[vtkusers] vtk4.2 + Python2.3 + win = works,but repr(<vtkobject>) crashes

[Andy Cedilnik]

Hi Frank,

I am looking at similar behavior on Valhalla:
http://www.vtk.org/Testing/Sites/valhalla.kitware/Windows-bcc32/20030922-0300-Nightly/Results/__Common_Testing_Python_PythonSmokePython-image.html

Looks like when it tries to print on the object it breaks. I think the
borland python combination is not there yet. Thanks for looking into it
though. If you figure it out, let me know.

Thanks.

			Andy

On Mon, 2003-09-22 at 10:32, Frank Conradie wrote:
> Actually I have no idea where the part of the stack trace into the Borland
> stuff comes from, but I can tell you what I did. I re-built vtk with
> debugging on (using the free BCC32 compiler), then used gdb through cygwin
> to run python.exe. I also have dev-c++ and it's Windows-only gdb also works.
> 
> By the way, I've definitely tracked down the problem I'm having - the Print
> method of vtkBaseObject is never entered, as if it's pointing to some random
> code location. For the time being I've simply copied the code from repr into
> str, so I now get the same output for repr and str of vtk objects.
> 
> The reason why this is important is because I use PyCrust with my app, and
> it crashes when it tries to str() a vtk object.