The shift key overrides NumLock

Perhaps not as well-known today as it was in the days when the arrow keys and numeric keypad shared space is that the shift key overrides NumLock.

If NumLock is on (as it usually is), then pressing a key on the numeric keypad while holding the shift key overrides NumLock and instead generates the arrow key (or other navigation key) printed in small print under the big digits.

(The shift key also overrides CapsLock. If you turn on CapsLock then hold the shift key while typing a letter, that letter comes out in lowercase.)

Perhaps you might decide that this little shift key quirk is completely insignificant, at least until you try to do something like assign Shift+Numpad0 as a hotkey and wonder why it doesn't work. Now you know.

Comments (22)
  1. Mack says:

    A propos *: can somebody explain why different symbols for multiplication are used in different countries?

    In Germany, we write ordinary (real) multiplication as a small centered dot. Cartesian set product and vector cross product is a little cross (like a floating x).

    Americans seem to write a large x for real multiplication, and a smaller cross for the vector cross product.

  2. Steve Loughran says:

    Laptops implement numlock as an overlay to the normal keys. Shift num key on laptop will give you the normal keys back, not the cursor stuff, which is off to the side.

    There is nothing like the joy of aterminal-services login from a laptop that sets the numlock option, you sit there hitting your password "say mmmm4" and being rejected, because you are really entering "0004" but cannot see it because you are getting * chars up instead. Really tsclient on laptops should ignore numlock settings for the local keyboard (but not external ones, sigh)

  3. Anonymous Coward says:

    IMHO what should be done for passwords is to show you the character typed for a second or two and then replace it with a *. Many WAP phones do this.

    Numlock is yet another one of those things that Microsoft code seems to go out of its way to turn off despite the user turning it on (status bars is the other). Ultimately I end up turning it on in the bios and in teh registry:

  4. Anonymous Coward says:

    Re: Laptops implement numlock as an overlay to the normal keys. Shift num key on laptop will give you the normal keys back, not the cursor stuff, which is off to the side.

    I just tried this on my Asus M6N notebook, and I get the arrow keys when I do a "Shift-Numpad x"…

    Still this is useful to know! Thanks Raymond.

  5. James Mastros (theorbtwo) says:

    The arrow keys and the numeric keypad still share space. It’s just that there’s also extra arrow keys that don’t share space, just like there have always been extra number keys that don’t share space (but do share space with puncuation).

    Hint: You can move diagonaly using the page up/page down/home/end keys in many games.

    Oddity: non-numlock numeric-key-pad 5 is normally meaningless, as opposed to the operator keys, which mean the same thing numlocked or un-numlocked.

    More oddities: In contential europe, using shift with capslock on commonly doesn’t temporarly override caps-lock, but rather turns it off. Also, the / and * "grey" keys are marked with the bar-with-two-dots and x operators you learned in elementry school, not with / and * like they are on US keyboards. They have exactly the same meaning, though.

  6. Raymond Chen says:

    "show you the character for a second or two".

    I hope that was a joke. Think of all the passwords you could steal with a video camera in a public library. (Or try using your hand to cover a twenty-foot tall screen in the front of a large meeting hall.)

  7. Anonymous Coward says:

    It isn’t a joke. If you can already observe that then you can observe where their fingers go on a keyboard, or install something physical or even software.

    Typing in passwords is one of the most user unfriendly things in computing today. The systems go out of their way to avoid giving you feedback. Fortunately most give you a clue as to how many characters you have typed. Some even notice that you have caps lock down and may alert you in various interesting ways. (It is fun to try on both Windows and Unix).

    So most people sit there typing in stuff, and the computer will tell you it is wrong. Of course you have no idea if what you typed is different than what you think you typed or if the password is genuinely wrong.

    (As an aside most people di type fairly badly. The easy way to verify this is in IM conversations were most just type away without corrections).

  8. Nope says:

    "If NumLock is on (as it usually is)". I have to disagree on this one. It is on if the user turns it on; on a fresh 2000 – XP install it is off. Not to talk about Linux or stuff.

  9. Adelle Hartley says:

    If you can already observe that then you can observe where their fingers go on a keyboard, or install something physical or even software.

    I have heard of a scam here in Australia where someone installs their own card-reader into an automatic teller machine, and a hidden camera watches the user to capture their secret number. The card-reader enables the scammer to make a copy of the card.

    However, I don’t think having the actual characters pop up on the screen and then disappear would be a good feature. How would the user know whether a newly encountered log-on screen has this feature? Maybe its just that I type quickly enough that I don’t have trouble hiding my password from casual observers, but I think a better solution would be to have a "Reveal" command.

  10. Cooney says:

    Americans seem to write a large x for real multiplication, and a smaller cross for the vector cross product.

    We did this in grade school, but starting in middle school (7th grade at the time), we went to the centered dot. Of course, with algebra, you typically just leave it out entirely unless you’re multiplying a bunch of constants.

  11. Merle says:

    I would expect shift to override caps lock. I mean, caps lock is like holding shift down, so the second shift should de-toggle it. But my "intuition" may be based on the fact that I have been used to it for so long.

    Ditto for numlock. But I never use the numberpad for numbers. Never really learned the 10-key pad (and would be very confused as some of them are upsidedown — compare your numpad to the phone). It really confuses me to use systems with numlock on. (I’m trying to scroll, why isn’t it? … oh….)

    It actually irritates me some that caps lock does *not* act like a global shift. For example, it does not shift numbers or punctuation. Then again, just as I do not spend all day entering numbers, it is a rare day when I need to capitalize more than a few words in a row.

    I love the accessibility feature that beeps if you (accidentally) hit the lock keys. That’s saved me from vi accidents so many times. A great feature.

  12. Axolotl says:

    This reminds me of another puzzle that has afflicted almost every version of Windows.

    Why does the insert key keep switching back to overwrite mode, seemingly at random (although it tends to happen more often if you scroll back a paragraph, click the cursor in the middle of a sentance and start to type)?

  13. Norman Diamond says:

    Base note:

    > If NumLock is on (as it usually is),

    That "usually" is true only in countries where desktop machines outnumber laptops.

    9/7/2004 10:49 AM Merle:

    > It actually irritates me some that caps lock

    > does *not* act like a global shift.

    IBM used to make some terminals where Shift Lock operated exactly that way, the same as on a physical typewriter. That feature was unpopular. Although Caps Lock is less backwards compatible, it was so much more convenient than Shift Lock was, that this feature became wildly popular.

  14. foxyshadis says:

    The one major blunder I see in ms password protection is the cruel and unusual hiding and then requiring double entry for a WEP key. I guess just typing 26 characters once isn’t enough, even though it goes against every other password input – this isn’t creating a new one, it’s just inputting an old one. At that point the chance of error is probably 80% for both entries, and since I can’t tell which was wrong or which character I’m screwed. So I just type it in an open field and copy it out twice anyway. </rant>

  15. Merle says:

    Ah, Norman, I had forgotten about shift lock. I did use IBM typewriters when I was growing up; maybe that’s why I find caps lock so different.


  16. Norman Diamond says:

    9/8/2004 9:25 AM Merle

    > did use IBM typewriters when I was growing up

    Actually all makers of mechanical typewriters had shift lock keys not caps lock keys. By the way these had other weird characteristics, for example backspacing and typing a replacement character didn’t erase the first character, it left both characters displayed in the same character cell. ‘Course that’s not the worst that can happen to you. If a car crashes and it’s a real car, it doesn’t restart when you press Ctrl-Alt-Del. Such user-unfriendly mechanical devices ^u^

    My reference to an IBM terminal was a computer terminal which wasn’t a mechanical typewriter. That had a backwards compatible shift lock key instead of caps lock key, and that’s what sucked. That’s the only terminal that I’m aware of with that misfeature.

  17. ööé says:

    Shift does not always override Caps-Lock. At least not when you are using a swiss keyboard layout.

    There are three keys with umlauts.

    normal shifted

    ö é

    ü è

    ä à

    With Caps-lock on

    Ö É

    Ü È

    Ä À

  18. I agree with foxyshadis. Why does a WEP key have to be protected by stars? There should at least be an "unhide" option. Whenever I am entering a WEP key I am usually reading it off a piece of paper that I have printed out – it is secure enough, because it isn’t like the guy sitting outside in his car with a laptop can see the bit of paper sitting on my desk.

  19. Ben Cooke says:

    The manufacturer-supplied configuration software for my wireless card lets you type the WEP key(s) in the clear but once you click OK or Apply it’s starred out.

    Of course, all it takes is a Windows message to the text field to un-star it again…

  20. yeah yeah yeah says:

    Remind me – does the shift lock/caps lock key affect the number keys (top row, beneath the F keys) on US/UK kbds? It does on German ones and it is damned annoying.

  21. Norman Diamond says:

    The difference between shift lock and caps lock key should be that shift lock does and caps lock doesn’t. In my experience with caps lock on US and Japanese keyboards, caps lock correctly refrains from affecting other than Italian letters.

    Hmmmm. If the IME is active then caps lock doesn’t even affect Italian letters — the IME toolbar still shows CAPS as active, but caps are not active and shift key operations are ordinary instead of inverted. (At least in Windows XP SP2.)

  22. yeah yeah yeah says:

    "The difference between shift lock and caps lock"

    I only have the one key :)

Comments are closed.

Skip to main content