CGAL users discussion list

[Bruce Duncan <>]

Sebastien Loriot (GeometryFactory) wrote:
> Bruce Duncan wrote:
>> On Lunes 09 Agosto 2010 19:43:47 Sylvain Pion escribió:
>>> Le 06/08/10 17:37, Bruce Duncan a écrit :
>>>> Sylvain Pion wrote:
>>>>> Le 06/08/10 14:02, Bruce Duncan a écrit :
>>>>>> For one of the assertions I'm hitting, I've created what I hope is a
>>>>>> small testcase. The code and the output on my machine is attached.
>>>>> The assertion means that an uninitialized variable has been used.
>>>>> And indeed the plane equations in the Polyhedron of CGAL needs manual
>>>>> construction.  Seehttp://www.cgal.org/FAQ.html#polyhedron_normals .
>>>> I have tried to modify my program to use the Traits class as shown in
>>>> the example program.
>>>>
>>>> I am now having trouble converting this Polyhedron to a Nef_polyhedron,
>>>> which I want to use for the boolean operations. Is this possible?
>>>>
>>>> I haven't attached the compiler output this time, as it's very long.
>>>> Suffice to say it starts with "Nef_3/polyhedron_3_to_nef_3.h:218:
>>>> error:
>>>> 'class Vector_3' has no member named 'is_degenerate'"
>>>>
>>>> What am I doing wrong?
>>> I am not well versed enough into the Nef polyhedron to answer.
>>> I hope someone else can follow up.  In any case, I advice to send
>>> as much of the error message as possible : I'm not sure it's
>>> possible for anyone to see what's going wrong here.
>>
>> Hi,
>>
>> Sorry for the delay in replying.
>>
>> I have attached my test case and the compiler output. The code
>> compiles if line 20 is changed from Polyhedron_traits_with_normals_3
>> to just Polyhedron_traits_3.
>>
>> I looked at the difference between CGAL/Polyhedron_traits_3.h and
>> CGAL/Polyhedron_traits_with_normals_3.h. This hunk seemed interesting
>> (line 32):
>>
>> -    typedef typename Kernel::Plane_3  Plane_3;
>> +    typedef typename Kernel::Vector_3 Plane_3;
>>
>> I changed the code to:
>>
>>     typedef typename Kernel::Plane_3 Plane_3;
>>     typedef typename Kernel::Vector_3 Vector_3;
>>
>> and it seems to compile. Is this a bug in CGAL? Perhaps the author
>> intended both Plane_3 and Vector_3 to be defined?
>>
>> Thanks,
>> Bruce
>>
> Hello Bruce,
> 
> I had a look at the code and the problem is in my opinion in
> polyhedron_3_to_nef_3 which assumes that the type Plane defined by the
> polyhedron traits is a of type Plane<K>. I will have a closer look at it
> to find a solution.
> 
> Can you use Polyhedron_traits_3 instead?

Hi Sebastien,

I think I really need the plane normals. I eventually hope to export
these data to a fortran code which will need them. Any hope that this
bug can be fixed? Am I allowed to call it a bug? :)

Bruce
-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

Attachment: signature.asc
Description: OpenPGP digital signature