One of the major annoyances of case sensitive passwords is accidentally having the capslock key on. This can easily eat up several entry attemps, locking you out of the system. Sure, some newer software will warn you that capslock is on, but that doesn't always help, especially if you haven't had your morning cofee yet.
My question is, why doesn't software allow inverse passwords? If AbC123 is your password aBc123 would also be acceptable. This only cuts the total number of possibilites in half, not a significant reduction compared to the total number of possibilities.
Edit: Alternate version after realizing not all systems de-capitalize with both shift and capslock pressed. Have the software ignore or correct for the capslock while entering passwords, rather than just telling you it's on. Since the OS is still receiving all keystrokes from the keyboard, it should be possible to interpret what those keystrokes would be without capslock on.-- MechE, Oct 20 2009 My caps-lock only capitalizes, it doesn't invert - capslock, shift-A still comes out as A, not a. Hm. But if it did toggle case as it should, wouldn't the counterpart to AbC123 be aBc!@# ?-- jutta, Oct 20 2009 [MechE] means holding the shift key while the caps lock is active. This has the effect of reversing the caps lock action. "ABC123" becomes "abc!@#".
(Presumably one would only be pressing the shift key with the expectation of generating the shifted character. The author doesn't appear to be shifting the "123" part of the example in the idea, although that's not made clear.)-- phoenix, Oct 21 2009 Okay, I just realized the having shift de-capitalize capslock is not a standard as I had thought. Windows machines do invert, my Mac does not. The example in the text were typed on windows at work, with the capslock on for the second, same keystrokes otherwise.
That being said, I was trying to enter an alphanumeric mixed case password, no special characters. On every keyboard I've ever used capslock does not toggle numbers to special characters, the shift key is required. Maybe the correct version of this idea is to have the software disable capslock instead of accepting both versions, I'll append that to the main idea.-- MechE, Oct 21 2009 This fully baked. Most Domain controllers allow for ignoring case.-- shahcat, Oct 21 2009 This is not ignoring case, that wouldn't care what the case is. This is still accepts dual case passwords, that is you still have to hit shift and the letter where needed, but it won't care what that combination actually produces on the screen.-- MechE, Oct 21 2009 The fact that some systems do not invert case when you have caps lock on means that you will always have to accept ALL CAPS versions; meaning that the case specificity is entirely lost.-- vincevincevince, Oct 21 2009 [V^3] Not neccesarily. Even if I have capslock on, the keyboard is still sending signals saying that shift was pressed/held in combinaation with another key. It should be possible for the operating system to neglect the capslock key, and just accept the input as if it was off, regardless of its actual state.-- MechE, Oct 21 2009 MechE, that is not only OS but more importantly platform dependant. The password checking mechanism used by websites is based upon the final string of text which is sent to a remote server for verification. The remote server has no access or way to access specific shift sequences used during the typing of the password.-- vincevincevince, Oct 21 2009 Most browsers, however, are now able to tell if something is a password (they ask if you want to save it). If they can do this, then they can ask the OS to pass them the corrected version, rather than the characters. This would handle it locally and pass the remote server the correct information. Definitely some programming involved, but shouldn't be impossible.-- MechE, Oct 21 2009 MechE, you are right; however that is likely to require rewriting the browser... which is something that web developers would love to do, but cannot yet achieve!-- vincevincevince, Oct 21 2009 I think if the feature was added to the OS, then the browsers would be re-written to support it in short order.-- MechE, Oct 22 2009 You are right [MechE], and in about ten years you will be able to start relying on users have a new OS and a new browser that supports it :)-- vincevincevince, Oct 22 2009 random, halfbakery