlC than the CtlAltDel

        1. Yet Another Commentard

          Re: "Oops. Did hitting that mess something up for you?"

          Esc stopped the programme (error 17, easy to stop) Break did a "soft" reset, ctrl+break did a "hard" reset IIRC.

          The Amiga had a similar one I think - was it ctrl+amiga+amiga?

      3. HippyFreetard

        *KEY 10 "OLD|MRUN|M"

        Can't believe I remember that - was one of the ways to annoy BASIC programmers. Ctrl+Break did the trick :)

        1. Anonymous Coward
          Anonymous Coward

          Re: *KEY 10 "OLD|MRUN|M"

          Of course the BBC Master had a "Break Lock" which could physically prevent you from pressing the Break key.

    3. NullReference Exception

      Re: "Oops. Did hitting that mess something up for you?"

      The original Apple ][ and ][+ had a "RESET" key on the top right of the keyboard, right above the Return (i.e. Enter) key. It was very easy to hit it by mistake and lose all your work. Many users would make it harder to hit RESET by putting rubber washers under the keycap or using various other tricks. Eventually, someone at Apple realized that single-key RESET was NOT a good idea, and from the Apple //e onwards the design was changed so you had to press Ctrl+RESET to do a reset.

      Bill & Co used to write stuff for the Apple... guess he forgot about this!

    4. deshepherd

      Re: "Oops. Did hitting that mess something up for you?"

      And I would like to add that asking for a single dedicated "BORK YOUR COMPUTER" key on a computer keyboard may be the single stupidest thing I have ever heard a multizillionaire brag about.

      Around the time MS-DOS was appearing there were computers with single key "resets" - e.g. the BBC Micro had the "Break" button which was "conveniently" sitting at the top right of the keyboard right next to keys that you would use in normal usage. This meant it was quite possible to press it by mistake .... think as a result a market developed for plastic covers which sat over the break key preventing it being pressed.

      I always though that having to press 3 keys (and 3 keys that needed both hands on most keyboards) was an eminently sensible idea to make it clear you were really meaning to reset the PC. Later when C-A-D became the way to bring up the login menu in later versions of windows was probaly more problematic as (a) you wanted to login in - what else where you going to do at that point and (b) using a key combination that was (at least subconciously) associated with "help, my PC has gone wrong" with an action that said "I've just turned my PC on, lets get started" seemed counter intuitive.

  3. Joe User
    Megaphone

    Ignore the obvious choice

    "Guy who did IBM keyboard design didn't want to give us our single button."

    Was there something wrong with using the SysRq (System Request) key? It was the obvious choice, and that key has been around since the PC/AT days.

    1. Don Jefe

      Re: Ignore the obvious choice

      I think Gates wanted a single, dedicated, button because it would make the keyboards different. Great branding strategy, kind of dumb from systems design standpoint though. Computers, by default, are the great reuse kings of everything. Building single use, most anything, into a computer is kind of dumb. It goes against everything a computer represents.

    2. jaywin

      Re: Ignore the obvious choice

      That's used under Windows though.

      Pause/Break on the other hand... "Hello, IT" "My computer has broken / paused. Fix it" "Just push the Pause / Break button" "Many thanks good sir"

      1. Tannin

        Re: Ignore the obvious choice

        Joe User : Was there something wrong with using the SysRq (System Request) key? It was the obvious choice, and that key has been around since the PC/AT days.

        jaywin: That's used under Windows though.

        This was 1981. Windows did not exist. DOS ruled. The dreadful first release of the Windows operating environment (i.e., a type of DOS application) came along just before 1986, and it did not become truly usable and widely popular until v 3.1 in 1992 - 11 years after the standard IBM keyboard was first manufactured.

        1. John Smith 19 Gold badge
          Unhappy

          Re: Ignore the obvious choice

          "The dreadful first release of the Windows operating environment (i.e., a type of DOS application) came along just before 1986, and it did not become truly usable and widely popular until v 3.1 in 1992 - 11 years after the standard IBM keyboard was first manufactured."

          And still being co-operativemulti-tasking, like the Archimedes IIRC, but without MS being very honest about it one rogue application could still hang your computer.

          Thanks Microsoft.

    3. Don Jefe
      Joke

      Re: Ignore the obvious choice

      Or used the PrntScrn button instead. They could have made some big time partnerships with ink manufacturers at the same time.

    4. HCV

      Re: Ignore the obvious choice

      Er, yes, but the PC/AT wasn't around in the PC days. The PC-101 keyboard came out several years after the original PC.

      1. Steve the Cynic

        Re: Ignore the obvious choice

        "Er, yes, but the PC/AT wasn't around in the PC days. The PC-101 keyboard came out several years after the original PC."

        Read. History. Carefully.

        The original PC (the 5150) came out in late '81, with an 83-key keyboard. It was superseded in 1983 by the PC/XT (the 5160) which used the same keyboard.

        The PC/AT (the 5170) came out in 1984, using an 84-key keyboard, the 84th key being SysRq. The keyboard connection was not compatible in either direction with the PC / PC/XT keyboard connection, except that many third-party keyboards had converter switches.

        The 101/102-key layouts reached us in 1987 with the arrival of the PS/2s.

        1. Nick Ryan Silver badge

          Re: Ignore the obvious choice

          It was worse when MS decided that Windows keys should be added to PC keyboards. But no ****ing help key, instead an arbitrary function key was "commonly" used instead. On the other hand, prior to this time the VT100 style system keyboard had specific Help keys... probably from the earlier as well, but I daren't try to remember pre VT-100 layouts.

          So along with the left and right windows keys and the menu key, they really missed the boat on specifying something useful, the Help key. And the "Any" key as well, that would have saved a lot of time hunting around for it... :)

Page:

This topic is closed for new posts.

Other stories you might like