is not available.
*Add default empty struct for Get_arithmetic_kernel. That way we can test if
a specialization is avaible. I used it for the default case of Root_of_traits.
When a number type not known by CGAL is used to instanciate Root_of_traits,
the field type used to instanciate Sqrt_extension is made using Quotient.
*Patch of Lazy_exact_nt constructor from another Lazy_exact_nt type.
In revision 61559, that constructor was retricted to the case the exact type
is implicitely constructed from the other exact type. Now the old behavior is
back but only if it is explicitely called (patch from Marc Glisse).
*Update To_double specialization of Sqrt_extension to use the same formula
than To_interval so that the double value is contained in the interval.
*update root_of_2 test-suite to use Sqrt_extension and that they behave
similarly
move Get_arithmetic_kernel specialization into MP_Float.h
reorder declaration of to_double(Sqrt_extension<MP_Float>)
Sqrt_extension need specialization of Real_embeddable_traits
and Algebraic_structure_traits.
We declare
cmake_minimum_required(VERSION 2.6.2)
but we also use
cmake_policy(VERSION 2.8.4)
to declare that our CMake scripts are OK with all the defaults of CMake policies
as of CMake-2.8.4. That shuts down the warnings of CMake-2.8.4.
That way, we no longer need any declaration of specific policies.
Those two lines must be present and maintained in all our CMakeLists.txt
files (the one for the libraries, and also the one for examples and demos,
and maybe tests).
The documentation of CGAL says that the minimal version of CMake must be
CMake-2.6.2.
- change cmake_minimum_required to VERSION 2.6.2
- no longer any need for the policy CMP0003 (was for 2.4.x)
- no longer any need for CMAKE_ALLOW_LOOSE_LOOP_CONSTRUCTS (was for 2.4)
Changed the calls mpfr_get_z_exp -> mpfr_get_z_2exp for MPFR>2 (the old
function name may be deprecated in future versions).
Added comments in the Gmpfr code about Gmpfr destructors.
Updated my mail address in source files.
On Windows, with MSVC, std::fabs is an intrinsic function with /Oi (implied
with /O1, /O2, and so on) *if* /fp:strict is not used. We should reevaluate
the need to use /fp:strict (and maybe use /fp:precise, the default,
combined with #pragma fenv_access).
First step: bench! This macro will help.
::test.
The error message was:
/home/lrineau/CGAL/boost/1.44-beta1/include/boost/type_traits/has_new_operator.hpp: In function 'int main()':
/home/lrineau/CGAL/boost/1.44-beta1/include/boost/type_traits/has_new_operator.hpp:24: error: 'template<class U, U x> struct boost::detail::test' is not a function,
/home/lrineau/CGAL/CGAL-3.7-I-139/cmake/platforms/x86-64_Linux-2.6_g++-4.4.4_F13/test/Number_types/to_interval_test.cpp:30: error: conflict with 'template<class NT> void test(const NT&)'
/home/lrineau/CGAL/CGAL-3.7-I-139/cmake/platforms/x86-64_Linux-2.6_g++-4.4.4_F13/test/Number_types/to_interval_test.cpp:73: error: in call to 'test'
| ------------------------------------------------------------------------
| r58141 | lrineau | 2010-08-18 15:08:04 +0200 (Wed, 18 Aug 2010) | 5 lines
| Changed paths:
| M /branches/CGAL-3.7-branch/Algebraic_kernel_d/include/CGAL/Algebraic_kernel_d/Float_traits.h
| M /branches/CGAL-3.7-branch/Number_types/include/CGAL/CORE_BigFloat.h
|
| Fix the "bug" of CORE-1.7 in 64 bits. The bug was actually in CGAL
| Number_types and Algebraic_kernel_d! The basis of CORE::BigFloat is not
| 2^14: it is 2^CORE::CHUNK_BIT. CORE::CHUNK_BIT is 14 in 32 bits, but *30*
| in 64 bits!
|
| ------------------------------------------------------------------------
Update: See revision 58145.
-- Laurent Rineau Wed Aug 18 16:26:15 CEST 2010
Sébastien Loriot
Laurent Rineau
Andreas Fabri
Luis Peñaranda
Marc Glisse
Eric Berberich
Michael Hemmer
Daniel Russel
Sylvain Pion