[vtkusers] [VTK4.x + Python + wxPhyton] wxVTKRenderWindow maturity

[Prabhu Ramachandran]

>>>>> "DG" == David Gobbi <dgobbi at irus.rri.ca> writes:

    DG> Hi Eric, Thanks for the tests.  I've patched the VTK 4.0
    DG> wxVTKRenderWindow so that it should work fine in a notebook or
    DG> a splitter under Linux.  It was the SetWindowInfo() before

Thanks Eric and David!

    DG> resize & paint events are handled, and many other things 2)
    DG> the wxGTK code base isn't very stable, huge changes seem to
    DG> happen in the code with each minor release which makes it hard
    DG> to work around any wxGTK bugs 3) the SetWindowInfo() trick

I'll second that.  Maybe using wxPython 2.2.6 (which is the stable
release) is a better option.  Under linux it seems to work ok.

    DG> that connects VTK-Python and wxPython is an ugly hack that
    DG> only works if window creation and rendering occur in the
    DG> correct order (this can very hard to guarantee).

    DG> The main issue is simply that wxWindows is not as mature as
    DG> Tk, and there's not much that we can do about that.

I'm not sure its wxWindows that has the trouble.  AFAIK its the GTK
port that really has the trouble.  I dont use wxWindows on win32 so I
wouldnt know how bad it is there but so far, based on all the reports
I've heard it seems pretty well behaved under win32.  I'm hoping that
GTK 2.0 will fix some of the wxGTK related problems.  IIRC there are
plans for a well behaved double buffered widget with GTK 2.x.

Anyway, it does look like Davids widget is the most stable for the
present.

prabhu