[Paraview] endUndoSet called without a beginUndoSet.

Robert Maynard robertjmaynard at gmail.com
Tue May 5 14:57:53 EDT 2009


Odly enough I just finished fixing this bug today, with a patch for the 
visibility issue at http://www.vtk.org/Bug/view.php?id=8958

The locking bug it self can be fixed by simply changing one line.

Index: C:/Work/ParaViewGeo/VTK/Widgets/vtkBorderWidget.cxx
===================================================================
--- C:/Work/ParaViewGeo/VTK/Widgets/vtkBorderWidget.cxx    (revision 3953)
+++ C:/Work/ParaViewGeo/VTK/Widgets/vtkBorderWidget.cxx    (revision 3954)
@@ -275,7 +275,9 @@
   reinterpret_cast<vtkBorderRepresentation*>(self->WidgetRep)->MovingOff();
 
   // stop adjusting
-  self->EventCallbackCommand->SetAbortFlag(1);
+    //by setting the abort flag to 0 we allow the end interaction event 
to be
+    //caught be other observers so that mouse buttons don't get stuck 
being down.
+  self->EventCallbackCommand->SetAbortFlag(0);
   self->EndInteraction();
   self->InvokeEvent(vtkCommand::EndInteractionEvent,NULL);
 }

Dominik Szczerba wrote:
> Many thanks for such a quick reaction!
> I will give it a try next week.
> regards,
> Dominik
>
> Utkarsh Ayachit wrote:
>> Dominik,
>> I just committed a fix for this issue. Feel free to give it a try and 
>> let me know if it still doesn't work.
>>
>> Thanks,
>> Utkarsh
>>
>> /cvsroot/ParaView3/ParaView3/Qt/Core/pqScalarBarRepresentation.cxx,v  
>> <--  Qt/Core/pqScalarBarRepresentation.cxx
>> new revision: 1.11; previous revision: 1.10
>> /cvsroot/ParaView3/ParaView3/Servers/ServerManager/Resources/rendering.xml,v  
>> <--  Servers/ServerManager/Resources/rendering.xml
>> new revision: 1.345; previous revision: 1.344
>> /cvsroot/ParaView3/ParaView3/Servers/ServerManager/vtkSMScalarBarWidgetRepresentationProxy.cxx,v  
>> <--  Servers/ServerManager/vtkSMScalarBarWidgetRepresentationProxy.cxx
>> new revision: 1.10; previous revision: 1.9
>> /cvsroot/ParaView3/ParaView3/Servers/ServerManager/vtkSMScalarBarWidgetRepresentationProxy.h,v  
>> <--  Servers/ServerManager/vtkSMScalarBarWidgetRepresentationProxy.h
>> new revision: 1.6; previous revision: 1.5
>>
>>
>> On Wed, Apr 15, 2009 at 1:49 PM, Dominik Szczerba 
>> <dominik at itis.ethz.ch <mailto:dominik at itis.ethz.ch>> wrote:
>>
>>     -----BEGIN PGP SIGNED MESSAGE-----
>>     Hash: SHA1
>>
>>     Even simpler way to reproduce:
>>
>>     just move the (sliced) dataset with MMB onto the visible colorbar: a
>>     lock results for me. Invisible datasets hide their colorbars, but 
>> they
>>     are somehow still there, locking the navigation: a clear bug, at 
>> least
>>     on my system.
>>
>>     - - Dominik
>>
>>     Dominik Szczerba wrote:
>>      > No, it does not have to do with the mouse outside the render 
>> window.
>>      > This is a bug with colorbar. To reproduce:
>>      >
>>      > Load a dataset, slice it, color it with a scalar, display the
>>     colorbar.
>>      > Load another dataset, slice it, color it.
>>      > Now hide the first slice (unclick the eye).
>>      > The color legend is gone, but stays there hidden, faking the
>>     mouse. Move
>>      > the cursor to see false mouse-over effects. This now often 
>> results in
>>      > the mentioned "locking" while dragging with the middle button. It
>>     is the
>>      >  invisible colorbar that gets moved instead of the dataset and
>>     the whole
>>      > thing gets eventually stuck. It is extremely annoying, if not
>>     impossible
>>      > in case of complex pipeline states, to go through all entries and
>>     switch
>>      > off the colorbars. Is it possible to fix this?
>>      >
>>      > - Dominik
>>      >
>>      > Utkarsh Ayachit wrote:
>>      >> That generally happens when you start dragging but move the
>>     mouse out
>>      >> of the renderwindow before releasing the mouse button. Is that
>>     indeed
>>      >> the case?
>>      >
>>      >> Utkarsh
>>      >
>>      >> On Tue, Apr 14, 2009 at 6:19 AM, Dominik Szczerba
>>     <dominik at itis.ethz.ch <mailto:dominik at itis.ethz.ch>> wrote:
>>      >> Hi,
>>      >> I am getting annoying critical problems when moving datasets
>>     (here, a
>>      >> result of cutting tool): the dataset keeps moving even after I
>>     release
>>      >> the middle mouse button and I have to click around blind like
>>     crazy to
>>      >> "unlock" it. I see the following in the output window:
>>      >
>>      >> endUndoSet called without a beginUndoSet.
>>      >> endUndoSet called without a beginUndoSet.
>>      >> endUndoSet called without a beginUndoSet.
>>      >> endUndoSet called without a beginUndoSet.
>>      >> endUndoSet called without a beginUndoSet.
>>      >> endUndoSet called without a beginUndoSet.
>>      >> endUndoSet called without a beginUndoSet.
>>      >> endUndoSet called without a beginUndoSet.
>>      >
>>      >> Can someone please advise how to fix it?
>>      >
>>      >> regards,
>>      >> Dominik
>>      > _______________________________________________
>>      > Powered by www.kitware.com <http://www.kitware.com>
>>      > Visit other Kitware open-source projects at
>>      > http://www.kitware.com/opensource/opensource.html
>>      > Please keep messages on-topic and check the ParaView Wiki at:
>>      > http://paraview.org/Wiki/ParaView
>>      > Follow this link to subscribe/unsubscribe:
>>      > http://www.paraview.org/mailman/listinfo/paraview
>>      >
>>
>>     - --
>>     Dominik Szczerba, PhD
>>     Computational Physics Group
>>     IT'IS Foundation for Research
>>     www.itis.ethz.ch <http://www.itis.ethz.ch>
>>     -----BEGIN PGP SIGNATURE-----
>>     Version: GnuPG v2.0.9 (GNU/Linux)
>>     Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org
>>
>>     iEYEARECAAYFAknmHiYACgkQ/EBMh9bUuzJwwQCfZSHSpQ99zTAxS6tiIYAhZyvm
>>     DqgAoKFZHkNGdmIro+9P4Zel4P/U7eC4
>>     =d9xA
>>     -----END PGP SIGNATURE-----
>>
>>
>


More information about the ParaView mailing list