BKor

Hello everybody, I installed Rad Studio Athens ( with some graphic 3rd parties) and now I am trying to open some projects that I have done in Alexandria ( Rad Studio11). Unfortunatelly I am not succesfull - (access violation at address 00876830 in module "Project1.exe" (offset 6830). Read of address 000000000.) I am working with C++ Builder and before I had no problems with migrating projects from previous versions of Rad Studio. Any advice that could help me running older projects?

Thank you both for your help, reinstalling the gr32 and using the actual constructor&destructor did help. Just one question about "This isn't necessary in this particular case; The TImage32 owns the layers and will automatically destroy them." May I delete pLayer[j] (TImage32) in Form destructor anyway (although not necessary as you say)? Thanks again.

Thanks a lot. I will try it in the evening.

If it is not too much, can I kindly ask you if you can show me how to ( in a simple example?), use the Form's actual constructor/destructor in c++ builder?

Yes, I mean running the code. I will remove the 3rd party stuff and see what happens. And then also reinstall the 3rd parties. Thx...

I was wrong. Rewriting the code in RadStudio 12 did not work. When compiling it gives: Access violation at address 6CE9D636 in module 'rtl290.bpl' (offset 11D636). Read at address 006C005E. When debugging it stops at loop, as before. I would like to delete some of my previous messages (ones with tons of images), is this possible?

Yes, I did. I have a feeling that if I would newly rewrite the same code in Athens - that it would work normally (as it does in Alexandria). I will try that tomorrow and let you know...

I have tried both suggestions, but that did not make any difference. The debugger again stops with a blue arrow at the loop line. If I press F8 to go to the next line of code - a blue arrow goes to the end of void __fastcall TForm1::FormCreate(TObject *Sender) - as before. Thx!

I was refused by "Clean Talk"... I will try to send code as .txt in attached file.... MelanderCode.txt

This morning I tried to send the code (via code block) and was refused to submit as potential spam.... I will try to send you this message and afterwards another one with code. In case I will not be allowed to send the code via code block - is there any other way for sending the code? I am sending the header and upper part of the code. I hope it helps somehow. pLayer is part of TImage32 ( graphics32 , 3rd party), defined as array, and c is defined as const int c=9; Pressing F8 does not make any difference. I wonder why the debugger stops with a blue arrow anyway. The same code is compiled and runs without problems in Alexandria. Thx. (also for pointing out the code block ...)

If I press Shift F7 (trace to next source line) it throws me to the end of "void __fastcall TForm1::FormCreate(TObject *Sender)" and the j stays unchanged. This happens even If I place breakpoints in the code. Thx for any advice.

I am sending the CPU window after debugging pointing to ntdll.dll

I tried to migrate to Rad Studio 12 Athens another older project from Rad Studio 11. I run it with debugger and it shows: 1) still within c++ builder window: 2) then continues to pop up c++ builder window with: 3) After I close all pop up windows it shows in application window another series of alerts: Access violation remains in first case in module rtl290.bpl and then in module Project1.exe I could not monitor anything with debugger, because closing the alerts led me to the application with another series of alerts.... Hmmm, thx for any advice

I tried with debugging, it looks like there is problem with for loop: for (int j=0; j<c; j++) The initial value of j after running with debugging is 12 (j=12, see the image) , although it should start the loop with value j=0. There seems to be another problem with c, which shows no value at all. j and c are defined as follows: TForm1 *Form1; const int c=9; TBitmapLayer* pLayer[c]; int i,j=0; int r; bool slip[c]={ false }; bool bing[c]={ false }; bool bong[c]={ false }; int alpha[c]={ 0 }; ..... If I replace in the loop c with cc ( which is defined solely as int (int cc=9 ; ), not as const int (const int c=9;), then the value of cc shows as expected, that is 9. I might be totally wrong, never worked with a debugger before. I attached the image of debugging window below. Is it possible that some kind of "compile settings", which are different from previous version of RAD Studio make this problem? Thanks for any advice.

I am not used to work with debugger or stack of called methods. I will try it anyway. Is it of any help if ( in case of not able to interpret debugger window...) I make a screenshot of debugger window and post it?