CGAL users discussion list

[Monique Teillaud <>]

Hi Tom,

I am trying the new list

tom wrote:
> I see nobody is reacting, 

the only non answered question I can remember was about cgal 3.1

> so I`ll try to write it more straightly:
> I`m inserting points into regular triangulation and when I run the program it 
> stops whith this announcement:
>
> CGAL error: precondition violation!
> Expr: orientation(p0, p1, p2, p3) == POSITIVE
> File: /home/xskoupi1/CGAL/include/CGAL/Regular_triangulation_3.h
> Line: 745
>
> And in this file on this exact line (745) is some precondition about 
> orientation.

> Why is it testing orientation, when I`m inserting points into regular 
> triangulation?
> Or why is it executing function side_of_oriented_power_sphere (in which this 
> line and so this precondition is) when I`m inserting points into regular 
> triangulation?

The algorithm is building a correct regular triangulation, so it checks 
at each insertion that new points are tested against valid cells (for 
instance in case users build cells 'by hand'). Valid cells must be well 
oriented and must satisfy the 'empty sphere' condition.

These preconditions should not fail if the triangulation is valid.

I might be an arithmetic problem, if you are using a non exact kernel 
(typically Cartesian<double>). You can have a look at the FAQ.

> Please, at least tell me this is a stupid question, but tell me something. I 
> really don`t know what to do.

I will not dare say so ;)

best
        Monique