November 2013 - Boost - lists.preview.boost.org

milliseconds UTC from discrete year, month, day, hour, minute, second
by Victor Yankee 22 Nov '13

22 Nov '13

Hello, Can someone should me how to calculate the milliseconds since jan 1, 1970 in UTC, given separate variables holding the year, month,day,hour,minute and seconds? Thanks,

3 2

[MPL] Reason for mpl::quote's name
by Louis Dionne 22 Nov '13

22 Nov '13

Hi, I was just wondering about the origin of the `mpl::quote` metafunction class. Why was this name chosen? Thanks, Louis Dionne

4 5

[Boost-build] Android tests run
by Antony Polukhin 22 Nov '13

22 Nov '13

Hi all, With help of Steven Watanabe, Vladimir Prus and Boost-for-Android project, I've finally run the regression tests on Android ARM. Results can be already seen on the regression test page for trunk. Here is a description of some common issues that can be seen in regression tests results: "arm-linux-androideabi/bin/ld.exe: error: cannot find -lrt" Some of the Android libraries (for example librt) are merged with libc. Some patches to b2 required not to add -lrt fot the platform. "soinfo_link_image(linker.cpp:1635): could not load library "libboost_unit_test_framework-gcc-d-1_55.so.1.55.0" needed by "./copy_if_test1"; caused by load_library(linker.cpp:745): library "libboost_unit_test_framework-gcc-d-1_55.so.1.55.0" not found" This is because of cross compilation - tests are compiled in Win and then copied to the emulator. This is an error in my copying script that will be fixed someday. Also please note, that sources are compiled with -DBOOST_ASIO_DISABLE_STD_ATOMIC and -DBOOST_AC_USE_PTHREADS because there are some issues with std::atomic on android (See https://code.google.com/p/android/issues/detail?id=42735#makechanges ) This must be possibly added to the config of the libraries by default. Some hints: * You can check for android platform in your code using defined(__ANDROID__). * __ANDROID__ can run on different hardware, so do not assume that it will be always ARM. * Having problems with Jamfile and cross compilation? Use <target-os> instead of <host-os> in your jamfiles. * Boost-for-Android project ( https://github.com/MysticTreeGames/Boost-for-Android) may contain some useful fixes for your library. Author of the library "would love" to see the issues integrated, see discussion at https://github.com/MysticTreeGames/Boost-for-Android/pull/15 This letter is cc'ed to the boost-build because of: * "-lrt" issues * Is there a way to install/call_script for all the boost libraries before running the regression tests? * specifying "<threadapi>pthread" in user-config.jam does not work -- Best regards, Antony Polukhin

1 0

[chrono] boost::chrono::floor() and negative durations
by Krzysztof Czainski 22 Nov '13

22 Nov '13

Hello, The function boost::chrono::floor() is documented: "This function round down the given parameter." Therefore I expect the assert in the following program to pass, but it fails: #include <boost/chrono/floor.hpp> #include <cassert> int main() { boost::chrono::nanoseconds const nsec( -1 ); boost::chrono::seconds const sec = boost::chrono::floor<boost::chrono::seconds>(nsec); assert( sec.count() == -1 ); } Is this intended behavior? If yes, could the docs be updated to clarify this? Regards, Kris

3 6

Fwd: Ask for a solution regarding to the function intersection() in boost geometry
by Jun Zhang 21 Nov '13

21 Nov '13

Dear Sir/Madam, Many thanks for your work on boost library. I am a user of the library. But now I meet a problem regarding to the function intersection. Could you please see the attached file. When I get the intersection of the polygon "red" and "blue" with the function, the output is the "green" (only one polygon). However, "green" should be two polygons. And if the green is regarded as one polygon, there is some error to get the intersection of it with others. That means I cannot calculate the intersection of "green" and "blue" with intersection(). The error is as following: This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information. terminate called after throwing an instance of 'boost::geometry::overlay_invalid_input_exception' what(): Boost.Geometry Overlay invalid input exception All the best -- Dr.-Ing. Jun Zhang Jülich Supercomputing Centre Institute for Advanced Simulation phone +49 2461 61 1995 E-Mail: ju.zhang(a)fz-juelich.de<mailto:ju.zhang@fz-juelich.de> Internet: http://www.fz-juelich.de/ias/jsc ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------ Forschungszentrum Juelich GmbH 52425 Juelich Sitz der Gesellschaft: Juelich Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender), Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt, Prof. Dr. Sebastian M. Schmidt ------------------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------------------

1 0

Problems building boost.python against python 3.x on Windows
by Fredrik Orderud 21 Nov '13

21 Nov '13

There seem to be a problem building boost.python against python 3.x on Windows. Steps to reproduce: * Download & unzip boost 1.55 * bootstrap.bat * set PYTHONHOME=C:\Python33 * b2 link=shared runtime-link=shared toolset=msvc-11.0 b2 will then output the following: --- warning: No python installation configured and autoconfiguration note: failed. See http://www.boost.org/libs/python/doc/building.html --- This does not happen if building against Python 2.7 with set PYTHONHOME=C:\Python27. I suspect that the reason is due to tools\build\v2\engine\build.jam only containing logic for python 2.2-2.7. Could anyone please look into this? Thanks in advance, Fredrik

1 0

[Config] N3276 decltype support detection missed for VC++12 (Ticket #9410)
by Michel Morin 20 Nov '13

20 Nov '13

As you know, VC++12 (aka 2013) completed N3276 decltype support. (See, e.g. http://msdn.microsoft.com/en-us/library/vstudio/hh567368.aspx http://thread.gmane.org/gmane.comp.lib.boost.devel/233752/focus=239118 ) Unfortunately the detection code for VC++12's N3276 decltype support hasn't been updated yet, as reported in #9410 (decltype N3276 for MSVC12 + compile fix for tests) https://svn.boost.org/trac/boost/ticket/9410 Ticket #9410 has a patch for this issue. The patch also fixes the test code of Boost.Config. Regards, Michel

1 0

Re: [boost] decay.hpp problem
by Tim Burgess 20 Nov '13

20 Nov '13

OK, Here's the output from the Log Navigator, showing how includes are being stacked. I've placed it at: https://dl.dropboxusercontent.com/u/29578720/Build%20log.rtf I noted the references to wxAny, which I was using for handling variants, but I've now replaced this with boost::any and I'm not including wxAny in any of my modules, despite the log seeming to implicate this class. Again, the new version of the code builds fine using VS2010, but fails using Xcode. Best wishes. Tim Burgess -----Original Message----- From: Steve Ramsey [mailto:steve@lucena.com] Sent: 18 November 2013 18:59 To: tim(a)raisedbar.net Subject: Re: [boost] decay.hpp - anonymous enum problem On Nov 18, 2013, at 7:37 AM, Tim Burgess <tim(a)raisedbar.net> wrote: > I'm pretty much a newbie when it comes to Xcode, so I've no idea how > to get you that trace - the Issue Navigator gives me no clues as to > what's including/calling decay.hpp. Any clues/links to reading > materials would be most welcome. I find the Log Navigator using the Assistant editor pane to be more flexible than the Issue Navigator when dealing with failed build detritus. So, from the Log Navigator, select your failed build. In the editor pane, you'll find an expandable list of build issues broken down by file. On the far right side of each file-level entry, there is a "show transcript" widget; popup text shows the internal command that was issued for the file and when it was issued. Clicking the widget will show a complete dump of the compiler/linker output for that particular command. It's also possible to right-click on any given issue entry and copy the transcript for either just the current issue or for the entire filtered list of issues. Steve

3 3

VC2013 1.55 beta failures
by Olaf van der Spek 19 Nov '13

19 Nov '13

Hi, Building 1.55 beta keeps failing on VC2013 RC. Trunk fails too IIRC. compile-c-c++ bin.v2\libs\date_time\build\msvc-12.0\debug\link-static\threading-multi\gregorian\greg_month.obj greg_month.cpp C:\VC\boost_1_55_0b1\boost/iterator/detail/facade_iterator_category.hpp(166) : error C2039: 'assert_not_arg' : is not a member of 'boost::mpl' C:\VC\boost_1_55_0b1\boost/mpl/eval_if.hpp(41) : see reference to class template instantiation 'boost::detail::facade_iterator_category_impl<CategoryOrTraversal,ValueParam,Reference>' being compiled with [ CategoryOrTraversal=boost::forward_traversal_tag , ValueParam=std::basic_string<char,std::char_traits<char>,std::allocator<char>> , Reference=const std::basic_string<char,std::char_traits<char>,std::allocator<char>> & ] C:\VC\boost_1_55_0b1\boost/iterator/detail/facade_iterator_category.hpp(193) : see reference to class template instantiation 'boost::mpl::eval_if<boost::detail::is_iterator_category<CategoryOrTraversal>,boost::mpl::identity<boost::forward_traversal_tag>,boost::detail::facade_iterator_category_impl<CategoryOrTra versal,ValueParam,Reference>>' being compiled with [ CategoryOrTraversal=boost::forward_traversal_tag , ValueParam=std::basic_string<char,std::char_traits<char>,std::allocator<char>> , Reference=const std::basic_string<char,std::char_traits<char>,std::allocator<char>> & ] C:\VC\boost_1_55_0b1\boost/iterator/iterator_facade.hpp(104) : see reference to class template instantiation 'boost::detail::facade_iterator_category<CategoryOrTraversal,ValueParam,Reference>' being compiled with [ CategoryOrTraversal=boost::forward_traversal_tag , ValueParam=std::basic_string<char,std::char_traits<char>,std::allocator<char>> , Reference=const std::basic_string<char,std::char_traits<char>,std::allocator<char>> & ] C:\VC\boost_1_55_0b1\boost/iterator/iterator_facade.hpp(620) : see reference to class template instantiation 'boost::detail::iterator_facade_types<Value,CategoryOrTraversal,Reference,Difference>' being compiled with [ Value=std::basic_string<char,std::char_traits<char>,std::allocator<char>> , CategoryOrTraversal=boost::forward_traversal_tag , Reference=const std::basic_string<char,std::char_traits<char>,std::allocator<char>> & , Difference=ptrdiff_t ] C:\VC\boost_1_55_0b1\boost/token_iterator.hpp(40) : see reference to class template instantiation 'boost::iterator_facade<boost::token_iterator<TokenizerFunc,Iterator,Type>,Type,boost::forward_traversal_tag,const Type &,ptrdiff_t>' being compiled with [ TokenizerFunc=char_separator_type , Iterator=std::_String_const_iterator<std::_String_val<std::_Simple_types<char>>> , Type=std::basic_string<char,std::char_traits<char>,std::allocator<char>> ] C:\VC\boost_1_55_0b1\boost/date_time/date_parsing.hpp(132) : see reference to class template instantiation 'boost::token_iterator<TokenizerFunc,Iterator,Type>' being compiled with [ TokenizerFunc=char_separator_type , Iterator=std::_String_const_iterator<std::_String_val<std::_Simple_types<char>>> , Type=std::basic_string<char,std::char_traits<char>,std::allocator<char>> ] C:\VC\boost_1_55_0b1\boost/date_time/gregorian/parsers.hpp(30) : see reference to function template instantiation 'date_type boost::date_time::parse_date<boost::gregorian::date>(const std::string &,int)' being compiled with [

13 52

Interesting variadic template trick
by Louis Dionne 19 Nov '13

19 Nov '13

Hi, As some might know, I'm currently working on a reimplementation of the MPL for C++11. In doing so, I came across a technique that I did not know of and I thought I would share my discovery with the community. If this is already well known, please excuse my ignorance. The goal is to implement fast compile-time conjunction and disjunction. It goes like this: template <typename ...T> void allow_expansion(T&&...); template <typename ...T> struct or_ { static constexpr bool value = !noexcept( allow_expansion((T::value ? throw : 0)...) ); }; That's it! Note that disjunction is easily obtained in a similar way. The benefit of this technique over traditional implementations is that it's several times faster according to my benchmarks. However, you might have noticed that short-circuit evaluation is not respected so the above technique is only suitable when short-circuit evaluation is not required. If you find this interesting, see [1] for a collection of similar techniques. The one I posted above seems to be the fastest with Clang and GCC. Regards, Louis Dionne [1] https://gist.github.com/ldionne/7522393

7 12