GetAsyncKeyState(VK_LBUTTON) true after DoModal()
in [MFC]
|
From: Simon on 7 Apr 2010 01:23 On 2010/04/06 07:39 PM, David Ching wrote: > "Simon" <bad(a)example.com> wrote in message > news:#4ARD0Y1KHA.4724(a)TK2MSFTNGP02.phx.gbl... >> Because if I click on the OK button the dialog box only closes when I >> release the left button, so why would (GetAsyncKeyState(VK_LBUTTON)& >> 0x8000) == 0x8000 ) return true. >> > > VK_LBUTTON is always the physical left mouse button, so if you have > swapped mouse buttons, you need to check VK_RBUTTON instead: > http://msdn.microsoft.com/en-us/library/ms646293%28VS.85%29.aspx Thanks, I did not swap the buttons. > > Also, it should be OK the way you have it, but just do: > > if ( GetAsyncKeyState(VK_LBUTTON & 0x8000) ) // don't compare == 0x8000 > ; I am curious, why do you suggest that? I agree that both essentially mean the same thing, but why do you suggest not comparing to 0x8000? Also are you sure it is if ( GetAsyncKeyState(VK_LBUTTON & 0x8000) ) and not if ( GetAsyncKeyState(VK_LBUTTON) & 0x8000 ) ? > > If this still does not work, what if you close the dialog by setting > focus to the OK button and use the keyboard? Does it then work? There > might be something going on with the dialog's DoModal() as that is a > separate message loop running, which might screw up the cached keyboard > state (although I understand GetAsyncKeyState doesn't use a cached > state, it is supposed to be what the hardware state is at the moment you > call it.) When I select the Ok button with the keyboard then the left button does not report as been down. Regards, Simon
From: David Ching on 7 Apr 2010 02:50 "Simon" <bad(a)example.com> wrote in message news:Ok9B0Jh1KHA.5996(a)TK2MSFTNGP05.phx.gbl... >> if ( GetAsyncKeyState(VK_LBUTTON & 0x8000) ) // don't compare == 0x8000 >> ; > > I am curious, why do you suggest that? I agree that both essentially mean > the same thing, but why do you suggest not comparing to 0x8000? > > Also are you sure it is > if ( GetAsyncKeyState(VK_LBUTTON & 0x8000) ) > and not > if ( GetAsyncKeyState(VK_LBUTTON) & 0x8000 ) > ? > Yes, you are right, it should be the latter one: if ( GetAsyncKeyState(VK_LBUTTON) & 0x8000 ) Essentially GetAsyncKeyState() returns a bitfield, you are masking the most significant bit, and you need to test whether that is non-zero. That intent is most clearly expressed by not doing == 0x8000, which implies comparison to an integer (not a bitfield) and does not imply testing for bits set. >> >> If this still does not work, what if you close the dialog by setting >> focus to the OK button and use the keyboard? Does it then work? There >> might be something going on with the dialog's DoModal() as that is a >> separate message loop running, which might screw up the cached keyboard >> state (although I understand GetAsyncKeyState doesn't use a cached >> state, it is supposed to be what the hardware state is at the moment you >> call it.) > > When I select the Ok button with the keyboard then the left button does > not report as been down. > I don't exactly know what's going on, but try using GetKeyState() instead of GetAsyncKeyState(). I'm not sure if this works with VK_LBUTTON, so also try GetKeyboardState(). These functions sync the results with the message loop. -- David
From: Goran on 7 Apr 2010 08:25 On Apr 6, 8:13 pm, Joseph M. Newcomer <newco...(a)flounder.com> wrote: > Perhaps because you had to click the OK button? And the mouse button is still down? You > are essentially seeing the result of the difference between computer speed and human > speed; within a few tens to a small integer number of hundreds of microseconds after you > have clicked the button, the mouse button state is examined, and according to the > instantaneous state, the mouse button is still down, because the message that it has come > up has not yet been processed. I find the basic idea of this test more than a little > weird, actually. Why would you care if the mouse button was down a few nanoseconds after > the DoModal() returns? No, there really is something fishy (I tried yesterday). You have to __release__ the mouse button for the UI button to fire "clicked" event (I am guessing, I didn't trace through that, I just called DoModal()). That's normal behavior. So when I release my mouse button, UI button should fire "clicked" and mouse button it should exactly be released. But GetAsyncKeyState says it isn't! Amazing! Why?! How!? I have no explanation. Looks like a question for MS people. > >Is the left button really down? > > >Is there something else I could do to check if the left button is down. > > **** > Please explain why you think this is important. That, I would like to know why, too. Just out of curiosity, might learn about some use-case I never met in the wild. Goran.
From: David Lowndes on 7 Apr 2010 09:10 >Looks like a question for MS people. One for Raymond Chen maybe? I wonder what the value may be if a sleep were added before the GetAsyncKeyState call? Dave
From: AliR on 7 Apr 2010 09:11 Which OS is this on. I tried this on XP and GetAsyncKeyState said that the key was down. AliR. "Goran" <goran.pusic(a)gmail.com> wrote in message news:e9e9d8fa-1de1-4415-a311-87ccd1ed5f78(a)g10g2000yqh.googlegroups.com... On Apr 6, 8:13 pm, Joseph M. Newcomer <newco...(a)flounder.com> wrote: > Perhaps because you had to click the OK button? And the mouse button is > still down? You > are essentially seeing the result of the difference between computer speed > and human > speed; within a few tens to a small integer number of hundreds of > microseconds after you > have clicked the button, the mouse button state is examined, and according > to the > instantaneous state, the mouse button is still down, because the message > that it has come > up has not yet been processed. I find the basic idea of this test more > than a little > weird, actually. Why would you care if the mouse button was down a few > nanoseconds after > the DoModal() returns? No, there really is something fishy (I tried yesterday). You have to __release__ the mouse button for the UI button to fire "clicked" event (I am guessing, I didn't trace through that, I just called DoModal()). That's normal behavior. So when I release my mouse button, UI button should fire "clicked" and mouse button it should exactly be released. But GetAsyncKeyState says it isn't! Amazing! Why?! How!? I have no explanation. Looks like a question for MS people. > >Is the left button really down? > > >Is there something else I could do to check if the left button is down. > > **** > Please explain why you think this is important. That, I would like to know why, too. Just out of curiosity, might learn about some use-case I never met in the wild. Goran.
First
|
Prev
|
Next
|
Last
Pages: 1 2 3 4 Prev: MSMQ problem Next: Setting PropertySheet Title (Wizard mode) |