typedef struct DOUBLE_MATRIX { double **ppdDoubleMatrix; MIINT32 iRows; MIINT32 iColumns; }DOUBLE_MATRIX; static DOUBLE_MATRIX *New(MIINT32 iRows, MIINT32 iColumns) { DOUBLE_MATRIX *pxThis = NULL; pxThis = calloc(1, sizeof(mxSystemDOUBLE_MATRIX)); if (pxThis && iRows > 0 && iColumns > 0) { MIINT32 iRowsCounter = 0; pxThis->iRows = iRows; pxThis->iColumns = iColumns; pxThis->ppdDoubleMatrix = (double **) calloc(iRows, sizeof(double *)); for (iRowsCounter = 0; iRowsCounter < iRows; iRowsCounter++) { pxThis->ppdDoubleMatrix[iRowsCounter] = (double *) calloc(iColumns, sizeof(double)); } } return pxThis; } Papa while (TRUE) Papa.WillLove ( Bebe ) ;

If your app uses MFC, you can use the DEBUG_NEW macro, and memory leaks will be outputted to your Output windows in Visual Studio. This even tells you the line of code where the object was "new"ed. At top of all .cpp files: #ifdef _DEBUG #define new DEBUG_NEW #undef THIS_FILE static char THIS_FILE[] = __FILE__; #endif

you're right, sorry for this stupid mistake and thanks for your help

I don't know exactly why, but there is an obvious place to start. In debug mode, newly allocated memory is filled with CC in each byte. In Release, it is filled with 00 in each byte. So you may have a variable that you have not explicitly initialised, but Release sets it to 0 / NULL. Maybe you've missed a DDX_Control? You'll need to put breakpoints in your dialogs constructor, OnInitDialog, etc and see where it dies in more detail. If you are really stuck, email me your project, and I'll have a look see. Iain.

sorry to be late to answer you. Just to tell you that your solution works, thanks a lot

then use #import "excel8.olb" And you will have to carry excel8.olb as a part of your build. I am not sure if it is legal though, you might have to have Excel8 on your build machine, check the user agreement.

When you call DeleteHistoNorm() you check to see if pHistoDBNorm is NULL or not. The thing is that when you declare double *pHistoDBNorm, just to be safe in your class constructor you should set pHistoDBNorm = NULL. I don't know if you are using this pointer somewhere else in your source code without initializing it and then all of a sudden you try to delete it. none

I try to follow first the allocation describe by the function called AllocComptStartEndMin, and you know as usual when you do the allocation of memory for a pointer of "double" for example, p=new double[X]; when you try to see what are the value, for each i from 0 to to X-1 of p[i], (before any affectation), in the debug window, you have in this window: p[i] in the column "Name" and -6.2774385...e+066 in "Value" column. So when i execute step by step AllocComptStartEndMin, it works even if i>120. So, why when i want to affect value to the pointers i can't affect value for the first pointer when i>120 ( and what is very strange is this value:120, because in my sample RowSize =866 and ColSize=255!)

The sample detailed in Q186120 works perfectly well. I have just tested it with Excel 97. She's so dirty, she threw a boomerang and it wouldn't even come back.

I think your only settings the first value. Maybe the approch with the array is not the easiest way. If done this (where is the code???) with writing to a cell and the changing the position, it worked well. Tip: if you write in a csv (comma separated file) with ";" as delimiter excel can open it easily. If you need you can also deploy a excel macro to some strange things. If also read in CP an article about writing to Excel via ODBC Functions, it is working.

if you want to fill cells with a for loop, how we could do that? for(i=1;i<= Bound;i++) { . . } where the term "Bound" could be different for different test( it could be equal to 20 or 10 or another number) If the size of your set of data is always the same, i know i could do it: range.GetRange(COleVariant("A1")),COleVariant("B20")) range.SetValue(COleVariant("2")) and you will see in each cell of the area A1:B20, the number "2". So, know if the width and/or the height change dynamically, how could i fill the set of cells i need? :confused:

Please, build a debug version and run it under the debugger. Since it faults in the runtime library, it should be obvious why it fails when run in the debugger. Common reasons for access violations are str...() functions with null pointers or invalid strings (source or target, length wrong), memcpy/memmove with invalid source or target locations. Also overwrites of allocated memory or stack may cause this error. int x=1, y=5; x^=y^=x^=y; // whats the content of x and y now? ClickHereForHelp();

you were right, in fact i declare a wrong memory space for only one pointer that have to be deleted by the class destructor, but there was a problem as when i put value to this pointer, i did it in a for loop, in which i would like to put more that ten times memory spacxe than i have allocated. thanks a lot gerald

in fact in my code i change 2 things: for(i=1;i<=pow(2,m_dlgAsset)-1;i++) { k=m_dlgAsset-1; while(*(pTabBinPort+k)==1) { *(pTabBinPort+k)=0; Table.ConvertBinToNum(k,pTabBinPort,pTabNumPort); if(k!=0) k--; } *(pTabBinPort+k)=1; Table.ConvertBinToNum(k,pTabBinPort,pTabNumPort); *(pTabVar+i)=Table.CalcVar(pTabNumPort,m_dlgAsset); } you could see in the for statement that i going from 1 to pow(2,m_dlgAsset)-1 now and i've put a test in the while loop just for the case of k=0, so that we couldn't be in the situation of k=-1. But now i have another problem, when i test my application with m_dlgSector=30 and m_dlgAsset=15, at the end of the function i have another message error which says: Unhandled exception in CalcRiskport.exe (MFC42.DLL):0xC0000005:Access Violation. It's happened at this line: strValue.Format("%.5f",m_dlgValue); You could see the whole code in the precedent post.

Gérald Mercet wrote: so you could see the following code C++ and VBA, for the moment it works, but perhaps you could think that it needs modifications?!? Because you are declaring the function in this way.. Public Declare Function Maximini Lib "D:\essai2\Debug\essai2.dll" Alias "_Maximini@8" (ByVal m1 As Long, ByVal m2 As Long) As Long If you use a .def file you will can declare the function in the following way: Public Declare Function Maximini Lib "D:\essai2\Debug\essai2.dll" (ByVal m1 As Long, ByVal m2 As Long) As Long If your code works fine, great, but with my way you don't need to look for the internal name in the dll. That is the unique difference... Gérald Mercet wrote: as you ask me yesterday, sorry to be late, but when you post it i believe that it wasn't the same hour in my office than in yours! Not problem: I live in South America, And we have several hours of diffence, and plus I'm in vacation, only I am in CP, because I can't live without CP.:(( Regards... :-D Carlos Antollini. Sonork ID 100.10529 cantollini

You can call a DOS program using the VBA Shell statement in sync or async. An application can be called using the VBA AppActivate statement. So you can implement your automation functions in VC++ apps. Hope it helps... Atul Sonork ID : 100.13714

Hi, A very good doc is: "Automating Microsoft Office 97 and Office 2000" ( written by Lori Turner , 17 February 2000), it´s free you can search for it in MSDN in the web or the cd. You can find a lot of help in MSDN, just search for this keywords: kbAutomation kbExcel kbMFC sample But anyway you are right most of the samples are made for VBasic, but it´s easy to transform it to VC code. About docs explaining more on Word Automation: http://www.codeproject.com/com/xoffice.asp and in codeguru ( www.codeguru.com) you can find some help as well. Some more samples you can find in the Kruglinski, Shepher, Wingo book "Inside Visual C++". And one more tip... sometimes the easiest way to find how to implement something for Excel or Word is just to record a macro and see the code generated in VBA. Good Luck Braulio