[vtkusers] RE: Question asked by many, answered by none. (Help!)

[Goodwin Lawlor]

The problems is also that the search facility for the VTK users archive
doesn't work very well... Very often a result will link to the wrong
page.

I've put a google site search for the user archive here: 
http://www.bioengineering-research.com/vtk/

I've included the naked html... It would be nice to have this on vtk.org

Goodwin



-----Original Message-----
From: jarv0075 [mailto:jarv0075 at umn.edu] 
Sent: 23 March 2004 01:43
To: goodwin.lawlor at ucd.ie; vtkusers at vtk.org
Cc: jarv0075 at umn.edu
Subject: Re: Question asked by many, answered by none. (Help!)


I think this kind of reiterates the usefulness of a giant FAQ of sorts
written in plain-english / VTK newbie speak...  I've spent hours
searching the VTK users archives but hadn't found anything.. but as you
all know, 95% of solving a problem is figuring out how to phrase the
question. (i.e., knowning the lingo).  I realize you all do this on your
free time out of sheer benevolence, but if anyone is feeling exceedingly
benevolent, and somewhat bored.. it would be a noble undertaking.

Anyways, I will definitely look into this at work on Weds.  

Thank you for answering

On 22 Mar 2004, Goodwin Lawlor wrote:
> Using vtkCleanPolyData before vtkStripper should help.
> 
> hth
> Goodwin
> 
> Ps this was (sort of) answered last year... 
> http://public.kitware.com/pipermail/vtkusers/2003-February/016061.html
> 
> -----Original Message-----
> From: vtkusers-admin at vtk.org [mailto:vtkusers-admin at vtk.org] On Behalf

> Of vtkusers-request at vtk.org
> Sent: 23 March 2004 00:25
> To: vtkusers at vtk.org
> Subject: vtkusers digest, Vol 1 #2508 - 6 msgs
> 
> 
> Message: 5
> Date: Mon, 22 Mar 2004 19:21:15 -0500
> From: Mathieu Malaterre <mathieu.malaterre at kitware.com>
> Organization: Kitware Inc.
> To: jarv0075 <jarv0075 at umn.edu>
> Cc: vtkusers <vtkusers at vtk.org>
> Subject: Re: [vtkusers] Question asked by many, answered by none.
> (Help!)
> 
> jarv0075,
> 
> 	Why didn't you used a vtkContourFilter ? I don't see any other
way to
> deal with such difficult polydata
> 
> Sorry,
> Mathieu
> 
> jarv0075 wrote:
> > The demo I refer to is the standard 'cow.tcl' demo under the 
> > vtkCutter
> 
> > examples.  In this, they use the edges from vtkCutter to generate a
> > polyline loop, which works fine for simple polygons but once it 
> > encounters holes it seems to be still trying to form one large 
> > polyline loop?  When it triangulates, all triangles originate from a

> > single point, and thus it triangulates out to the edges of the
holes, 
> > and the edges past the holes, making things a total mess (see screen

> > shot)
> > 
> > The screen shot can be found at 
> >   http://www.cs.umn.edu/~tjarvis/tri_fail.jpg
> > 
> > The error message is as follows:
> >   ERROR: In vtkPolygon.cxx, line 1179
> >   vtkPolygon (0x818da20): Possible triangulation failure
> > 
> > I want to be able to cut with the plane, and fill in the slice with
> > some color.  Basically, to be able to triangulate the output of 
> > vtkCutter even when holes are present.
> > 
> > 
> > Any help you can offer would be greatly appreciated!
> > 
> > 
> > On 22 Mar 2004, Mathieu Malaterre wrote:
> > 
> >>jarv0075,
> >>
> >>	You are loosing me, what 'wonderfully written samples' are you
> >>talking
> >>about. Could you please send us a demo + dataset to demonstrate the 
> >>problem, or a screenshot.
> >>
> >>	Thanks
> >>Mathieu
> >>
> >>jarv0075 wrote:
> >>
> >>> 
> >>>The issue is cutting polygonal data with a plane that results in a
> >>>'polygon' with holes in it.  Thus, the wonderfully written samples
do
> > 
> > not
> > 
> >>>apply, in that the 'sneaky shortcut' of using the edges to make
> > 
> > polylines
> > 
> >>>sends the triangulation filter into a tizzy.
> >>>
> >>>Could someone point me in the right direction on how to triangulate
> cut
> >>>polydata with holes present?   This does NOT seem like it should be
> >>>difficult, but still I've seen the question asked by multiple 
> >>>others
> > 
> > and
> > 
> >>>haven't seen one answer surface yet.
> >>>
> >>>Thank you for your time and patience.
> 
>