Comments by Eugen Podsypalnikov

Eugen Podsypalnikov 18-Dec-14 9:49am View

Thank you :)

Eugen Podsypalnikov 10-Dec-14 3:21am View

No, but the answer can be deleted
since it does not fit to your question :)

Eugen Podsypalnikov 10-Dec-14 2:43am View

Sorry, it was not the subject of my answer :)
Maybe you will discover it...

Eugen Podsypalnikov 10-Dec-14 2:41am View

You are right, in you case it is another list :)

Eugen Podsypalnikov 21-Nov-14 7:29am View

Deleted

first class: virtual void Sort() = 0; // all methods may be also protected :)
third class: virtual void Sort() {...}

Eugen Podsypalnikov 17-Dec-13 6:06am View

Deleted

As it already said: void pushElement(std::vector(int)& stack, int iElement) { stack.push_back(iElement); } :)

Eugen Podsypalnikov 13-Dec-13 4:31am View

You are wellcome, Stefan :)

Now we know: "why ?" and "why not ?"

Ciao.

Eugen Podsypalnikov 13-Dec-13 4:26am View

You and OP know
why it must be the same to the implicit form by Mr. Stroustrup
as well :D

Eugen Podsypalnikov 13-Dec-13 4:20am View

Here is the answer for you, Stefan, and for OP:

The explicit form of the Standard copy constructor
MUST BE THE SAME to its implicit form, by Mr. Stroustrup :)

Eugen Podsypalnikov 13-Dec-13 4:07am View

Are you meening
the implicitly implemented constructor by Mr. Stroustrup
in the class above, Stefan ? :D

Eugen Podsypalnikov 13-Dec-13 3:52am View

// "Because Stroustrup said so" is not an Explanation

Realy ? I think so :) Please see below:

{
class A
{
int i;
public:
A() : i(0) {}
};

A a1;
A a2(a1); // Here is the hidden Speach by Mr. Stroustrup
}

All other links just use the Speach, not more :)

Eugen Podsypalnikov 13-Dec-13 3:31am View

There is no explanation about any wrong Information, sorry :)
class B
{
public:
B();
B(const B&); // Why ? -Since it is a standard by Mr. Stroustrup :)
B(const B*); // Why not ? -Yes, you may do it as well :)
};

Eugen Podsypalnikov 13-Dec-13 3:25am View

class B
{
public:
B();
B(const B&); // Why ? -Since it is a standard by Mr. Stroustrup :)
B(const B*); // Why not ? -Yes, you may do it as well :)
};

Eugen Podsypalnikov 13-Dec-13 3:22am View

But it ("per pointer") may be an "own copy constructor" :)

Eugen Podsypalnikov 13-Dec-13 3:08am View

I think Mr. Stroustrup is correct, it is the begin in my answer :)
Stefan is correct as well :)
And OP may use an own copy constructor per pointer, even so :D

Eugen Podsypalnikov 12-Dec-13 11:21am View

I have answered the question of OP (please see it in bold above).
My code is just for the question part "Why not ?"
And the text - for the part "Why ?"

You can have many copy-constructors
but only one of them - Default Copy Constructor, by Mr. Stroustrup :)

Eugen Podsypalnikov 12-Dec-13 11:00am View

I'm passing a comment above about some own copy-constructors as well :) :
class B
{
public:
B();
B(const B&);
B(const B*);
};

Eugen Podsypalnikov 11-Dec-13 10:18am View

Try to include "stdafx.h" in sources (.cpp) only, like the CDTXBBAR.cpp case :)
The stdafx.h is unique for every project.

Eugen Podsypalnikov 11-Dec-13 9:57am View

1. Is there "#pragma once" at the top of TextControl.h as well ? :)
2. Why is there '#include "StdAfx.h"' in a header ?

Eugen Podsypalnikov 11-Dec-13 8:17am View

Because of the placement of std::map::insert(..)
The call can be in a function scope only.
The declaration of the map object may be global :)

Eugen Podsypalnikov 11-Dec-13 7:57am View

It would be a "global" insert, before the object's usage in any functions :) :
class charMap : public std::map<char,char>
{
public:
charMap() {
insert(make_pair('{','}'));
}
} g_map;

Eugen Podsypalnikov 21-Mar-13 11:39am View

It is good to know ! Thank you Stefan :)
Have you already tested the using of the "__forceinline" keyword
for a header inline function/operator - as well ?

Eugen Podsypalnikov 21-Mar-13 10:42am View

Hi Stefan !
Could you test this model for the perfomance too ? :)

Eugen Podsypalnikov 21-Mar-13 9:17am View

Maybe... It is just shown how to make an operator "pure" inline :)

Eugen Podsypalnikov 20-Mar-13 11:20am View

// But Outlook crash when i try ti initialize _ApplicationPtr pApp
But... there is no any initializing here... :)

Eugen Podsypalnikov 20-Mar-13 9:04am View

Hm... I have understood that you are thinking and wanting :)

Eugen Podsypalnikov 20-Mar-13 9:04am View

Deleted

Hm... I have understood that you are thinking and wanting :)

Eugen Podsypalnikov 20-Mar-13 8:47am View

The best way is your way... Of course :)

Eugen Podsypalnikov 20-Mar-13 7:59am View

You are right ! The init part (list) must be injected firstly, if necessary :)

Eugen Podsypalnikov 20-Mar-13 7:59am View

You are right ! The init part (list) must be injected firstly, if necessary :)

Eugen Podsypalnikov 20-Mar-13 4:04am View

Probably in the same manner :)
class twoGlobalUints
{
static UINT& GetFirst()
{
static UINT uiFirst(0);
return uiFirst;
}
static UINT& GetSecond()
{
static UINT uiSecond(0);
return uiSecond;
}
};
// any.cpp
UINT test()
{
twoGlobalUints::GetFirst() = twoGlobalUints::GetSecond() = 11;
twoGlobalUints::GetFirst()++;
return twoGlobalUints::GetFirst() - twoGlobalUints::GetSecond();
}

Eugen Podsypalnikov 20-Mar-13 3:13am View

// implement the copy operation here,
// same as in the assignment operator.
pre_iterator(const pre_iterator& __original)
{
*this = __original; // ;)
}

Eugen Podsypalnikov 19-Mar-13 9:55am View

No, sorry, there is nothing more to chat for me :)

Eugen Podsypalnikov 19-Mar-13 9:04am View

EXE-directory is the directory of "hw0-windows.exe"
I have copied the "shaders" directory and the "wood.ppm" file -
from your project directory int the EXE-directory, then started the "hw0-windows.exe" :)

The debugging (F5 from the VS solution) went just fine,
immediately after unrar and without any exception :)

Eugen Podsypalnikov 19-Mar-13 8:25am View

WOW ! Very nice example ! Thank You :)
I have seen your teapot !!! :)
(just copy the directory "shaders" and the file "wood.ppm" into the EXE-directory)

Eugen Podsypalnikov 19-Mar-13 7:13am View

// http://stackoverflow.com/questions/12329082/glcreateshader-is-crashing
Most likely :)

Eugen Podsypalnikov 19-Mar-13 7:02am View

// what's story ?
A "story" would be the calling sequence including its used parameters :)

// what should i do ?
I would debug the code and then improve it :)

Eugen Podsypalnikov 14-Mar-13 12:00pm View

OK, in this case the String must handle its ref/unref/clone needs by itself... :)

Eugen Podsypalnikov 14-Mar-13 11:25am View

OK... Then we will need a memory manager for all such strings, to register and handle their references(inc/dec), sharing and modifing to/of the data :)

Eugen Podsypalnikov 14-Mar-13 11:24am View

Deleted

OK... Then we will need a memory manager for all such strings, to register and handle their references(inc/dec), sharing and modifing to/of the data :)

Eugen Podsypalnikov 14-Mar-13 3:40am View

// Do I need to use WM_SETFONT for every single control in every single dialog?
Please try it, for a single control. What is the result of this test, please ? :)

Eugen Podsypalnikov 13-Mar-13 3:31am View

Hi Thomas !
Sorry, I have never used it too :
http://msdn.microsoft.com/en-us/library/windows/desktop/aa373163%28v=vs.85%29.aspx .

Eugen Podsypalnikov 12-Mar-13 7:14am View

The UNICODE version of the solution may copy not all characters... :)

Eugen Podsypalnikov 7-Mar-13 9:28am View

Hm... why not just the using of a temp file from a temp directory ? :)

Eugen Podsypalnikov 7-Mar-13 9:21am View

Very likely... :)
Thank you !

Eugen Podsypalnikov 7-Mar-13 9:21am View

Deleted

Very likely... :)
Thank you !

Eugen Podsypalnikov 6-Mar-13 10:27am View

Yep, or :) :
BOOL CFlashingProgressBar::OnInitDialog()
{
BOOL bRes(CDialogEx::OnInitDialog());

SetTimer(EndDialogTimer,40*1000,NULL);
SetTimer(CounterTimer,5*1000,NULL);

return bRes;
}

Eugen Podsypalnikov 1-Mar-13 8:08am View

Just try it, set a some region to a bordered window :)
See also: http://www.codeproject.com/Articles/31734/SonicUI-A-Convenient-GUI-Engine-You-ve-Never-Seen .

Eugen Podsypalnikov 18-Dec-12 2:56am View

// worse cache performance, because things will be all over the place in memory
First - please remember that the all objects are in the memory (once or twice ;-))

I think, it (worse cache performance) is wrong :) ,
because a map will work with pointer- (or reference-) keys in its hash-table _more_faster_
as with full full-dumped-copies keys.

Above all, your code will be more canny (in the memory usage context).

Eugen Podsypalnikov 11-Dec-12 9:49am View

Try it :) :
for (i = 0; i < _countof(abc); i++) {
//..
}

Eugen Podsypalnikov 11-Dec-12 5:59am View

Some entries of the virtual functions table of an abstract class
can not be created by its owner object
(that is why "ISample aSample" would be an error for the compiler),
they must be defined by the derived objects.

The instances of the abstract classes
are only accessable per such derivations.

In C++, any not-private base-constructor's call
is possible from any its derivation-constructor :)

Eugen Podsypalnikov 11-Dec-12 2:55am View

The daemon (in Windows - service) of the HW communication
could be started by the operating system up

and can provide an interface (OpenPort(), ClosePort(), ReadPort(), WritePort(), ..)
for a shell script's (terminal program) calls.

Here we have defined two SW-partners: daemon(1) and shell client program(2).

Eugen Podsypalnikov 11-Dec-12 2:43am View

Maybe, I could :)
An abstract class (for example: an animal)
could hold the all centered data (age, weight, ... of an animal)
or centerd functions (get_age(), ...)
but not implement one or more of the "needed functions"
(for example: "go to", "jump to", "swim to", "say hello", ...).
Any "not abstracted" animal must reimplement the "personal" functions in its own manner :)

Eugen Podsypalnikov 11-Dec-12 2:40am View

Maybe, I could :)
An abstract class (for example: an animal)
could hold the all centered data (age, weight, ... of an animal)
or centered functions (get_weight(), ...)
but not implement one or more of the needed functions
(for example: "go to", "jump to", "swim to", "say hello", ...).
Any "not abstracted" animal must reimplement the "personal" functions in its own manner :)

Eugen Podsypalnikov 11-Dec-12 2:31am View

// my requirement is to use it as a daemon (by calling it from shell script)
That is an answer: your application could communicate with the daemon :)

Eugen Podsypalnikov 11-Dec-12 2:09am View

// but the first one fails...
What code is it ? :)

Eugen Podsypalnikov 30-Nov-12 9:58am View

Hm... :)
Now the function is visible for the compiler
and invisible for the linker :)
Try to transport it (its correctly form):
CString CFILicenseSettings::GetCodeStartup1()
{
...
}
into the file of usage (FIServerDlg.cpp ?)
then - try again :)

Eugen Podsypalnikov 30-Nov-12 9:52am View

Now try to make an error inside the function's body,
for example: write "hello world" instead of "return".
Then try again, please, to see the output :)

Eugen Podsypalnikov 30-Nov-12 9:47am View

Now try to remove your comments (make any changes)
and to compile the file from its context menu in the project tree:
can you see any output (at the output window) ? :)

Eugen Podsypalnikov 30-Nov-12 9:42am View

Try comment the function itself in the FILicenseSettings.cpp,
not its call... :)

Eugen Podsypalnikov 30-Nov-12 9:38am View

Have you the same linker error
if you comment its cpp-implementation ? :)

Eugen Podsypalnikov 30-Nov-12 9:27am View

Can you see the function,
if you select "Open" in the file's context menu at the project tree ? :)

Eugen Podsypalnikov 30-Nov-12 9:21am View

Can you compile the file from its context menu
(and see the output of #pargma error) ? :)

Eugen Podsypalnikov 30-Nov-12 9:15am View

Can you see "Yes" in the entry "Excluded From Build"
in the general properties of the FILicenseSettings.cpp
(context menu for the file in the project tree) ? :)

Eugen Podsypalnikov 30-Nov-12 9:07am View

=> the function is invisible for the compiler.
Have you some
#ifdef
or
#if
in the FILicenseSettings.cpp, above the function ? :)

Eugen Podsypalnikov 30-Nov-12 9:01am View

Has the pragma (set it):
#pragma error
any compiling effect inside the function's body ? :)

Eugen Podsypalnikov 30-Nov-12 8:55am View

...and how it (the function) is declared in its header ? :)

Eugen Podsypalnikov 30-Nov-12 8:51am View

Are the (FIServerDlg.cpp and FILicenseSettings.cpp) files in the same project ? :)

Eugen Podsypalnikov 16-Nov-12 3:08am View

OK :)
- then use an additional call of ::IsWindowVisible(hwnFound) after the hwnFound = ::FindWindow(..)

Eugen Podsypalnikov 16-Nov-12 2:31am View

OK :)
- try to set the HWND variable to NULL before the second call of ::FindWindow(..)
- try to pass also the window class name (shown by Spy++) to the call of ::FindWindow(..)
- check that your searching window title is the same as the shown by Spy-tool (copy it there and then paste to your code)

Eugen Podsypalnikov 16-Nov-12 2:29am View

Deleted

OK :)
- try to set the HWND variable to NULL for the second call of ::FindWindow(..)
- try to pass also the window class name (shown by Spy++) to the call of ::FindWindow(..)
- check that your searching window title is the same as the shown by Spy-tool (copy it there and paste to your code)

Eugen Podsypalnikov 3-Nov-12 16:12pm View

As Jochen said, you could try to
- manipulate your manifest to provide the administrator rights for the application, or
- write a service started by the SYSTEM user and delegate such tasks to it

Eugen Podsypalnikov 2-Nov-12 6:42am View

Good point for a classical base class,
not for the interfaces, which are nomally defined to hide the objects behind them :)

Eugen Podsypalnikov 1-Nov-12 14:52pm View

What types of the variables and
what kind of the access (read/write) are wanted, please ? :)

Eugen Podsypalnikov 1-Nov-12 7:45am View

Please post here the buttons code (their .rc or/and .cpp parts) :)

Eugen Podsypalnikov 1-Nov-12 7:40am View

A refactoring of the code that you have could change the situation.
For example, you could pass a reference to an object of a matrix-class,
that could answer any question about its dimensions :)

Eugen Podsypalnikov 31-Oct-12 7:20am View

In the current context (list of the structure-copies),
you have to iterate the list to find the needed structure firstly
(for example, it could be identified by a member inside the structure).
The searching process can be slowly ("magabytes of data") inside the list :)
That is why, you could also use a std::map<DWORD, MYSTRUCT*> as the data container
and then access a needed MYSTRUCT* by the using of std::map::operator[] :
void MyContext::EditById(DWORD dwId, int iNewValue)
{
MYSTRUCT* pMS(m_map[dwId]);
if (pMS) {
pMS->iData = iNewValue;
}
}

Eugen Podsypalnikov 30-Oct-12 18:15pm View

Very understandable solution :)

Eugen Podsypalnikov 30-Oct-12 17:26pm View

Yes, I have got inclination to write the structure destructor properly :)
http://www.codeproject.com/Questions/486436/Howpluscanplusipluspassplustheplusthisplusstlplusl

Eugen Podsypalnikov 30-Oct-12 15:52pm View

3 stars - for three memory leaks :)

Eugen Podsypalnikov 29-Oct-12 12:35pm View

// Will this imply that pointer->data variable will also be deleted?
- Yes, the pointer will be deleted, in its uninitialized case as well :)

// I seem to get a run time error in class A destructor once is being called in this code
- There is no call of ~A() in this (posted above) code :)

Eugen Podsypalnikov 29-Oct-12 2:57am View

Only 3 as a help for the beginner :) :
- there is no explanation about the wrong used term "pointer to a member function"
- there are no deallocations in the provided code (= memory leaks)

Eugen Podsypalnikov 26-Oct-12 13:51pm View

That's right, see also the reset-toolbar usage here:
http://www.wretch.cc/blog/kuolun1003/13910292 :)

Eugen Podsypalnikov 28-Aug-12 7:19am View

Please look for example at the Windows File Explorer :) , there are:
- Directories pane (customers)
- File pane (customers orders)
- File preview pane (order details)

Eugen Podsypalnikov 21-Aug-12 9:00am View

Sure :) , that is why you can see the "~cSon()" output. Try it without the keyword virtual in the base, and you will win a memory leak :)

Eugen Podsypalnikov 16-Aug-12 16:58pm View

// When the status bar is first created, there are no strings to cancel
-- When the status bar is first created, m_nCount == 0, the loop will be not executed

// Reversing the two blocks of code solves that problem
-- Reversing the two blocks of code would produce memory leaks, at the second call

Good luck ! :)

Eugen Podsypalnikov 15-Aug-12 19:15pm View

Please try the last solution, ist this that you wanted ? :)

Eugen Podsypalnikov 15-Aug-12 11:34am View

Please try the update of the solution :)

Eugen Podsypalnikov 15-Aug-12 3:32am View

Sure, it is a way :)

Eugen Podsypalnikov 15-Aug-12 3:27am View

Yes, that is right :)

Eugen Podsypalnikov 14-Aug-12 16:16pm View

// But if some other process has already written into xyz.txt file,
// how do i know if my xyz.txt file is corrupted or not.
I think, the traced time could answer this question :)

Eugen Podsypalnikov 14-Aug-12 9:28am View

I will try it, see above :)

Eugen Podsypalnikov 14-Aug-12 9:27am View

Deleted

I will try it, see above :)

Eugen Podsypalnikov 14-Aug-12 6:13am View

Sorry, could not understand it from the first reading :) I will improve the solution...

Eugen Podsypalnikov 14-Aug-12 6:12am View

Deleted

Sorry, could not understand it from the first reading :) I will improve the solution...

Eugen Podsypalnikov 14-Aug-12 3:50am View

Nice, without some keys "virtual" as well :)

Eugen Podsypalnikov 13-Aug-12 16:35pm View

Why in c ? :)))))

Eugen Podsypalnikov 18-Jul-12 11:09am View

// then you don't need RPC at all
OK, thank you :)
I did not know about BE-LE relations and even about RPC as technique as well :)

Eugen Podsypalnikov 18-Jul-12 10:30am View

// But the question is about RPC
You are right :)

Eugen Podsypalnikov 18-Jul-12 10:27am View

// reimplement again when changing the structure
Why ? Only sizeof() will be recompiled... In case of an object the same serializing that is used for HDD exchange could do it... :)

Eugen Podsypalnikov 18-Jul-12 9:51am View

// It is unstructured, and it deppends on the implementation
Yes, the serializing code must be known (shared) for both sides,
as it's shown at the Figure 2 in the link above :)

Eugen Podsypalnikov 18-Jul-12 9:46am View

// This most likely first means that you will have to initialize COM
Yes, IStream is COM interface :)

Eugen Podsypalnikov 18-Jul-12 9:34am View

// And marshaller knows nothing about how should be used an unstructured buffer
Why nothing ? The in-/out- code could be shared for both sides... :)

Eugen Podsypalnikov 18-Jul-12 8:05am View

// Anyway, IStream is for serialization, but you need marshalling
IStream pointers can be marshaled by type library marshaling, and can be accessed by C++ clients (see the link above) :)

Eugen Podsypalnikov 17-Jul-12 11:13am View

Yes, it can be equivalent to another loop :)

Eugen Podsypalnikov 17-Jul-12 10:54am View

The copying after reallocation can take a loop as well...
My result is N^2 (+1 - for copying of the return value) :)

Eugen Podsypalnikov 4-Jul-12 15:28pm View

// how its work ?
As OriginalGriff said,
the value will be read always from the address of the variable directly :)

Eugen Podsypalnikov 3-Jul-12 9:14am View

You should debug the code to determinate its error line,
set a breakpoint there and execute it step by step... :)

Eugen Podsypalnikov 3-Jul-12 1:56am View

Please post that error text here :)

Eugen Podsypalnikov 2-Jul-12 9:58am View

// but what about bigger numbers like 3456 0r 4567

Very c00l ! Nearly all digits are the same ! 5ed :D

Eugen Podsypalnikov 2-Jul-12 9:30am View

{ BUTTONINFO* pButtonInfo(new BUTTONINFO);
pButtonInfo->xPos = 12;
//..
pButtonInfo->szText = new TCHAR[80];
_tcscpy_s(pButtonInfo->szText,
_countof(pButtonInfo->szText),
_T("Hello world !"));
pButtonInfo->Length = 120;
//..
::SetWindowLong(hwnd, 0, (LONG) pButtonInfo);
}

case WM_DESTROY:
{
BUTTONINFO* pButtonInfo = (BUTTONINFO*) ::GetWindowLong(hwnd, 0);
delete[] pButtonInfo->szText;
delete pButtonInfo;
//..
}
break;

Eugen Podsypalnikov 1-Jul-12 14:21pm View

Your structure and all its members should be allocated in heap.
When it could be passed to the SetWindowLong(..) function per its pointer.
After this action you can get its pointer per GetWindowLong(..) function.

Eugen Podsypalnikov 22-Jun-12 3:24am View

In the context of your dialog:
bool bDroppedOnSrcBox(GetDlgItem(IDC_SRC_BOX)); // IDC_SRC_BOX is the resource ID of the box
bool bDroppedOnTrgBox(GetDlgItem(IDC_TRG_BOX)); // IDC_TRG_BOX is the resource ID of the box

Eugen Podsypalnikov 21-Jun-12 10:18am View

Please do not try to call it (free(..)), just comment it out :)

Eugen Podsypalnikov 21-Jun-12 6:54am View

// In MSDN it is written that, this function returns 0 on failure.

Please try it :) :
assert(12 == atoi("12 months"));

Eugen Podsypalnikov 21-Jun-12 2:11am View

Take it :)

Eugen Podsypalnikov 21-Jun-12 1:55am View

At your "dropped" event, call this test for the both boxes (pass their pointers). The function will return true, if a testing box is under the mouse :)

Eugen Podsypalnikov 21-Jun-12 1:50am View

Hi Maximilien ! Yes, you will need IMPLEMENT_SERIAL instead, if your button has own persistent members and must be serialized dynamically by the framework, of course ! :)

Eugen Podsypalnikov 20-Jun-12 10:08am View

Deleted

Please post your class declaration and the call of ReplaceButton here :)

Eugen Podsypalnikov 19-Apr-12 10:35am View

Thanks,
I agree with you :)

Eugen Podsypalnikov 19-Apr-12 10:25am View

Thank you ! :)

Eugen Podsypalnikov 19-Apr-12 10:05am View

Thank you,
but I can not see any copy operation here, for <char> as well... :)

Eugen Podsypalnikov 19-Apr-12 2:45am View

Thanks, 5-ed :)

Eugen Podsypalnikov 2-Mar-12 1:57am View

For example:
(ACE)99, (P)1, (DF)99, (P)1, (ACE)1000, ...
could meen:
ACE-chord plays for 99 milliseconds, then
Pause for 1 millisecond, then
DF-chord for 99 milliseconds, then
Pause for 1 millisecond, then
ACE-chord for 1 second, then ...

Eugen Podsypalnikov 21-Feb-12 9:56am View

OK :) Then please try it:
{
//..
CCreateContext cContext;
cContext.m_pNewViewClass = RUNTIME_CLASS(CConfigViewPlugin);
m_cYourSplitter.CreateView(.. , &cContext);
//..
}

Eugen Podsypalnikov 9-Feb-12 8:10am View

Thank you,

but 3*(fWB * fWB) = (fR * fR + fG * fG + fB * fB) , isn't it ? :)

Eugen Podsypalnikov 9-Feb-12 7:47am View

So the casting has just "extend" (incorrect)
the boundary of a smaller object by the data member of the bigger type :)

There could lie another application data,
but your test had a lucky end (not a low) :)

When you try to extend the smaller object by a float member as well,
the "casted" set method will set it (not the original member) :)

Eugen Podsypalnikov 9-Feb-12 7:06am View

Hm... It was just an abstract example...

We could call it
as we want to add a sound file link to the collection...

Eugen Podsypalnikov 9-Feb-12 4:20am View

static CStringArray arSoundFiles;

void CYourCode::AddSound(const CString& cszSoundFile)
{
arSoundFiles.Add(cszSoundFile); // :)
}

const CString& CYourCode::GetLastSound()
{
static CString cszNone;
if (arSoundFiles.GetCount()) {
return arSoundFiles.GetAt(arSoundFiles.GetCount() -1);
}
return cszNone;
}

Eugen Podsypalnikov 9-Feb-12 4:09am View

Why not to hold a link to a sound file (instead) ? :)

Eugen Podsypalnikov 30-Nov-11 9:53am View

// How can I make my C programm run faster and use less memory?
Is there a function in your C programm
that does make you not sure ? :)

Eugen Podsypalnikov 23-Nov-11 6:25am View

Hm... A static instance in the server code could call the all connected clients by its destructor... :)

Eugen Podsypalnikov 7-Nov-11 2:29am View

I would see the labor task (research only, probably your 1 up to 10 positions) to find out the k and d for the linear law: y = k*x + d ...where y is needed interpetation by the second program of the values x from the first program. Once they are found (in a labor(research) workaround), they could be used in the hard code of the second program (probably your 11 up to 13 positions) :)

Eugen Podsypalnikov 7-Nov-11 2:26am View

Deleted

I would see the labor task (research only, probably your 1 up to 10 positions) to find out the k and d for the linear law:
y = k*x + d
...where y is needed interpetation by the second program of the values x from the first program.
Once they are found (in a labor(research) workaround), they could be used in the hard code of the second program (probably your 11 up to 13 positions) :)

Eugen Podsypalnikov 9-Sep-11 8:33am View

Please, try to invalidate the bottom by resizing, overlapping, or minimizing/reparing :)
Is it then visible ?

Eugen Podsypalnikov 9-Sep-11 8:20am View

Deleted

Then try it :) :
if( cScrollFlag == 'Y' ) // if scrolling required //
{
TextOut( hdc, cxChar, iVertPos - cyChar, szBuffer, lstrlen( szBuffer ) );
}

Eugen Podsypalnikov 9-Sep-11 7:58am View

Have you already tried to scroll without the clipping (last two parameters =NULL) ? :)

Eugen Podsypalnikov 5-Sep-11 11:27am View

You can try the Spy++ tool (inside the MS VS set) to observe the styles and the received messages (generated by the mouse, which - maybe - could be simulated by your code as well) of the child... :)

Eugen Podsypalnikov 5-Sep-11 6:41am View

An int is not a CObject yet... :)
But you can control for example the VARIANTs in your list
or the pointers of CMyGenericValueBase. Then you could dirive some real values, for example CMyString : public CMyGenericValueBase and others... :)

Eugen Podsypalnikov 5-Sep-11 6:34am View

...it is also comfortable to use a type predefinition, for example :) :
typedef CList<int, int> CIntEntries;
class CMyRegs : private CIntEntries
{
..
void Reset() {
//.. some leaving actions
//..
CIntEntries::RemoveAll();
}
};

Eugen Podsypalnikov 26-Aug-11 5:35am View

Are you
using the LoadLibrary(..) function
or linking the DLL directly ? :)

Eugen Podsypalnikov 26-Aug-11 3:41am View

Not sure, sory...
I think, only the drawing in the context of OnDraw(..) will "repare" the tracker after the scrolling :)

Eugen Podsypalnikov 26-Aug-11 2:52am View

Would be QT thinkable for you ? :)
( http://www.youtube.com/watch?v=jfHM4pQSqt4 )

Eugen Podsypalnikov 26-Aug-11 1:43am View

How is the Dll "attached", please ? :)

Eugen Podsypalnikov 24-Aug-11 3:37am View

Hi Laxmikant Yadav,
the remaining bytes (in the processor registers) are not using in the second case.
This behavior (machine code) is the result of a (32- or 64-bit) compiler and its appropriate linker.
So a 64-processor can receive the 32- or 64-bit instructions (commands).
A processor is a HW part.
An OS or an executable file are SW.
The SW can be build by 32- or 64-compiler/linker.
It is also possible to provide (send) 32-bit instructions of an executable by a 64-bit OS on (to) a 64-bit processor. This process (of that executable) will take just a "shorter cord" for its (memory) space addressing :)

Eugen Podsypalnikov 29-Jul-11 4:45am View

Deleted

Please look for "ES_READONLY" and "WS_DISABLED" in your *.rc file and remove them just for test :)

Eugen Podsypalnikov 28-Jul-11 7:21am View

0. Probably you should increment i1 (i1++) instead i (i++)
1. See also at my modified first function: the count will be "serialized" as well
:)

Eugen Podsypalnikov 28-Jul-11 3:17am View

Please see above :)

Eugen Podsypalnikov 19-Jul-11 4:29am View

// call to force a specific doc/view to undock from the tab bar
I could find the versa-way only: a dock-pane can fly to the client area and back :)
(see the caption context menu at the "file view dock-pane" in the general wizard's example)

Eugen Podsypalnikov 13-Jul-11 6:41am View

0. Thanks !
1. Try nSize *= 2 before its passing.
2. Please provide the returned value of wcstombs()

Eugen Podsypalnikov 13-Jul-11 6:35am View

0. A "life" message from a client with its ID - once per one minute
and a "death" timeout for two - as a communication low ? :)

1. Try to hold the client's communication object as global static object
with the CreateInstance(..) in its constructor ant Release(..) in its destructor :)

Eugen Podsypalnikov 13-Jul-11 5:57am View

What is a "connection" in your case,
why does it need any maintenance ? :)

Eugen Podsypalnikov 13-Jul-11 5:00am View

Please provide your test code and the returned value :)

Eugen Podsypalnikov 6-Jul-11 10:48am View

Try to find (or generate) an *.?dl (in my case an *.idl) file :)

Eugen Podsypalnikov 6-Jul-11 10:29am View

Yes, I see it so (too) :) The COM+ is an extension (not a replacement) for the COM...

Eugen Podsypalnikov 6-Jul-11 10:23am View

static const CLSID CLSID_MYLAYER = {
0x9912EF9F, 0x0FF0, 0x42c1, { 0xAE, 0x6B, 0x74, 0x0C, 0xC8, 0x3D, 0x6D, 0xA8 } };

static const IID IID_MYLAYER = {
0x03A1EE13, 0x2CB6, 0x4b79, { 0xA4, 0xF6, 0x40, 0x65, 0xBB, 0xBE, 0x9B, 0x2B } };

void CSomeUser::SomeStart(..)
{
// IMPILayer* m_pIMPILayer = NULL; // Interface' pointer

HRESULT hr = CoCreateInstance(CLSID_MYLAYER,
NULL,
CLSCTX_INPROC_SERVER,
IID_MYLAYER,
(void**) &m_pIMPILayer);
if (S_OK == hr && m_pIMPILayer) {
m_pIMPILayer->SomeMethodeCall(..)
...
}
}

Eugen Podsypalnikov 6-Jul-11 10:10am View

Probably I would use some interface(s) of an COM+ object in the same (COM) way... :)

Eugen Podsypalnikov 6-Jul-11 9:54am View

In general case
you could begin at CoCreateInstance(..) function :)
See also: http://www.jj54.com/73850-CoCreateInstance-Fails-On-COM-43-Object-With-Object-Construction-Enabled-And-No-Constructor-String-Present/

Eugen Podsypalnikov 6-Jul-11 9:39am View

Then try to modify your free-function to the following form :) :

DYNLIB_API void fn1(int **p)
{
if (p) {
delete *p;
*p = NULL;
}
}

...and its call :

int main(int argc, char* argv[])
{
int *p=NULL;
p=new int(100);
fn1(&p);
return 0;
}

Eugen Podsypalnikov 6-Jul-11 9:26am View

Already tried to use delete and not free ? :)

Eugen Podsypalnikov 6-Jul-11 9:17am View

I would put the resul in rbbi.hbmBack at stake... :)

Eugen Podsypalnikov 10-May-11 6:41am View

It is a variable, as your array's const-reference :)
It could be written also:
const cArrItems& items = doc->GetArray();

Eugen Podsypalnikov 10-May-11 5:00am View

Just modify your usage like this :) :
const cArrItems& items(doc->GetArray());
Please note, that the const reference does provide the "const" access only.
So you will be not able to call the not const functions on such refernce...

Eugen Podsypalnikov 10-May-11 4:52am View

Please post the both lines now:
- declaration of your function
- usage of your function
We will try to compare them :)

Eugen Podsypalnikov 10-May-11 4:41am View

Please try to use the last posted function :)

Eugen Podsypalnikov 10-May-11 4:36am View

Please post the declaration of the GetItems() function here :)

Eugen Podsypalnikov 10-May-11 4:27am View

Please try it :) :
CTypedPtrArray <cobarray,coverlayitems*>& myArrayRef = doc->GetItems();

Eugen Podsypalnikov 10-May-11 3:45am View

Please post the code line of the error here :)

Eugen Podsypalnikov 8-Apr-11 6:43am View

You could try to implement your derived function :) :
void CDerivedOne::DoDataExchange(CDataExchange* pDX)
{
CSampleDlg::DoDataExchange(pDX);
}

Eugen Podsypalnikov 25-Mar-11 12:34pm View

1) Then (0 == buff[0]), in other words the buffer's content is empty
2) You should try to subclass the edit box by your own control with its own reaction for WM_GETTEXT
3) A control does receive the WM_GETTEXT (LPARAM=buffer, WPARAM=its length) to fill the buffer out (it is an answer, so the customer of the text has no answer when the reaction does nothing (!))
4) ...says LPTSTR (not LPSTR)

Eugen Podsypalnikov 25-Mar-11 12:25pm View

Deleted

1) Then (0 == buff[0]), in other words the buffer's content is empty
2) You should try to subclass the edit box by your own control with its own reaction for WM_GETTEXT
3) A control does receive the WM_GETTEXT (LPARAM=buffer, WPARAM=its length) to fill the buffer out (it is an answer, so the customer of the text has no answer when the reaction does nothing (!))
4) ...says LPTSTR (not LPSTR)

Eugen Podsypalnikov 16-Feb-11 1:38am View

Yes, you could also pack your edited fields into a class, for example:
class CDataEntry
{
CString m_cszFieldA,
m_cszFieldB;
int m_iFieldC,
m_iFieldD;
..
};

...and then implement void CYourView::UpdateView(CDataEntry& cData); as well :)

The (any) DB-exchange (open/save/update/...) could be managed by the view only.

Sorry, I have not understood why you are saying about a DB-type switch (1),

and my DB-skills (speed/comfort/safety/...) are very poor (2)... :)

Eugen Podsypalnikov 14-Feb-11 1:58am View

I do not know it yet, sorry.
Anyway I would try to exclude any "unneeded" rowsets concurrency...
The edited set by the dialog would be ready for the view directly, without any DB exchange :)

Eugen Podsypalnikov 26-Jan-11 17:33pm View

Yes, it can :)
I was just attracted to the same "ready" type of the member and its get function... :)

Eugen Podsypalnikov 15-Dec-10 6:41am View

Please check the redrawing
under the overlapping by a normal cmd-window shown by ist own (not a child) process... :)

(I would remove also the last two lines 'CloseHandle(..)' :) )

Eugen Podsypalnikov 15-Dec-10 3:14am View

Understood. So you are looking for a library that provides the PNG-bits access, aren't you ? :)

Eugen Podsypalnikov 15-Dec-10 2:34am View

Have you already tried it ? :) :
{
CImage cImage;
if (S_OK == cImage.Load(_T("C:\\transp.png"))) {
void* pBits = cImage.GetBits();
if (pBits) {
//..
}
}

Eugen Podsypalnikov 13-Dec-10 15:26pm View

Please see the steps 6 and 7 here:
http://www.codeproject.com/KB/DLL/XDllPt1.aspx :)

Eugen Podsypalnikov 13-Dec-10 3:00am View

A DLL does cover the executing linkage,
during a LIB - the building linkage (your fault case) :)

Eugen Podsypalnikov 22-Nov-10 16:43pm View

The both sequences below are re-using the constructor code only not an object :) :
{
CArchive* pA(new CArchive(&f, CArchive::store));
...
delete pA;
pA = new CArchive(&f, CArchive::load);
...
delete pA;
}

bool CYourArray::ExchangeData(bool bLoad, ...)
{
...
CArchive ar(&f, bLoad ? CArchive::load : CArchive::store);
...
}

Eugen Podsypalnikov 4-Oct-10 1:45am View

Hm... I thought, you know them :)
- Please validate your received hWnd and hToolbar by Spy.
- Try it (yours) also with the index 0 (the separators have their positions as well).
- Please observe the returned value by SendMessage(.., TB_GETBUTTON, ..)
- Try also the message TB_GETBUTTONINFO (set the field cbSize before you will pass a TBBUTTONINFO)
- Usually, the commands handling frame of a bar is its parent or the main window
- Does the window "Test window" lie in an external process (started by the same rights) ?

Eugen Podsypalnikov 30-Sep-10 8:35am View

Thank you, Sauro ! :)

Eugen Podsypalnikov 30-Sep-10 5:16am View

// newbie like me

http://www.youtube.com/watch?v=QgbntWU7pG8 :)

Eugen Podsypalnikov 28-Sep-10 12:20pm View

Try also to transform your application to UNICODE,
it will be more comfortable :)

Eugen Podsypalnikov 28-Sep-10 12:14pm View

0. OS->NonUnicodeLaguage=Russian
1. In your .rc file, opened as text:
...
/////////////////////////////////////////////////////////////////////////////
// Russian resources
#if !defined(AFX_RESOURCE_DLL) || defined(AFX_TARG_RUS)
#ifdef _WIN32
LANGUAGE LANG_RUSSIAN, SUBLANG_DEFAULT
#pragma code_page(1251)
#endif //_WIN32

IDD_RUSTEST_DIALOG DIALOGEX 0, 0, 320, 200
STYLE DS_SETFONT | DS_MODALFRAME | DS_FIXEDSYS | WS_POPUP | WS_VISIBLE | WS_CAPTION
EXSTYLE WS_EX_APPWINDOW
CAPTION "RusTest"
FONT 8, "MS Shell Dlg", 0, 0, 0x1
BEGIN
DEFPUSHBUTTON "OK",IDOK,209,179,50,14
PUSHBUTTON "Cancel",IDCANCEL,263,179,50,14
LTEXT "12,78 M/C",IDC_TEXT,54,53,57,24
END

STRINGTABLE
BEGIN
IDS_STRING129 "%.02f M/C"
END

...
2. in your .cpp :
GetDlgItem(IDC_TEXT)->SetWindowText("34,76 M/C");
or :
char szBuffer[MAX_PATH];
::LoadStringA(NULL, IDS_STRING129, szBuffer, MAX_PATH);
CString cszText;
cszText.Format(szBuffer, 12.98);
GetDlgItem(IDC_TEXT)->SetWindowText(cszText);
:)

Eugen Podsypalnikov 28-Sep-10 9:15am View

The "text box" seems to be coded in CP1252 (not in 1251)...
Can you put _this_ string at a button there (at the surface) properly ? :)

Eugen Podsypalnikov 28-Sep-10 5:54am View

Be also sure,
that your CMDIChildWnd::Create(..) is called after the ExeApp::MainFrame was allocated :)

Eugen Podsypalnikov 28-Sep-10 5:06am View

...or try your model without any AFX_MANAGE_STATE, please :)

Eugen Podsypalnikov 28-Sep-10 4:58am View

Please see the "second part" above... :)

Eugen Podsypalnikov 28-Sep-10 4:02am View

Very good :)
Why do you want to use any managers (and AFX_MANAGE_STATE too)
and not just export/import your classes by DLL/EXE ? :)

Eugen Podsypalnikov 28-Sep-10 3:56am View

But your combination EXE&DLL will need a main frame,
is there any frame at the exe-app ?.. :)

Eugen Podsypalnikov 28-Sep-10 3:55am View

Deleted

But your combination EXE&DLL will need a main frame,
is there any frame at the exe-app ?.. :)

Eugen Podsypalnikov 25-Sep-10 7:35am View

1) It would be OK, when the validity of the pointers is granted for the whole time their usage.
2)
class CCsTrendPoint : public CObject // a point's data
{
..
}

class CDisplayItems : public CObList // collection of CCsTrendPoint* for a pen, second dim
{
..
};

class CDisplayData : public CObArray // collection of CDisplayItems* indexed by pens, first dim
{
..
};

:)

Eugen Podsypalnikov 24-Sep-10 10:16am View

An operator 'new' returns a pointer of the constructed typed object allocated in the heap. Such pointers are valid until they will not be deleted by the operator 'delete'.
Othersides, you can get any typed pointer by 'ampersand', for example: { int i = 2; int* pointer = &i; }. Such pointers are valid only for the live-period of their "master-variables".
So you can use '&trendpoint', only when you are sure the trendpoint's live period is long enough... :)

Eugen Podsypalnikov 24-Sep-10 1:44am View

What is your simulation context ?
At a valid window handle,
you could try to post the messages WM_MOUSEWHEEL, WM_RBUTTONDOWN, ... :)

Eugen Podsypalnikov 18-Sep-10 15:06pm View

// Except you never see y...
...and its address as well... :)

Eugen Podsypalnikov 18-Sep-10 14:55pm View

It's correctly, the "cout << x" is "precompilled",
but the debugger shows the modified value of x... :)
See also:
const int iSize(3);
int iSize2(3);
int ai[iSize]; // OK
int ai2[iSize2]; // error C2057: expected constant expression

Eugen Podsypalnikov 13-Sep-10 8:53am View

Please post your code here :)

Eugen Podsypalnikov 9-Sep-10 17:06pm View

The address of test[5] does not lie at the integer-boundary,
the testptr can not be used to dereference its value directly on some systems :)

Eugen Podsypalnikov 9-Sep-10 9:11am View

1. I have not used any term of a "pointer address", but one of a "pointer" :)
2. You can not "correct" an data address, but you can copy the addressed data :) :
{
char* szTestBuffer = "content";
WORD* pwData = (WORD*) &szTestBuffer[1]; // an odd address !
#ifdef _X86_
(*pwData)++; // it is OK on x86
#elif defined(MIPS)
// (*pwData)++; // it is not OK on MIPS
WORD wTemp(0); // an even address
memcpy(&dwTemp, pwData, sizeof(WORD));
wTemp++;
memcpy(pwData, &dwTemp, sizeof(WORD));
#endif
ASSERT('p' == szTestBuffer[1]); // :)
}

Comments by Eugen Podsypalnikov (Top 200 by date)