C / C++ / MFC

Re: GetFocus() returns NULL

7-Mar-04 19:19

Simply write out the html tags, a technique followed by many application.

MSN Messenger.
prakashnadar@msn.com

Re: html generation

George L. Jackson8-Mar-04 8:01

George L. Jackson

8-Mar-04 8:01

You can use XML/XSLT to create HTML dynamically.

Re: html generation

basementman8-Mar-04 13:13

basementman

8-Mar-04 13:13

See here [^]

onwards and upwards...

GetFocus() returns NULL

Swinefeaster7-Mar-04 18:58

Swinefeaster

7-Mar-04 18:58

Hello,

Just wondering if any of you would know the answer to this:

For the win32 api function, ::GetFocus(), the docs say "If the calling thread's message queue does not have an associated window with the keyboard focus, the return value is NULL."

So how can you check if your application is active or not?

Thanks,

Swine

[b]yte your digital photos with [ae]phid [p]hotokeeper - www.aephid.com.

Prakash Nadar7-Mar-04 19:07

Thread programming question

7-Mar-04 19:07

Swinefeaster wrote:
So how can you check if your application is active or not?

you mean has focus or not ?
Well if it returns null it means its not active, rite ?
If the window returned is other than ur window then again ur window/application is not focused.

If you want to explicitly make ur application active call
SetFocus(...)

MSN Messenger.
prakashnadar@msn.com

directx 8 / 9 issues

siniani7-Mar-04 16:24

siniani

7-Mar-04 16:24

am a chinese online-game programmer. Now there is a problem suffering me. In game programming, how can I display the ime windows in fullscreen mode when using the directx 3d technology. Waitting for your helpful solutions.
Thanks!

Madmaximus7-Mar-04 15:41

Madmaximus

7-Mar-04 15:41

I have a question about the proper way of closing a thread with Windows APIs.
This is pseudo code. I left out details for clarity.

HANDLE m_hConEvent};
class ThreadA;
void main(void)
{
ThreadA *pthreadA = new ThreadA(); //Do I need to remove pointer from the heap.

do some stuff...
Wait();

}

class ThreadA
{
public
ThreadA()
{
CreateThread(FuncPtr);
SetEvent(m_hConEvent);

}
~ThreadA() {};

static int FuncPtr()
{
bool bExit;
HANDLE arHandles[] = {m_hTerminatorX, m_hConEvent};

while(bExit)
{
dwWait = WaitForMultipleObjects(2, arHandles, FALSE, INFINITE);

switch(dwWait)
{
case WAIT_OBJECT_0 + 1 :
{
//Do some stuff. Now done with thread or errored out.
SetEvent(m_hTerminatorX);
}
break;
case WAIT_OBJECT_0 :
{
bExit = false;
}
break;
}
}
//Now I'm done with thread do I need to remove *pthreadA pointer?
// Like use "delete this" command to remove the pointer from the heap?
delete this;
}
};

Thanks in advance. Wink | ;)

Prakash Nadar7-Mar-04 15:48

7-Mar-04 15:48

How about deleteing it in the main function itself ?

This space is empty.

Madmaximus7-Mar-04 16:19

Madmaximus

7-Mar-04 16:19

Thanks for the reply. Unfortunately in this scenario I can’t determine in the main routine when to delete the object.

Prakash Nadar7-Mar-04 16:38

7-Mar-04 16:38

Opps sorry for the stupid answer... You are creating a thread.
But i dont think you have access to the this pointer in the static method.

To delete the object in the thread function you need to pass the this pointer to the thread. Then you delete it when the thread has done its job.

But what if the same object initiates 2 threads???? then the second call to delete of the same object will fail.
You have to make sure that the same object does not create another thread.

MSN Messenger.
prakashnadar@msn.com

hot to set the step value in spin

Curi0us_George7-Mar-04 16:43

Curi0us_George

7-Mar-04 16:43

The simplest answer is yes, you definitely need to delete the ThreadA that pThreadA points to.

There are a few other things that strike me as notable/strange.

1. FuncPtr is static, and thus has no "this" context. "delete this" will result in a compile time error. Any member functions/variables will also be inaccessible from FuncPtr.

2. If you want to use "delete this", you need FuncPtr to accept a pointer to a the ThreadA object and use it to call an instance function.

e.g.:
int FuncPtr2()
{
// do stuff
return 1;
}

static int FuncPtr(void * param)
{
return ((ThreadA *)param)->FuncPtr2();
}

Now FuncPtr2 has an appropriate context to use "delete this". However, this is probably a bad idea. In my experience, it sometimes leads to strange bugs, and it's just not worth it. In addition, the behavior of "delete this" is undefined if the object "this" points to was allocated on the stack.

My suggestion would be to put some sort of checking mechanism into main. (I assume this is what Wait() is doing.) At the end of the function (or at the end of the relevant section), use this checking mechanism to wait for the thread to complete, and then delete the ThreadA object in the main method.

Mughi7-Mar-04 15:26

Mughi

7-Mar-04 15:26

I had set the min value and max value to spin by
m_spin.SetRange(250,500);
But then, how to set the step value to spin.
It should increment like 250, 300, 350, 400, 450, 500
Can anyone help me out, please.
It is so urgent. I thank in advance.

Mughilan

Prakash Nadar8-Mar-04 1:10

8-Mar-04 1:10

Could not find anything that will do that what you are asking for by default. You have i guess customise it.

MSN Messenger.
prakashnadar@msn.com

David Crow8-Mar-04 4:47

David Crow

8-Mar-04 4:47

Will SetStep(50) not do what you require?

"The pointy end goes in the other man." - Antonio Banderas (Zorro, 1998)

Roger Wright8-Mar-04 9:21

Roger Wright

8-Mar-04 9:21

CSpinButtonCtrl doesn't have a SetStep() method that I can find. I'd try providing a handler for the NM_RELEASEDCAPTURE message that calls GetPos(), adds or subtracts 50, then calls SetPos(). Of course, someone more knowledgable might have a better solution. Blush | :O

Will Build Nuclear Missile For Food - No Target Too Small

David Crow8-Mar-04 9:28

David Crow

8-Mar-04 9:28

Indeed. I missed the reference to CSpinButtonCtrl. He'll need to use SetAccel() instead.

"The pointy end goes in the other man." - Antonio Banderas (Zorro, 1998)

Need opinions on proposed Plugin framework

Roger Wright8-Mar-04 10:09

Roger Wright

8-Mar-04 10:09

DavidCrow wrote:
use SetAccel()

I wondered about that... I haven't used it, but is it possible to set the first element of the UDACCEL structure to [0,50]? If it doesn't choke on the 0 value for elapsed time I would expect it to do the job adequately, but the documentation doesn't elaborate on it much.

Will Build Nuclear Missile For Food - No Target Too Small

Anonymous7-Mar-04 15:13

Anonymous

7-Mar-04 15:13

I am currently working on a very simple plugin framework for a C++ application. My proposal is as follows:

1 - All plugins are DLL's, and are uniquely identified by a string GUID constant within the DLL.

2 - All Plugins must contain a function called "InitDLL". InitDLL receives a structure defined as follows:

struct plugin_info
{
char* plug_guid;
char* plug_name;
char* plug_desc;
plug_send send_ptr;
};

In InitDLL, plug_guid, plug_name, and plug_desc are filled out. A local copy of send_ptr is kept in every plugin DLL.

3 - plug_send is a pointer to a function in the main program that receives events from the dll. It is defined as follows:

typedef void (*plug_send)(char* plug_guid, char* event, void* event_params_struct);

The local copy of send_ptr stored in the DLL can be called at any time to send events to the host application. Several structs will be defined for passing parameters (as the event_params_struct), and the struct used will depend on what event is being passed to the main application.

4 - Certain structs will also have pointers to functions in the DLL themselves. These functions will be used as callbacks, to allow the host application to send events back to the DLL.

5 - Both the host application and DLL's will have access to a public API, in the form of a C++ header file.

Does anyone see any major problems with this proposal, or will it work decently for a very basic plugin system? All feedback appreciated.

Re: Need opinions on proposed Plugin framework

Prakash Nadar7-Mar-04 15:43