[vtk-developers] [vtkusers] vtkCellPicker question

[Stephen Langer]

Hi David --

I managed to create a small test that recreates the problem.  I can't promise that it's minimal, but it's pretty close.  I uploaded it to http://vtk.org/Bug/view.php?id=10349.

As the original bug report says, the effect is very sensitive to the exact parameters used by the Camera, Locator, and Picker, which made it difficult to extract the right set of lines from the program in which I first saw the bug.

 -- Steve


On Apr 17, 2013, at 3:26 PM, David Gobbi wrote:

> Hmm, it looks like this patch changes the result of one of the
> existing VTK tests:
> 
> http://open.cdash.org/testDetails.php?test=186097519&build=2878639
> 
> My guess is that some faces that this test used to miss are now
> being found, as a result of the patch.  But it will require further
> investigation.
> 
> 
> On Wed, Apr 17, 2013 at 12:17 PM, David Gobbi <david.gobbi at gmail.com> wrote:
>> Hi Steve,
>> 
>> I've submitted a patch for consideration:
>> http://review.source.kitware.com/#/c/10931/
>> Would it be possible for you to put together a minimal test example
>> that fails on your system?  That will improve the chances of the patch
>> being accepted.
>> 
>> - David
>> 
>> 
>> On Wed, Apr 17, 2013 at 11:28 AM, Stephen Langer
>> <stephen.langer at nist.gov> wrote:
>>> Responding to my own question:  This is due to a bug in vtkCellLocator, reported at http://www.vtk.org/Bug/view.php?id=10349.
>>> 
>>> -- Steve
>>> 
>>> On Apr 12, 2013, at 3:51 PM, Stephen Langer wrote:
>>> 
>>>> Hi --
>>>> 
>>>> I'm using a vtkCellPicker to select cells in a tetrahedral mesh.  Sometimes the point returned by vtkCellPicker::GetPickedPositions is on the near side of the picked tetrahedron, and sometimes it's on the far side.  Changing the position passed into vtkCellPicker::Pick by only a pixel or two will cause the returned point to switch sides.  Is it possible to force the picker to return the point on the near side, or at least to return both points?  I'm really interested in identifying the nearer triangular face of the tetrahedron, but my vtkUnstructuredGrid contains only tetrahedral cells.
>>> 
>>> --
>>> -- stephen.langer at nist.gov                    Tel: (301) 975-5423 --
>>> -- http://math.nist.gov/~SLanger/             Fax: (301) 975-3553 --
>>> -- NIST, 100 Bureau Drive, Stop 8910, Gaithersburg, Md 20899-8910 --
>>> 
>>> "I've just had the best idea. I'm going to make hats out of Nutella."
>>>                                       -- Nadia Langer, 10 Mar 2012
>>> (Not intended as a product endorsement or statement of suitability.)

--
-- stephen.langer at nist.gov                    Tel: (301) 975-5423 --
-- http://math.nist.gov/~SLanger/             Fax: (301) 975-3553 --
-- NIST, 100 Bureau Drive, Stop 8910, Gaithersburg, Md 20899-8910 --

"I've just had the best idea. I'm going to make hats out of Nutella."
                                       -- Nadia Langer, 10 Mar 2012 
(Not intended as a product endorsement or statement of suitability.)