[vtkusers] Re: GenericDataSetTessellator question

[Thompson, David C]

> Hi Soeren,
>
> I suspect that this problem comes from the implementation of:
>
> int vtkBridgeCell::GetNumberOfBoundaries(int dim)
>
> For dim=0, the number of boundaries is NOT the number of points but the 
> number of "corner" points, for a quad it is always 4, whatever is the 
> number of points. The current implementation just returns the actual 
> number of points for a linear cell, but for a none linear cell it 
> returns the number of points minus the number of edges, which is right 
> for the VTK quadratic cells (because there is one extra point per edge), 
> BUT which is wrong in your case. So you should add your case here by 
> returning 4 in the quad case and see what happens.

This interface may change in the future to accommodate
higher order cells whose higher order modes aren't
associated with interpolated points; in the future,
the point iterator will most probably only loop over the
"corner" points. The mid-edge and mid-face points will
be accessible through an as-yet-undefined DofNodeIterator
which will return mode-shape coefficients which may or may
not be interpolated exactly.

        David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20050325/e11a06a5/attachment.htm>