|
I have some problems with overloading + operator I think??? What I have is a static library with 2 classes array is the base class and just has the array basic functions and the overloaded[]. The Vector class has the arithmatic functions. with the following code i get the error cannot add 2 pointers but shouldn't the overload take care of that?
here is what i have
int Main
{
Vector* Test1 = new Vector(3);
Vector* Test3(NULL);
Test3 = Test1 + Test1;
}
Vector header
friend Vector& operator+(const Vector& LHS, const Vector& RHS);
|
|
|
|
|
What are you trying to do? Adding two pointers together isn't going to yield anything useful
except a bad pointer.
Mark
|
|
|
|
|
I am trying to use the overload to do vector addition. That code should call my overload right?
|
|
|
|
|
Your "+" operator overload takes two Vector references and you are passing it two Vector
pointers.
|
|
|
|
|
How do I pass it 2 references? I have already tried *Test1 + *Test1 and &Test1 + &Test1 and neither will work.
|
|
|
|
|
I'm still not clear what you are trying to do...
What is a "Vector" and why do you pass 3 to its constructor?
Maybe this...(?)
int Main
{
Vector* Test1 = new Vector(3);
Vector Test3(NULL);
Test3 = *Test1 + *Test1;
}
|
|
|
|
|
I got it
it was (*Test3) = (*Test1) + (*Test1);
|
|
|
|
|
Test3 was an unallocated pointer in your original code. That's not good ((*Test3) = ...) 
|
|
|
|
|
First, ask yourself this question: What is the sum of two addresses? (Hint: A similar question is: What is the sum of two offset values?)
To solve your particular problem:
Vector Test1(3);
Vector Test3 = Test1 + Test1;
--
Not Y3K Compliant
|
|
|
|
|
Now I get
TestLibMain.obj : error LNK2019: unresolved external symbol "public: float & __thiscall Array::operator[](int)" (??AArray@@QAEAAMH@Z) referenced in function _main
ArrayLib.lib(Vector.obj) : error LNK2001: unresolved external symbol "public: float & __thiscall Array::operator[](int)" (??AArray@@QAEAAMH@Z)
ArrayLib.lib(Vector.obj) : error LNK2019: unresolved external symbol "public: float const & __thiscall Array::operator[](int)const " (??AArray@@QBEABMH@Z) referenced in function "class Vector & __cdecl operator+(class Vector const &,class Vector const &)" (??H@YAAAVVector@@ABV0@0@Z)
|
|
|
|
|
Hello i have a problem with event sink. I need to contol a NewWindow when it pops-up (to change if its visible toolbar)
i'm not sure about pDispParams->rgvarg[1].pdispVal; parametr. SOmehow i dont have handle to the new window.
Please help...
Code for NEWWINDOW2 event:
HRESULT __stdcall DWebBrowserEventsImpl::Invoke(DISPID dispIdMember,
REFIID riid,
LCID lcid,
WORD wFlags,
DISPPARAMS __RPC_FAR *pDispParams,
VARIANT __RPC_FAR *pVarResult,
EXCEPINFO __RPC_FAR *pExcepInfo,
UINT __RPC_FAR *puArgErr)
{
if (dispIdMember == DISPID_NEWWINDOW2)
{
//event sink is working because i can get a test message every time a new window pops
//but somehow i cant control a new wondow ?
MessageBox(NULL,"TEST","TEST",NULL);
//////////////////////////////////////////////////////////// this part wont work ?
CComQIPtr< IWebBrowser2> Nwin= pDispParams->rgvarg[1].pdispVal;
Nwin->put_Visible(TRUE);
Nwin->put_MenuBar(TRUE);
Nwin->put_ToolBar(FALSE);
}
return NOERROR;
}
|
|
|
|
|
I am currently designing a 2D/3D interface package based on Direct3D, but need to include a DirectShow VMR9 window in the interface which will be diplaying video from a capture card or pre-recorded videos.
The problem is that I have a render loop that constantly updates and redraws the interface, which causes the embedded video to flash. Is there a way that I can tell part of a surface not to be redrawn, i.e. the surface behind the vid window? It is on top of a large fullscreen size surface which holds a background image. Else is there a simple way to place individual frames of a my video onto a Direct3D surface which can then be drawn on top of the backgroung image.
I'm fairly new to Direct3D and very new to DirectShow. Any help is appreciated.
Thanks,
Dustin
|
|
|
|
|
|
I have a static boolean :
static bool booleanValue= true;
in my stdafx.h file.
I assumed this variable would have one instance across my application. But this assumption is apparently wrong. Within different objects, this boolean has separate instances, thus voiding its use.
I.E.
------------------------------
Have a class:
#include "stdafx.h"
class Blah {
int stuff;
void doStuff() {
if (!booleanValue) {
stuff++;
}
}
};
Have another class"
#include "stdafx.h"
class AnotherClass {
int stuffedPeppers;
void intoTheAbyss() {
if (!booleanValue) {
stuffedPeppers++;
}
}
};
------------------------------
The addresses for the same variable are different (same thread).
Where am I going wrong?
|
|
|
|
|
There's an instance for every module which includes the stdafx.h header file.
Try
In stdafx.h:
extern bool booleanValue;
And in one cpp file:
static bool booleanValue= true;
|
|
|
|
|
That worked. Thanks!
Had to use just:
bool booleanValue = true;
in cpp file.
So should this be done for functions as well?
|
|
|
|
|
switang wrote: Had to use just:
bool booleanValue = true;
Yeah, I suppose static is redundant in that case
switang wrote: So should this be done for functions as well?
Yes. Declare them in the header, define them in cpp files. Unless you need specific instances
for each cpp module. In that case (file local scope) you probably wouldn't want the declaration
in the header file anyway.
Mark
|
|
|
|
|
Thanks,
...Learn'n me some'm
Now I am a C++ Master...
|
|
|
|
|
Hello,
If you need global variables and functions just use class with static public members like this:
class GlobalWrap
{
public:
static int i1;
static double d1;
static void* pv1;
};
You can add constructor and destructor for initialization and memory cleaning and use it like this:
GlobalWrap::i1;
GlobalWrap::d1;
GlobalWrap::pv1->…
Regards,
Leonid
|
|
|
|
|
Thanks.
Never thought to do it that way...thanks again.
|
|
|
|
|
Hi
I'm working in a net monitoring system, and I sending on certain events packages to a server.
But, if the net falls I need to save the packages, and when the connection is restablished send them.
So, Im saving the packages to a text file, but when I recover the packages to send them, they are in reverse order, the newest is sent before the oldest.
Is any way I can get the packages from the beginning of the file and send them and then "move" the begining of the file?
Should I approach the problem from a different side?
In resume, I need to know how is easier to save a FIFO queue to a file.
Any idea?
Thanks 
|
|
|
|
|
You can fix this in 1 of 2 ways (easily). You can pop the items off to a stack and then write the stack to the file (and then just read it normally), or you can write it off in the same order you are doing it, but read it into a stack and then move it to your queue. Either case will fix your ordering problem.
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
But I'll have a lot of disk operations, and I wanted to avoid them
Well, seems that I'll have to do that
Thanks a lot Zac
|
|
|
|
|
You can avoid the excess disc operations by using a vector instead of a stack and then using a bulk write/read operation so the rest of the algorithm deals with data that is in memory already.
If you decide to become a software engineer, you are signing up to have a 1/2" piece of silicon tell you exactly how stupid you really are for 8 hours a day, 5 days a week
Zac
|
|
|
|
|
Hello,
Just use serialization for Array class.
Regards,
Leonid
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
