December 2014 - Boost-users - lists.preview.boost.org

[TEST] Tests registered manually never exectued....
by Witold E Wolski 23 Dec '14

23 Dec '14

Im trying to manually register unit tests following the example at : http://www.boost.org/doc/libs/1_57_0/libs/test/doc/html/utf/user-guide/test… My registration code looks like follows: boost::unit_test::test_suite* init_unit_test_suite(int argc, char* argv[]) { boost::unit_test::test_suite* ts2 = BOOST_TEST_SUITE("test_suite2"); ts2->add(BOOST_TEST_CASE(&liveViewTest)); ts2->add(BOOST_TEST_CASE(&checkImageFormats)); ts2->add(BOOST_TEST_CASE(&testRethrowException)); boost::unit_test::framework::master_test_suite().add(ts2); return 0; } The init_unit_test_suite function is entered and returns but non of the added tests is ever exectued. regards -- Witold Eryk Wolski

2 1

[boost] [compute] Review period extended till December 30 + sum up
by Antony Polukhin 23 Dec '14

23 Dec '14

Hi, Review period of Compute library extended till December 30. Some notes and answers to them that were pointed out during review so far: * Some of the algorithms could be tuned for a specific hardware - Library author is working on a auto-tuning solution and improves existing algos * Why OpenCL C API used instead of C++ API? - C++ API had issues, C API allows a bit more control. No C++ API for OpenCL 2.0 * API is not N4105 compatible, library is not N*** compatible - This could be easily fixed by a library that will use Compute as a backend * Some of the algorithms return futures while other work with command queues - This is because of the OpenCL design + async chains/futures would be probably added later * Not all the types of the Khronos API are supported - A few types are not documented yet, there's a patch in the works which would allow the Khronos C++ types to be passed into Compute algorithms * A few type related errors could be detected at compile time - This will be fixed, more asserts would be added to runtime only checkable places. Report issues to the tracker https://github.com/kylelutz/compute/issues * On small data sets CPU algorithm would work faster than a GPGPU. How about implicitly dispatching algo on CPU in those cases? - The call on whether to execute the algorithm should be left up to the user. While library author agrees that this would be a useful feature, he just don't think Compute is the right place for that logic. * How about providing way to do chains of async operations - This is a big task that will be solved some day. * How about providing Boost.ASIO like error handling via throw and error_code - Implementing an approach like ASIO's wouldn't be that difficult. Thanks to all the reviewers for spending their time and providing useful comments so far! -- Best regards, Antony Polukhin

1 0

interprocess::rbtree_best_fit assertion failure with vector resize
by John Harris 22 Dec '14

22 Dec '14

I encountered this with 1.56 and 1.57 under heavy allocation load where I repeatedly resize an interprocess vector<unsigned char>. The only way I was able to get around it was to switch to the simple_seq_fit allocation algorithm. The assertion is assert(FALSE), and so I'm wondering if the algorithm doesn't handle all cases where there is a single-byte type for the vector. The stack below is from a core dump, and so the "Cannot access memory" errors are not necessarily indicative of bad pointers. #3 0x0000003dc2a2b820 in __assert_fail () from /lib64/libc.so.6 #4 0x00007f869a166793 in boost::container::container_detail::insert_value_initialized_n_proxy<boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> >, unsigned char*>::copy_n_and_update (this=0x7f8698264b80) at /home/jharris/usr/local/boost/include/boost-1_57/boost/container/detail/advanced_insert_int.hpp:118 #5 0x00007f869a165fd6 in boost::container::vector<unsigned char, boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> > >::priv_forward_range_insert_expand_backwards<boost::container::container_detail::insert_value_initialized_n_proxy<boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> >, unsigned char*> > ( this=0x7f85adc00a40, new_start=0x7f85adc165a0 <error: Cannot access memory at address 0x7f85adc165a0>, new_capacity=199624, pos=0x7f85adc31974 <error: Cannot access memory at address 0x7f85adc31974>, n=101192, insert_range_proxy=...) at /home/jharris/usr/local/boost/include/boost-1_57/boost/container/vector.hpp:2783 #6 0x00007f869a164c0b in boost::container::vector<unsigned char, boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> > >::priv_forward_range_insert_no_capacity<boost::container::container_detail::insert_value_initialized_n_proxy<boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> >, unsigned char*> > (this=0x7f85adc00a40, pos=..., n=101192, insert_range_proxy=...) at /home/jharris/usr/local/boost/include/boost-1_57/boost/container/vector.hpp:2265 #7 0x00007f869a1639ec in boost::container::vector<unsigned char, boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> > >::priv_forward_range_insert<boost::container::container_detail::insert_value_initialized_n_proxy<boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> >, unsigned char*> > (this=0x7f85adc00a40, pos=..., n=101192, insert_range_proxy=...) at /home/jharris/usr/local/boost/include/boost-1_57/boost/container/vector.hpp:2328 #8 0x00007f869a16241a in boost::container::vector<unsigned char, boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> > >::priv_forward_range_insert_at_end<boost::container::container_detail::insert_value_initialized_n_proxy<boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> >, unsigned char*> > (this=0x7f85adc00a40, n=101192, insert_range_proxy=...) at /home/jharris/usr/local/boost/include/boost-1_57/boost/container/vector.hpp:2365 #9 0x00007f869a1604e5 in boost::container::vector<unsigned char, boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> > >::priv_resize<boost::container::value_init_t> (this=0x7f85adc00a40, new_size=178780, u=...) at /home/jharris/usr/local/boost/include/boost-1_57/boost/container/vector.hpp:2151 #10 0x00007f869a15de56 in boost::container::vector<unsigned char, boost::interprocess::allocator<unsigned char, boost::interprocess::segment_manager<char, boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family, boost::interprocess::offset_ptr<void, long, unsigned long, 0ul>, 0ul>, boost::interprocess::iset_index> > >::resize (this=0x7f85adc00a40, new_size=178780) at /home/jharris/usr/local/boost/include/boost-1_57/boost/container/vector.hpp:1213 #

1 0

[Random] normal distribution different behaviour 1.55 vs 1.57
by Thomas M 21 Dec '14

21 Dec '14

Hi, I have just switched an existent project from boost 1.55.0 to 1.57.0 and my test suites diagnosed a divergence in program (a simulation) results. I have tracked down the issue to different behaviour in the normal distribution random variates generator. Below is a simple program (run under VS 2012, Update 4) that outputs random variates for a uniform and a normal distribution; comparing the output between 1.55.0 and 1.57.0 it turns out that the first bunch of uniformly distributed variates is identical, and then all others variates (both bunches of normally distributed variates as well as the second bunch of uniformly distributed variates) diverge. Neither for 1.56.0 nor 1.57.0 does the change-log list an update to the random numbers library, so first I am puzzled why the libraries behave differently (though [Math] has undergone some changes -> propagation to [Random]?). Second, it is not clear to me why for the first bunch of uniformly distributed variates the results are identical, while for the second, after normally distributed variates were generated, the are not. It appears that the generation of normally distributed variates changes the whole state of the random numbers engine in a different manner [e.g multiple engine calls ??]. Any insight into what is going on here is much appreciated. And foremost: Is any of 1.55 or 1.57 bugged, that is one should be clearly preferred over the other? many thanks, Thomas #include <ostream> #include <iostream> #include <fstream> #include "boost/random.hpp" void test_rng() { typedef boost::mt19937 rnengine_t; typedef boost::uniform_01<double> uniform_distr_t; typedef boost::normal_distribution<double> normal_distr_t; typedef boost::variate_generator<rnengine_t &, uniform_distr_t> uniform_gen_t; typedef boost::variate_generator<rnengine_t &, normal_distr_t> normal_gen_t; uniform_distr_t uniformDistr; normal_distr_t normalDistr(0, 1); rnengine_t rnEngine; rnEngine.seed(100); uniform_gen_t uniformGen(rnEngine, uniformDistr); normal_gen_t normalGen(rnEngine, normalDistr); std::ofstream outFile("randomNumbers.txt"); outFile.precision(16); int n = 20; for (int outer = 0; outer < 2; ++outer) { for (int i = 0; i < n; ++i) outFile << "Uniform-distr variate #" << (n*outer) + i + 1 << ": " << uniformGen() << std::endl; for (int i = 0; i < n; ++i) outFile << "Normal-distr variate #" << (n*outer) + i + 1 << ": " << normalGen() << std::endl; } } int main() { test_rng(); return 0; }

5 8

Boost.container iterator_traits and flat_map
by gast128 21 Dec '14

21 Dec '14

Hello all, I have a problem here which boils down to the non-constness of iterator_traits for const_iterator of flat_map: void Foo() { typedef boost::container::flat_map<int, int> Int2IntFmap; Int2IntFmap::const_iterator it2; std::iterator_traits<Int2IntFmap::const_iterator>::value_type::first_type& r2 = it2->first; } 1>d:\develop\shared\main\test\testcommon\testgeneral\nctstgenhmhandlemap.cpp(59): error C2440: 'initializing' : cannot convert from 'const int' to 'int &' Anyone an idea or should I use decltype instead?

1 0

[multi_index] any interests extending ord_index_node to argumented ord_index_node
by fractal 19 Dec '14

19 Dec '14

dear all: i'm a game developer, and my projects used a lot of multi_index(thanks, it's a great library). one of main problems of game development is to rank players according to they level. usually we use boost::interprocess::map for the task. but more, players should know which rank they are and who is at rank 10000 (for example). these requirements lead to two new methods for map: iterator find_by_rank(int idx) int rank(iterator i) i implemented an argumented rbtree data structure according to "Algorithms and Data Structures " (see the attachment rank_tree.hpp), and latter i find coutertree which is in boost review schedule. but IMHO, the best way is to integrate it to multi_index, and if anyone interests, i can do this by some efforts. yours sincerely

1 0

RFC: library for fractal algorithms
by legalize+jeeves＠mail.xmission.com 17 Dec '14

17 Dec '14

[Please do not mail me a copy of your followup] I'm interested in talking with other folks about the best way to generalize various fractal algorithms in the style of the standard library: iterators, containers and algorithms. If there are any libraries that already exist that apply to this sort of code, I'd love to hear about them. Most fractals consist of iterating a sequence for a point in the complex plane. The iterating sequence is parameterized for each point in the complex plane and some characteristic of the iterating sequence is used to color that point. The result is an image. For instance, the classic Mandelbrot fractal is: z <- z^2 + c z0 := 0 The complex number 'c' changes for each point in the complex plane. This sequence is iterated until norm(z) > 4, or some maximum iteration count is reached. Pixels in the image correspond to each value of 'c' and are typically colored according to the number of iterations needed until norm(z) exceeded 4. There are a couple things that make for interesting variation in the algorithms used in rendering fractal images: - choice of formula to iterate in the complex plane - parameters for deciding the "bailout" criteria: - iteration counts - predicate function on z, i.e. [](complex<float> z){ return norm(z) > 4; } Iterating until the bailout test passesis the simplest mechanism, but there are gains to be made by applying more sophisticated mechanisms. A significant optimization is to look for cycles in the sequences. With very large iteration counts, detecting cycles in the iteration can save significant amounts of time. The cycle detection algorithm is basically from Knuth and isn't specific to fractals but ends up being mixed together with the Mandelbrot set computation in most programs. This seems to be another opportunity to use generic composition in order to apply this orthogonal feature to any iterated formula in the complex plane. I did a little experiment casting the Mandelbrot formula as an iterator and got the code at the end of this message. Ideally I'd like to explore approaches that result in code that is ignorant of the underlying representation of a complex number in order to have algorithms that can be applied to arbitrary precision numeric types. My little experiment works, but even just looking at this little experiment makes me think that the mandelbrot iterator is doing too much: it is performing the z <- z^2 + c business, but it is also doing the bailout test and these seem to be separate responsibilities. #include <complex> #include <iostream> #include <iterator> #include <vector> class mandelbrot : public std::iterator< std::input_iterator_tag, std::complex<float>, unsigned> { public: mandelbrot() : c_{}, iter_{}, max_iter_{}, bailout_{}, z_{}, bailed_out_{true} { } mandelbrot(std::complex<float> const& c, int max_iter = 256, float bailout = 2.0f) : c_{c}, max_iter_{max_iter}, bailout_{bailout*bailout}, iter_{0}, z_{0.f, 0.f}, bailed_out_{false} { } mandelbrot& operator++() { z_ *= z_; z_ += c_; bailed_out_ = ++iter_ > max_iter_ || std::norm(z_) >= bailout_; return *this; } bool operator==(mandelbrot const& rhs) const { if (bailed_out_ && rhs.bailed_out_) { return true; } if (bailed_out_ != rhs.bailed_out_) { return false; } return z_ == rhs.z_; } std::complex<float> const& operator*() const { return z_; } private: std::complex<float> c_; int const max_iter_; float const bailout_; int iter_; std::complex<float> z_; bool bailed_out_; }; bool operator!=(mandelbrot const &lhs, mandelbrot const &rhs) { return !(lhs == rhs); } int main() { float const r_min = -3.0f; float const r_max = 1.0f; float const i_min = -2.0f; float const i_max = 2.0f; std::vector<float> imag; imag.resize(24); for (int y = 0; y < 24; ++y) { imag[y] = i_min + (i_max - i_min)*(23 - y)/24.0f; } std::vector<float> real; real.resize(80); for (int x = 0; x < 80; ++x) { real[x] = r_min + (r_max - r_min)*x/80.0f; } std::string printables; for (int i = 32; i < 128; ++i) { printables.push_back(char(i)); } for (int y = 0; y < 24; ++y) { float const im = imag[y]; for (int x = 0; x < 80; ++x) { float const re = real[x]; mandelbrot end; mandelbrot it{std::complex<float>{re, im}}; unsigned count = std::distance(it, end); std::cout << printables[count % printables.size()]; } std::cout << '\n'; } } -- "The Direct3D Graphics Pipeline" free book <http://tinyurl.com/d3d-pipeline> The Computer Graphics Museum <http://computergraphicsmuseum.org> The Terminals Wiki <http://terminals.classiccmp.org> Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>

1 1

c++/cli & file.h
by It neophyte 16 Dec '14

16 Dec '14

hi quick question: could boost headers (.h) be complied in c++/cli project? I was told, no by someone, but need further feedback to be sure. Thanks

10 13

[endian] Heads up: breaking name changes coming
by Beman Dawes 15 Dec '14

15 Dec '14

Boost.Endian develop branch is about to have major changes merged in as part of the preparation for a mini-review. AFAICR, the breaking changes are limited to header, class, and function name changes, but there are a lot of hopefully non-breaking changes so I could easily be wrong about that. If you are currently using Boost.Endian develop branch, you will need to change your code accordingly. It would make sense to wait a few days until the mini-review merge is complete and any fixes required by regression test failures are fixed. Thanks for your patience, --Beman

1 0

Does shared_ptr support interface pointers without a virtual destructor
by John Selbie 15 Dec '14

15 Dec '14

What is the general guidance with using shared_ptr on interfaces that do not have virtual destructor? Take the following code sample: class IFoo{public: virtual void M1() = 0;}; class Foo : public IFoo{public: void M1() {} Foo() { std::cout << "Foo constructor\n";} ~Foo(){std::cout << "~Foo destructor\n";}}; I think we all know that this is not safe: { IFoo *pFoo = new Foo(); delete pFoo;} Since IFoo is missing a virtual destructor, then ~Foo() will not be invoked in the above case. So let's switch to using std::shared_ptr { std::shared_ptr<IFoo> spFoo(new Foo());} std::shared_ptr has a templated constructor that creates a "deleter". So when "spFoo" goes out of scope, the ~Foo destructor is properly invoked even though it's of type shared_ptr<IFoo>. But since my code targets older compilers, I have made use of Boost. But attempting this: { boost::shared_ptr<IFoo> spFoo(new Foo());} Generates a lot of scary warnings when compiling with -Wall. In file included from /home/jselbie/boost_1_57_0/boost/checked_delete.hpp:15:0, from /home/jselbie/boost_1_57_0/boost/smart_ptr/shared_ptr.hpp:26, from /home/jselbie/boost_1_57_0/boost/shared_ptr.hpp:17, from main.cpp:9:/home/jselbie/boost_1_57_0/boost/core/checked_delete.hpp: In instantiation of ‘void boost::checked_delete(T*) [with T = Foo]’:/home/jselbie/boost_1_57_0/boost/smart_ptr/detail/shared_count.hpp:134:38: required from ‘boost::detail::shared_count::shared_count(Y*) [with Y = Foo]’/home/jselbie/boost_1_57_0/boost/smart_ptr/shared_ptr.hpp:271:47: required from ‘void boost::detail::sp_pointer_construct(boost::shared_ptr<X>*, Y*, boost::detail::shared_count&) [with T = IFoo; Y = Foo]’/home/jselbie/boost_1_57_0/boost/smart_ptr/shared_ptr.hpp:349:58: required from ‘boost::shared_ptr<T>::shared_ptr(Y*) [with Y = Foo; T = IFoo]’main.cpp:36:44: required from here/home/jselbie/boost_1_57_0/boost/core/checked_delete.hpp:34:5: warning: deleting object of polymorphic class type ‘Foo’ which has non-virtual destructor might cause undefined behaviour [-Wdelete-non-virtual-dtor] Now despite the warning about the non virtual destructor, the correct behavior of having ~Foo() invoked works. The interesting thing is that marking the destructor of the *derived* class (virtual ~Foo) makes the warning go away. It's like it's warning me that if I ever inherit from Foo, then I might run into issues - not of any immediate problems. Should I never user shared_ptr with interface pointers that lack a virtual destructor? Always declare a virtual destructor? What is the guidance here? jrs

2 1

[boost] [compute] Review period starts today December 15, 2014, ends on December 24, 2014
by Antony Polukhin 15 Dec '14

15 Dec '14

Dear All, Review of the Compute library starts today on Mon 25st of December 2014 and will last for ten days. The Compute library provides a C++ interface to multi-core GPGPU and CPU computing platforms based on OpenCL. The project is hosted on GitHub at https://github.com/kylelutz/compute. Docs are available at http://kylelutz.github.io/compute/ Sources could be downloaded as a ZIP archive via https://github.com/kylelutz/compute/archive/master.zip Some pre-official reviews could be found in Boost Library Incubator: http://rrsd.com/blincubator.com/reviews/?library_id=669 Please, answer the following questions in your review: 1. What is your evaluation of the design? 2. What is your evaluation of the implementation? 3. What is your evaluation of the documentation? 4. What is your evaluation of the potential usefulness of the library? 5. Did you try to use the library? With what compiler? Did you have any problems? 6. How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? 7. Are you knowledgeable about the problem domain? And finally, every review should answer this question: 8. Do you think the library should be accepted as a Boost library? Be sure to say this explicitly so that your other comments don't obscure your overall opinion. -- Best regards, Antony Polukhin

1 0

[endian] Boost.Endian library added for 1.57.0
by Beman Dawes 14 Dec '14

14 Dec '14

The library provides integer and floating point types and conversion functions for portable byte ordering regardless of processor endianness. Although the primary uses are for file or network I/O, it is also useful for unaligned integers or floats, or for 3, 5, 6, and 7 byte integers. The library's predecessors have been in use for thirty years and developed independently multiple times. The repository is https://github.com/boostorg/endian The documentation is online at https://boostorg.github.io/endian/ See https://boostorg.github.io/endian/index.html#Release-history for changes since the Formal Review. No need to report "Inspection Report" problems; those will be cleaned up before 1.57.0 ships. Please do report any other problems or concerns, since the time between now and when the 1.57.0 beta ships in effect acts a mini-review period. If anyone has access to a big-endian platform with a modern C++ compiler, please, please run the endian test suite and report the results. All testing and use in recent years has been on little-endian platforms, so big-endian platform testing is a worry. Thanks for you patience, --Beman

3 2

:typedef mt19937, normal_distribution, variate_generator & data storage into vector
by It neophyte 14 Dec '14

14 Dec '14

hi all, does someone know why my vstore vector returns a 0 value? any suggestion to improve the code would be appreciated.cheers. *aim*: *generating 260 random numbers following a standard normal distribution, then store the values into a vector (here: vstore)* vector<double>vstore; vstore.resize(260); typedef mt19937 ALEA; typedef normal_distribution<double> STGAUSSIAN; typedef variate_generator<ALEA,STGAUSSIAN> GENALEA; ALEA ale; STGAUSSIAN stgauss(0,1); GENALEA gen(ale,stgauss); for (int i=0;i<=260;i++) { //cout<<gen()<<endl; //system("pause"); vstore.push_back(gen()); cout<<vstore[i]<<endl; }

2 2

[serialization] slow compile times with archive register_type
by Elizabeta 14 Dec '14

14 Dec '14

Hi I am using custom archive classes. Problem is BOOST_CLASS_EXPORT_GUID is not working with the custom archive classes, i.e I am getting unregistered class exceptions all over the place, when saving derived class with base pointer. Here is link to my question http://boost.2283326.n4.nabble.com/BOOST-CLASS-EXPORT-GUID-with-custom-arch… about this. So I found other way for exporting the classes, i.e with archive register type. Exporting the classes with register_type is working well, no more unregistered_class exceptions, but probelm is it increase the compile times of my project by more than half. Project compile time with register_type is 40-45 minutes, and without 18-19 minutes. Here is how I am using the register_type : //RegisterMyTypes.h template<class Archive> void register_all_types(Archive&); //RegisterMyTypes.cpp #include "A.h" #include "B.h" // other classes includes template<class Archive> void register_all_types(Archive& a) { ar.register_type (); ar.register_type(); ..... all other classes which is maybe 100 } // explicit instiantion of all my archive classess template void register_all_types(XmlWOArchive& ar); template void register_all_types(XmlWIArchive& ar); template void register_all_types(DbgXmlWOArchive& ar); template void register_all_types(DbgXmlWIArchive& ar); template void register_all_types(BinaryOArchive& ar); template void register_all_types(BinaryIArchive& ar); template void register_all_types(CloneBinaryOArchive & ar); template void register_all_types(CloneBinaryIArchive & ar); template void register_all_types(DuplicateBinaryOArchive & ar); template void register_all_types(DuplicateBinaryIArchive & ar); I am calling register_all_types before doing the serialization/deserialization. And I am doing the serialization/deserialization in multiple functions in my project not just in one place, but this should not be problem, I mean I have the registration done only in one cpp file i.e RegisterMyTypes.cpp, so the templates should be instantiated only in this file, and it should be slow, right ? Thank you for any help -- View this message in context: http://boost.2283326.n4.nabble.com/serialization-slow-compile-times-with-ar… Sent from the Boost - Users mailing list archive at Nabble.com.

2 4

Re: [Boost-users] Using static code checkers against the Boost code base
by Niall Douglas 13 Dec '14

13 Dec '14

On 11 Dec 2014 at 13:38, boost-users(a)lists.boost.org wrote: > I was given some time to implement automated nightly testing for > Boost.Thread which is now integrated and appears as a dashboard on > https://github.com/boostorg/thread/tree/develop. > > I'll mux in the clang and MSVC static analysers later today. This now done, and as can be seen on the Boost.Thread CI dashboard we currently see 30 of 90 build targets have problems. We'll get to fixing those over the next few days. After that we'll hopefully keep everything all blue from now on. Thanks for raising the problem, it caused resources to be made available. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

1 0

[Test] Reviving BOOST_*_CLOSE_COLLECTIONS request
by Ghyslain Leclerc 13 Dec '14

13 Dec '14

Hello, We are in the process of creating unit tests for our in-house application development and have selected Boot.Test as a library. I’ve been toying with the library for a while now and there is one feature that I feel is missing. It is the capacity to check a whole collection for close values, i.e. the “close” equivalent of BOOST_*_EQUAL_COLLECTIONS. From what I can tell, this has already been discussed (old threads from 2007 http://boost.2283326.n4.nabble.com/Boost-Test-request-BOOST-CLOSE-COLLECTIO…) and did not end-up in the library (at least from what I can tell from my version of boost (1.57). Other more recent threads (2012, http://boost.2283326.n4.nabble.com/test-BOOST-TEST-universal-testing-tool-t…) seem to suggest that the library is undergoing under the hood changes (I might be wrong here). So this request might be completely inappropriate, given the timing. Nevertheless, I have come up with an implementation for those macros. I wanted to share them here on the list (code at the end of the mail). It might be useful for some people. If Gennadiy (library maintainer) reads this mail and finds the macros adequate for use, great. If not, too bad. Anyhow, thanks. Ghyslain namespace boost_hlp { template <typename Left, typename Right> inline ::boost::test_tools::predicate_result close_coll_impl( Left lt_begin, Left lt_end, Right rt_begin, Right rt_end, ::boost::test_tools::percent_tolerance_t<double> tolerance ) { ::boost::test_tools::predicate_result res( true ); ::std::size_t pos = 0; for( ; lt_begin != lt_end && rt_begin != rt_end; ++lt_begin, ++rt_begin, ++pos ) { if( !(::boost::test_tools::check_is_close( *lt_begin, *rt_begin, tolerance )) ) { res = false; res.message() << "\nMismatch in a position " << pos << ": " << *lt_begin << " != " << *rt_begin; } } if( lt_begin != lt_end ) { std::size_t r_size = pos; while( lt_begin != lt_end ) { ++pos; ++lt_begin; } res = false; res.message() << "\nCollections size mismatch: " << pos << " != " << r_size; } if( rt_begin != rt_end ) { std::size_t l_size = pos; while( rt_begin != rt_end ) { ++pos; ++rt_begin; } res = false; res.message() << "\nCollections size mismatch: " << l_size << " != " << pos; } return res; } } // namespace boost_hlp #define CLOSE_COLLECTIONS_IMPL( L_begin, L_end, R_begin, R_end, tolerance, TL ) \ BOOST_TEST_TOOL_IMPL( \ check_impl,\ ::boost_hlp::close_coll_impl( (L_begin), (L_end), (R_begin), (R_end), \ ::boost::test_tools::percent_tolerance_t<double>((tolerance) )), "", TL, CHECK_EQUAL_COLL ), \ 5, \ BOOST_STRINGIZE( L_begin ), BOOST_STRINGIZE( L_end ), \ BOOST_STRINGIZE( R_begin ), BOOST_STRINGIZE( R_end ),\ BOOST_STRINGIZE( ::boost::test_tools::percent_tolerance_t<double>((tolerance) ) ) ) \ /**/ #define WARN_CLOSE_COLLECTIONS( L_begin, L_end, R_begin, R_end, tolerance ) \ CLOSE_COLLECTIONS_IMPL( (L_begin), (L_end), (R_begin), (R_end), (tolerance), WARN ) #define CHECK_CLOSE_COLLECTIONS( L_begin, L_end, R_begin, R_end, tolerance ) \ CLOSE_COLLECTIONS_IMPL( (L_begin), (L_end), (R_begin), (R_end), (tolerance), CHECK ) #define REQUIRE_CLOSE_COLLECTIONS( L_begin, L_end, R_begin, R_end, tolerance ) \ CLOSE_COLLECTIONS_IMPL( (L_begin), (L_end), (R_begin), (R_end), (tolerance), REQUIRE )

3 3

Re: [Boost-users] [Test] Reviving BOOST_*_CLOSE_COLLECTIONS request
by Ghyslain Leclerc 13 Dec '14

13 Dec '14

> Hello, > > We are in the process of creating unit tests for our in-house application > development and have selected Boot.Test as a library. ?I?ve been toying > with the library for a while now and there is one feature that I feel is > missing. ?It is the capacity to check a whole collection for close values, > i.e. the ?close? equivalent of BOOST_*_EQUAL_COLLECTIONS. Devel Boost.Test can do much better: BOOST_AUTO_TEST_CASE(foo) { vector<float> a, b; BOOST_TEST( a == b, tolerance(1e-5) ); BOOST_TEST( a <= b, percent_tolerance = 1e-5 ); } BOOST_TEST_DECORATOR( -tolerance = 1e-7 ) BOOST_AUTO_TEST_CASE(goo) { vector<float> a, b; BOOST_TEST( a > b ); BOOST_TEST( a != b ); } e.t.c. Gennadiy Gennadiy, thanks for pointing it out and taking the time to answer. I had read some of that in the archive threads. When is the devel Boost.Test expected to become the main version ? How can I get the devel version of the code ? Sorry if these seem like newbie questions. Again, thanks for the info. Ghyslain

2 1

Thread prioritization on OSX
by Alexander Carôt 13 Dec '14

13 Dec '14

Hello all, according to this: http://www.boost.org/doc/libs/1_57_0/doc/html/thread/thread_management.html #thread.thread_management.tutorial.attributes I tried to change the thread priority to SCHED_RR on OS X via: boost::thread::attributes attrs; pthread_attr_setschedpolicy(attrs.get_native_handle(), SCHED_RR); sendThread = boost::thread(attrs,&sender::sendItThreadFunction,this); In that context I noticed that get_native_handle is not known so I used attr.native_handle(), which compiled ok but finally threw an error here: /boost/boost/bind/bind.hpp:313:9: Type 'boost::thread_attributes' does not provide a call operator I am still using boost 1.45 and wonder if this is the actual problem or if something else is wrong. Any help appreciated, thanks Alex Am 20.11.14 10:53 schrieb "Ben Pope" unter <benpope81(a)gmail.com>: >On Thursday, November 20, 2014 05:27 PM, Alexander Carôt wrote: >> Hello all, >> >> is it correct that boost does not offer thread prioritization ? > >There's nothing directly on the interface. > >> I am dealing with an application with a number of threads, which all >> depend on the realtime-scheduled sound card callback thread. >> >> These threads must have the same realtime scheduling priority. >> >> Is there a way to achieve this ? > >Basically there's a platform specific back door: native_handle. > >http://www.boost.org/doc/libs/1_57_0/doc/html/thread/thread_management.htm >l#thread.thread_management.tutorial.attributes > >Ben > > > >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users

2 9

BOOST_CLASS_EXPORT_GUID with custom archive
by Elizabeta 13 Dec '14

13 Dec '14

Hi I am using boost 1.49. I have custom archives derived from the boost binary archives. However it seems that when using BOOST_CLASS_EXPORT_GUID the types are not actually registered for the custom archive types, and I am getting unregistered_class exception when serializing derived class from base pointer. If I am using the boost binary original archives this is not the case. My custom archive classes are defined in in its own header file and at the end of the file I am including BOOST_SERIALIZATION_REGISTER_ARCHIVE, according to this in the documentation : " The newly created archive will usually be stored in its own header module. All that is necessary is to include the header and construct an instance of the new archive. EXCEPT for one special case. Instances of a derived class are serialized through a base class pointer. Such instances are not "registered" neither implicitly nor explicitly. That is, the macro BOOT_CLASS_EXPORT is used to instantiate the serialization code for the included archives. To make this work, the following should be included after the archive class definition. #define BOOST_SERIALIZATION_REGISTER_ARCHIVE(Archive) Failure to do this will not inhibit the program from compiling, linking and executing properly - except in one case. If an instance of a derived class is serialized through a pointer to its base class, the program will throw an unregistered_class exception." here is my archive h file : ------------------- customarchive.h -------------------- #include <boost/archive/binary_iarchive.hpp> #include <boost/archive/binary_oarchive.hpp> #include <boost/archive/binary_oarchive_impl.hpp> #include <boost/archive/binary_iarchive_impl.hpp> #include <boost/archive/shared_ptr_helper.hpp> class CloneBinaryIArchive : // don't derive from binary_oarchive !!! public boost::archive::binary_iarchive_impl< CloneBinaryIArchive, std::istream::char_type, std::istream::traits_type >, public boost::archive::detail::shared_ptr_helper { typedef CloneBinaryIArchive derived_t; typedef boost::archive::binary_iarchive_impl< CloneBinaryIArchive, std::istream::char_type, std::istream::traits_type > base_t; public: CloneBinaryIArchive(std::istream & is,unsigned int flags = 0) : base_t(is, flags) {} CloneBinaryIArchive(std::streambuf & bsb, unsigned int flags = 0) : base_t(bsb, flags) {} int foo; }; BOOST_SERIALIZATION_REGISTER_ARCHIVE(CloneBinaryIArchive) and here is the cpp file in each I am registering all classes : -------------- register.cpp --------------- #include "customarchive.h" #include <boost/serialization/export.hpp> #include "A.h" BOOST_CLASS_EXPORT_GUID(A,"A") -- View this message in context: http://boost.2283326.n4.nabble.com/BOOST-CLASS-EXPORT-GUID-with-custom-arch… Sent from the Boost - Users mailing list archive at Nabble.com.

2 1

share_ptr and Container compilation error on new version
by JB 13 Dec '14

13 Dec '14

Hello all, I'm trying to build an application. This isn't an app that I wrote but rather inherited. I'm currently porting it from Fedora 15 to Fedora 20. It compiled well on gcc 4.6.3 with boost 1.46.0. However, Fedora 20 has gcc 4.8.3 and boost 1.54. I'm new to boost and have never used it before but it looks very comprehensive. As far as I know boost/shared_ptr is the only header this component of the application is including. I'm wondering if boost changes might be causing this or maybe gcc changes but it's difficult to tell from the documentation and I'm a little under the gun at the moment. This code: template<typename Container> class DefinableVectorFinalizer: public XmlFinalizer { typedef typename Container::value_type contained_value_type; typedef typename Container::value_type::value_type value_type; .... } g++ is throwing this error which seems to have some sort of problem with Container::value_type::value_type: ../../../../../devel/include/xcl/io/DefinableVectorFinalizer.h:31:54: error: no type named ‘value_type’ in ‘xclApp::DefinableVector<boost::shared_ptr<simApp::Project> >::value_type {aka class boost::shared_ptr<mainApp::Project>}’ typedef typename Container::value_type::value_type value_type; xclApp::DefinableVector looks like this: class DefinableVector { public: typedef typename std::vector<T>::size_type size_type; typedef typename std::vector<T>::value_type value_type; typedef value_type& reference; typedef const value_type& const_reference; typedef typename std::vector<T>::iterator iterator; typedef typename std::vector<T>::const_iterator const_iterator; .... } Any insight into this would be hugely appreciated!!!

3 4

intrusive bug: equal_range implementation vs bounded_range precondition
by Matei David 12 Dec '14

12 Dec '14

In the absence of a bug tracker, I'll post the potential bug here: I'm seeing some strange results out of `multiset::equal_range`. I didn't isolate a test case, but looking at the code something seems wrong: The `bstree_algorithms::equal_range(const KeyType &, KeyValueCompare)` method is currently implemented (in `bstree_algorithms.hpp`) as a call to `bstree_algorithms::bounded_range(header, key, key, comp, true, true)`. However, `bounded_range` states as prerequisite: If 'lower_key' == 'upper_key', ('left_closed' or 'right_closed') must be false. Either the `equal_range` implementation or the `bounded_range` precondition is wrong. Matei

2 3

[Math] advice using brent_find_minima()
by dariomt＠gmail.com 12 Dec '14

12 Dec '14

I'd like to find the root of a function f (find x such that f(x) = 0). Should I pass abs( f ) to brent_find_minima()? I can't find examples... Thanks in advance!

4 6

bug tracker not working
by Matei David 12 Dec '14

12 Dec '14

The bug tracker stopped working for me. I tried to post a new bug, and I'm getting a captcha error message below. However, the web interface doesn't display any captcha for me to solve (!?) I tried both Google Chrome & Firefox on Linux. (I disabled various ad blocker extensions.) === Captcha Error Submission rejected as potential spam (StopForumSpam says this is spam (ip)) Trac thinks your submission might be Spam. To prove otherwise please provide a response to the following. === I've seen this message before, with the exact same problem (no captcha), when in the past I had tried to include a (legitimate) link in a bug post. However, after obfuscating the link I was able to post without this error. Now I can't any more. Can somebody look into this? Thanks, Matei

1 0

Flags for strongly typed enum?
by TONGARI J 11 Dec '14

11 Dec '14

Hi there, Is there any library in Boost that provides type-safe enum bit flags? Something like below: enum class option { a = 1, b = 2, c = 4, .... }; DEFINE_FLAGS(options, option); options set1 = option::a | option::b; options set2 = set1 | option::c; options err1 = option::a | 2; // error options err2 = set2 | 2; // error if (set1 & option::a) {...} if (set1 & 2) {...} // error Thanks.

1 0

boost MSM
by Raj Sambasivan 10 Dec '14

10 Dec '14

Hi Christophe, I have been using boost msm, i have two state machines each with around 30 transitions / 4 regions , no submachine etc for now... it perfectly fits the bill on what im trying to do, thanks for the great library, one issue i face is ,the debug build takes for ever to compile and the size is very big (the optimized one with no debug symbol compiles pretty fast) and when errors happens the compiler (template barf) goes for pages and pages .. any hints on this please? thanks -raj

4 5