C / C++ / MFC

10-Apr-08 19:39

Hello everyone,

Here is the comments from the book ATL Internals about in-process server considerations,

--------------------
Any thread that calls CoGetClassObject or CoCreateInstance must already have joined an apartment, or the call will fail immediately. So, does that mean all objects exposed from in-process servers must be equally at home in single-threaded and multithreaded apartments so that random clients isn't cause harm? Fortunately, no.
--------------------

Does it mean COM client (thread) and COM server no need to be the same type of thread model (e.g. STA, MTA, etc.)?

thanks in advance,
George

Stephen Hewitt10-Apr-08 20:14

10-Apr-08 20:14

George_George wrote:
Does it mean COM client (thread) and COM server no need to be the same type of thread model (e.g. STA, MTA, etc.)?

No. If the threading models are compatible direct communication occurs. If the threading models are incompatible the object is created in an apartment it can "live in" and the interface pointer is marshaled back to the caller -- communication occurs via a proxy and stub. This is one of COM’s key features.

Steve

George_George10-Apr-08 21:37

10-Apr-08 21:37

Thanks Steve,

How do you understand the quoted statements from my original question?

regards,
George

Stephen Hewitt10-Apr-08 21:39

10-Apr-08 21:39

I'm just going off my knowledge of COM.

Steve

George_George10-Apr-08 21:58

10-Apr-08 21:58

Are you kidding, Steve? You are COM expert, I know. How do you understand the statement in the book?

regards,
George

Stephen Hewitt10-Apr-08 22:04

10-Apr-08 22:04

George_George wrote:
Here is the comments from the book ATL Internals about in-process server considerations,

--------------------
Any thread that calls CoGetClassObject or CoCreateInstance must already have joined an apartment, or the call will fail immediately. So, does that mean all objects exposed from in-process servers must be equally at home in single-threaded and multithreaded apartments so that random clients isn't cause harm? Fortunately, no.
--------------------

Any thread that calls CoGetClassObject or CoCreateInstance must already have joined an apartment, or the call will fail immediately.

This part is pretty straight forward.

So, does that mean all objects exposed from in-process servers must be equally at home in single-threaded and multithreaded apartments so that random clients isn't cause harm?

This just seems to be another way of stating the question you asked.

Fortunately, no.

This is saying that the threading models of the client and server do not have to match but goes into no details on how this is achieved.

Steve

George_George10-Apr-08 22:07

Error loading ActiveX control in secondary desktop ?

10-Apr-08 22:07

Thanks Steve,

Question answered. Congratulations to get all COM knowledge back. Smile | :)

have a good weekend,
George

phucbaby10-Apr-08 19:31

phucbaby

10-Apr-08 19:31

In my app, i switch to a new desktop and pop up a dialog with an ActiveX control (Flash object) but got an erre msg in new desktop:

Debug Assertion Failed!
...
File: f:\rtm\vctools\vc7libs\ship\atlmfc\src\mfc\oleinit.cpp
Line: 78

But if i pop up this dialog in current desktop,it's fine.

Pls help me !

Re: Error loading ActiveX control in secondary desktop ?

phucbaby10-Apr-08 19:37

phucbaby

10-Apr-08 19:37

I switch using CreateDesktop (GENERIC_ALL...) and SwitchDesktop().

Re: Error loading ActiveX control in secondary desktop ?

leenmie13-Apr-08 4:03

leenmie

13-Apr-08 4:03

Could you give some code ?

WIN32 API to find Mouse Button State

poda10-Apr-08 19:29

poda

10-Apr-08 19:29

Dear Friends,
How to know whether the mouse button is Down
even if it moving.The WM_LBUTTONDOWN message is fired
only for the first time left button is pressed.
But how to know whether it is pressed all the time
though it may be moving.

Mark Salsbery10-Apr-08 19:54

Mark Salsbery

10-Apr-08 19:54

Maybe GetAsyncKeyState()[^].

Mark

Mark Salsbery
Microsoft MVP - Visual C++
Java | [Coffee]

Stephen Hewitt10-Apr-08 20:19

10-Apr-08 20:19

Calling GetKeyState may be a better choice as it reflects state of the button when the current message was generated.

Steve

Re: WIN32 API to find Mouse Button State [modified]

Mark Salsbery10-Apr-08 21:09

Mark Salsbery

10-Apr-08 21:09

I'm sure that's what I meant Smile | :)

Thanks Steve!

Mark

Mark Salsbery
Microsoft MVP - Visual C++
Java | [Coffee]

Nibu babu thomas10-Apr-08 20:26

Nibu babu thomas

10-Apr-08 20:26

poda wrote:
How to know whether the mouse button is Down
even if it moving.

Add event handler for WM_MOUSEMOVE. If the WPARAM for this event has MK_LBUTTON set then left mouse button is down, or if it has MK_MBUTTON, MK_RBUTTON then the corresponding mouse button is down!

If you are using MFC then use void OnMouseMove( UINT nFlags, CPoint point ).

nFlags will contain appropriate flags to indicate whether left, middle, right buttons or whether shift key is down. Use bitwise and to find out.

E.g.

if(( nFlag & MK_LBUTTON ) == MK_LBUTTON )
   AfxTrace( "You are dragging the mouse\n" );

Nibu thomas
Microsoft MVP for VC++

Code must be written to be read, not by the compiler, but by another human being.

Programming Blog: http://nibuthomas.wordpress.com

modified on Friday, April 11, 2008 2:58 AM

poda10-Apr-08 22:58

poda

10-Apr-08 22:58

Thanks for your replies friends.
I used like

<br />
SHORT LButtonDown;<br />
LButtonDown=GetKeyState(VK_LBUTTON);<br />
<br />
if(LButtonDown & 0x8000)<br />
{<br />
...code here<br />
}<br />

It works.Thanks a lot.