"Jenbo" <eamonnjenni***@yahoo.co.uk> wrote in message
news:1178037344.143350.257590@y80g2000hsf.googlegroups.com...
> Hi, I have an application that displays forms on the users desktop,
> what I am doing with it is having a SystemEvents
> SessionSwitchEventHandler to capture the lock of the desktop, I would
> like to kill the forms when this event is captured, that piece is
> fine, but I would like to sleep the app until the
> SessionSwitchReason.SessionUnlock happens then I can wake it again and
> go the correct part of the code? Can I wake a sleeping thread, or
> should I not sleep the thread at all, is there another way to get the
> thread to sleep or hibernate and then wake up?
>
> Thanks
> Eamonn
>

On 1 May, 18:41, "Peter Duniho" <NpOeStPe***@nnowslpianmk.com> wrote:
> On Tue, 01 May 2007 09:35:44 -0700, Jenbo <eamonnjenni***@yahoo.co.uk> 
> wrote:
>
> > Hi, I have an application that displays forms on the users desktop,
> > what I am doing with it is having a SystemEvents
> > SessionSwitchEventHandler to capture the lock of the desktop, I would
> > like to kill the forms when this event is captured, that piece is
> > fine, but I would like to sleep the app until the
> > SessionSwitchReason.SessionUnlock happens then I can wake it again and
> > go the correct part of the code? Can I wake a sleeping thread, or
> > should I not sleep the thread at all, is there another way to get the
> > thread to sleep or hibernate and then wake u
>
> It's not clear to me that there's really any need to do any of the things 
> you want to do.  What are your forms doing that you can't just leave them 
> open when the desktop is locked?  Windows will take care of hiding them 
>  from the user and preventing user input, and when the desktop is unlocked 
> then they will still be there, ready to go.  Windows is also smart enough 
> to not dedicate any attention to the user interface of the application 
> while it's essentially inactivated by the desktop being locked (not that 
> the UI component of an application is generally a significant load on the 
> computer in the first place).
>
> That said, it is possible to explicitly get a thread to simply stop doing 
> anything.  Under Windows, a thread cannot "sleep" or "hibernate" per se 
> (not in the sense that the computer can do those things), but you can 
> create a waitable event and call the WaitOne() method on that.  The thread 
> that calls that method will stop and wait at that call until some other 
> thread sets the event, releasing the thread for further execution.
>
> If you use a ManualResetEvent, and you insert a call to the event's 
> WaitOne() method at some appropriate spot (generally a point in some 
> code's processing loop that is executed often enough that latency for any 
> pausing of the thread will be low, but not so often that calling WaitOne() 
> causes a significant performance problem), then when you want the thread 
> to pause you reset the event and when you want the thread to continue to 
> set the event.
>
> Of course, you still need at least one thread around to process the 
> SessionSwitchEvent and manage any other threads that are blocked.  But I 
> suppose if you have a thread that is doing something particularly i/o or 
> CPU intensive and you want it to stop taking system resources when the 
> desktop is locked, you could build in this sort of pausing mechanism 
> without affecting the main GUI thread.
>
> If the above doesn't help you, perhaps you could be more explicit about 
> why it is you believe you need to "put your application to sleep" when the 
> desktop is locked.  This isn't something that normally needs to be done, 
> so you should probably be more clear about what abnormal situation exists 
> in your scenario that justifies or demands that sort of behavior.
>
> Pete

On May 1, 8:36 pm, "Peter Duniho" <NpOeStPe***@nnowslpianmk.com>
wrote:
> On Tue, 01 May 2007 12:27:09 -0700, Jenbo <eamonnjenni***@yahoo.co.uk> 
> wrote:
>
> > [...] So maybe I need to let it
> > process as normal and go back to my windows handler for the covering
> > of the desktop and check that piece
>
> I agree, that's likely to be a more fruitful approach.  Though, it may 
> simply be that there's a better way to maximize your form.
>
> Note that the real issue you're having is that your form changes size when 
> the desktop is locked.  You don't really want to suspend processing per 
> se.  You just think that doing so would have the side-effect of preventing 
> the form from changing size.
>
> It's been years since I did anything with full-screen desktop applications 
> and I haven't played around with the various methods to accomplish that 
> recently.  So I probably can't offer much direct advice.  However, perhaps 
> if you explained the technique you're using to maximize your form on the 
> desktop and what exactly happens to it after the desktop is locked and 
> then unlocked, you might get some good advice from other people in this 
> newsgroup about how to do that correctly.
>
> If you do that, I would suggest starting a new thread, since the title of 
> this thread has very little to do with what you're actually asking about. 
> You'll probably get better response by doing so, and the thread will be 
> easier to look up later if someone else runs into the same issue.
>
> Pete

On May 2, 5:41 pm, "Peter Duniho" <NpOeStPe***@nnowslpianmk.com>
wrote:
> On Wed, 02 May 2007 03:36:34 -0700, Jenbo <eamonnjenni***@yahoo.co.uk> 
> wrote:
>
> > Thanks for the ideas. Although I am not sure why the last paragraph
> > was necessary as I would say that the title of the thread describes
> > pretty well what I wanted to do at first and where I was going with
> > this then. Strange to get such a pedantic end of message.
>
> I'm sorry you found it pedantic, and honestly I don't see where you get 
> that.  My intent was to offer you advice as to how you could post a 
> question that is more likely to get the results you want.
>
> Personally, I think you could've done just fine keeping your thoughts on 
> the matter to yourself.  I don't see the point in bringing it up, but 
> since you did...
>
> Frankly, I left out a whole paragraph providing advice about how your 
> original post really wasn't the right way to ask the question.  Further 
> more, as I have pointed out and as it *seemed* that you had agreed with, 
> the correct solution here isn't actually to use any sort of threading API 
> to deal with the problem.
>
> Since threading isn't relevant to the question at all, having a subject 
> that mentions threading is a problem if you are looking for for help on 
> the question.  It may mislead people who might be able to help you into 
> thinking that they don't have anything to offer (if they are unfamiliar 
> with threading issues, for example).  Additionally, since restoring your 
> form to full-screen after unlocking the desktop *is* the goal, the subject 
> of the thread ought to mention that.
>
> All that said, you are obviously free to post your questions as you like. 
> I appreciate you letting me know that you don't find my input helpful; 
> saves me the trouble of putting any more time or effort into more of your 
> questions.
>
> Thanks,
> Pete