[Paraview] EnSight multi-part loading

[Jean M. Favre]

While reading EnSight geom files composed of multiple parts, I was 
surprised by the large number of points shown in the 
Information.statistics field of my reader object.

my data was made of just over 2 million points and ParaView showed I had 
202 million points. For a moment, I thought I had beaten my own 
visualization record. :-)

The answer is that the statistics are calculated by adding the number of 
points for each EnSight part, i.e. each VTK Output, event though the 
vtkPoints is ref. counted, and only created once. Since I had 101 parts, 
I had 101 times the number of real points.

What is the justification for that? no need to boost performance like 
that. Isn't ParaView already good enough?

Somebody cares to comment?

Jean


-- 
Dr. Jean M. Favre,                            email: favre @ cscs . ch
http://www . cscs . ch/people/Favre . html
CSCS, Swiss Center for Scientific Computing | Tel:  +41 (91) 610.82.40
Via Cantonale, 6928 Manno, Switzerland      | Fax:  +41 (91) 610.82.82