Skip to Content.
Sympa Menu

cgal-discuss - Re: [cgal-discuss] Surface reconstruction compiler error

Subject: CGAL users discussion list

List archive

Re: [cgal-discuss] Surface reconstruction compiler error


Chronological Thread 
  • From: "Laurent Rineau (GeometryFactory)" <>
  • To:
  • Subject: Re: [cgal-discuss] Surface reconstruction compiler error
  • Date: Fri, 16 Oct 2009 19:43:50 +0200
  • Organization: GeometryFactory

On Thursday 15 October 2009 14:21:59 David Doria wrote:
> > Actually I was wrong. I have re-read the C++ norm, in the section
> > *8.5 Initializers [dcl.init]*, and even in the the initializers for
> > aggregates
> > and arrays, the expressions in the { } do not need to be constant.
> >
> > What compiler do you use? It is probably buggy, and not supported by the
> > CGAL
> > project.
> >
> > I'm using gcc 4.4. Why not use the non-initialization list method if
> > there
>
> is any question/confusion and it seems to not work on some compilers?

CGAL-3.5 has been tested with gcc-4.4. The problem must be elsewhere. Can you
give us the full error log?

> CGAL examples and demos are shipped with CMake scripts, that precisely
>
> > states
> > which libraries to link with, for each example or demo. If you do not use
> > CMake, at least you should have a look at those files. Sometimes, the
> > list of
> > libraries to link with depends on the platform. We cannot duplicate the
> > logic
> > of the CMakeLists.txt files in another place.
> >
> > I see the list now in the CMakeLists.txt file, but it seems to not
> > include
>
> metis, which was required.

Actually, the CMake macro FindTAUCS, called by find_package(TAUCS), handles
the dependencies of TAUCS.

> I don't think putting a comment in the classes
> is duplicating the logic - it is simply a more informative way to tell
> users "If you use this class, you must use this library". That is very
> different than picking a handful of examples and saying in the cmake file
> you must use X,Y and Z libraries for this example - then the user would
> still have to figure out which classes use which libraries before building
> their own projects.

You are right. But sometimes it is quite difficult to describe: that depends
on the platform. Anyway, I agree such things should be documented, as much as
possible.

> > TAUCS is a big library, and most of CGAL packages do not use it. What is
> > more,
> > TAUCS is an external library, with different copyright and license than
> > CGAL.
> > The CGAL project has the right to redistribute TAUCS, but if it was in
> > the same tarball as CGAL, the license scheme of the tarball would be more
> > complicated than now.
> >
> > --
> > Laurent Rineau, PhD
>
> Understandable. Maybe the website can at least be updated with this warning
> about the >3.3 versions not working with the current release of taucs in
> all the places the taucs links are given?

Yes, that will be done.

--
Laurent Rineau, PhD
R&D Engineer at GeometryFactory http://www.geometryfactory.com/
Release Manager of the CGAL Project http://www.cgal.org/



Archive powered by MHonArc 2.6.16.

Top of Page