[Paraview] PolyData bug??

[Sean Ziegeler]

What you propose is adding the tesselation to the reader, which doesn't 
make much sense design-wise (it would have to be added to every reader). 
  It sounds like what is needed is a filter.  If you know you're working 
with potentially non-convex polygons, simply apply the filter after the 
reader.

That said, wouldn't the Tesselate filter work, or am I missing something?

-Sean

Alejandro Aragon wrote:
> If it is an expensive operation, it could be done only once when reading 
> the input file and save the result as a flag for subsequent rendering. 
> Again, I don't think it's a big deal and it prevents completely from 
> doing scientific visualization if someone uses non-convex polygons. 
> Imagine for some reason that I do CFD with non-convex finite volumes. 
> How in the world am I going to see the elements????
> 
> aa
> 
> On Apr 1, 2008, at 10:44 AM, Moreland, Kenneth wrote:
> 
>> I understand your sentiments, but the solution is not practical.  If 
>> ParaView were primarily used for modeling, which deals with fewer, 
>> more complicated elements, then I would totally agree with you.  
>> However, scientific visualization typically deals with data comprising 
>> many simple elements.
>>
>> Checking for convexity is an expensive operation on data that is 
>> already tessellated.  This is a problem when using ParaView for large 
>> data visualization, which is on of the major things it was designed for.
>>
>> -Ken
>>
>>> -----Original Message-----
>>> From: Alejandro Aragon [mailto:alejandro.aragon at gmail.com]
>>> Sent: Tuesday, April 01, 2008 9:04 AM
>>> To: Moreland, Kenneth
>>> Cc: paraview at paraview.org
>>> Subject: Re: [Paraview] PolyData bug??
>>>
>>> I strongly disagree. The user shouldn't be aware of such a thing. If
>>> Paraview gives you the option of drawing a polygon, it should be
>>> independent on its convexity. What it should do, is to check if the
>>> polygon is convex to use the regular GL function, and if not use
>>> tessellation to draw it. I have done this myself in the past, so I
>>> don't think it's a big deal.
>>>
>>> aa
>>>
>>> On Apr 1, 2008, at 9:06 AM, Moreland, Kenneth wrote:
>>>
>>>> I don't think this is a bug.  OpenGL (and every other rendering
>>>> system that I am aware of) assumes that its polygons are convex.
>>>> You will have to tessellate your polygons into convex pieces in
>>>> order for them to be rendered correctly.
>>>>
>>>> -Ken
>>>>
>>>>> -----Original Message-----
>>>>> From: paraview-bounces at paraview.org [mailto:paraview-
>>> bounces at paraview.org
>>>>> ]
>>>>> On Behalf Of Alejandro Aragon
>>>>> Sent: Monday, March 31, 2008 10:31 PM
>>>>> To: paraview at paraview.org
>>>>> Subject: [Paraview] PolyData bug??
>>>>>
>>>>> Hello everyone,
>>>>>
>>>>> I was playing to see if I could write a complicated shape with
>>>>> PolyData, and I found out that the visualization is really bad. The
>>>>> surface and surface with edges visualizations don't look quite right.
>>>>> I attached a wireframe and the surface with edges plots so you can
>>>>> take a look. Some time back I got the same problem when I tried to
>>>>> draw a polygon using OpenGL. I was able to solve the problem by using
>>>>> tesselation, so I believe the same is happening here. I also attached
>>>>> the vtp file. I'm on Mac OS, but I don't think that would make any
>>>>> difference.
>>>>>
>>>>> Cheers,
>>>>>
>>>>> Alejandro Aragón
>>>>
>>>>
>>>
>>
>>
> 
> _______________________________________________
> ParaView mailing list
> ParaView at paraview.org
> http://www.paraview.org/mailman/listinfo/paraview
> 
>