>>>>Whenever I attempt to change the Zoom setting when Previewing 2 particular Reports with SET REPORTBEHAVIOR 90, see a brief flash of the re-scaled Preview followed by the Development Environment crashing. Other Reports work correctly and this does not occur in the compiled versions running the same 2 reports.
>>>>
>>>>I looked at the Windows Error Log and it appears to be an issue with the NTDLL.DLL file in the Windows/System32 folder. Stumped as to why this is happening for these 2 reports.
>>>
>>>You could open the reports as tables and remove any instances of CHR( 0 ) that may be present in character or memo fields.
>>
>>Hi Al
>>The problem I see with this, the frx itself could rarely be the problem. The report as such is run. This is SET REPORTBEHAVIOR 90, and I think Marshal will use a standard approach. So all pages are stored to memory before calling the preview.
>>The first page is rendered from memory stored GDI+ image. What fails is listeners OutPutPage re-rendering the same image from memory to the canvas in a different scaling. (it is unproven if rendering the very same page on same scale will work.) It's more like the complexity / size of the GDI+ image(s).
>>I have now idea how complex the stored GDI+ information is, if it holds layer like stuff like it is used while creating the picture, or if only the resulting bitmap exists.
>
>I suggested some system level checks and tests in a reply in the other thread, I don't know if they were tested, or the results (e.g. SFC).
>
>Speaking of GDI+, I remember Rich Simpson (developer of Mind's Eye Report Engine [MERE]) discussing an issue with it - some sort of capacity or complexity limitation. I did find this discussion: https://social.msdn.microsoft.com/Forums/en-US/c34f92fb-1fe0-4d70-baf8-31d9b0d6316f/reportlistener?forum=visualfoxprogeneral . Anyone else remember anything about this? The link talks about reports with thousands of pages of output but ISTR Rich was talking about complex reports using a lot of GDI resources (?)
>
>Update: you posted Thread#01476778
>
>Update 2: Rich's original post Message#01024004

Hi Al,

Thanks for pointing this out. I will check my listener for Rich's code. :)

Nice to see a thread by my alter ego. Time is passing by. Yeah, this is the feeling I have with this issue - only it is a different way of odd. I checked the code of the previewer - and there is not much more then creating a shape of the size needed and call OutputPage to draw the image to it. The zoom is just the size of the canvas. I think the "original" size is maxed to the preview form, so it is zoomed anyway.
Why the second call fails is the main question. It does the same as the first call, except a different size of the shape.