May 2018 - Boost - lists.preview.boost.org

dynamic_bitset needs some attention
by James E. King, III 16 May '18

16 May '18

I have been adding a common CI infrastructure to the CMT repositories. Occasionally I have to disable one or two build jobs to get a clean build and I file an issue. When I got to dynamic_bitset however, there's clearly something going on in there and it isn't good. I think it has to do with streaming. It's probably just one issue causing most of the failures. You can see the results from my latest dry-run PR here: https://github.com/jeking3/dynamic_bitset/pull/1 Right now both the main travis and all of the appveyor builds are failing, and they all look like they get stuck on test 4 in iostream processing. Something just isn't right in there - it could be the test itself or it could be something else. I may submit a PR into the repo for the CI additions anyway even if they are failing, since it points out this component has rotted and needs attention. Also if my PRs in pool and function can be merged, we'll have better CI on those. Aiming to get all the CMT repos covered with appveyor, travis, codecov, and eventually covscan. - Jim

1 0

Clarifying Boost evolution plans mentioned on Twitter
by Bryce Adelstein Lelbach aka wash 16 May '18

16 May '18

Hi all, Some pictures from the C++Now "Boost Library In A Week" end-of-week summary were posted to Twitter today by an attendee, along with some mistaken statements suggesting that "Boost plans to move to C++11". I want to clarify what is actually going on: 0.) During the week, the Library In A Week group discussed some ideas for how Boost might migrate to newer standards. I'm somewhat to blame as I initiated these discussions. 1.) Those ideas were collected into a draft document by Jeff Garland. 2.) During the closing panel, Jeff presented the draft document and stated the next step would be to share it with the mailing list. 3.) Due to an innocent misunderstanding, an attendee posted a picture of part of the document with the unclear message "Boost plans to move to C++11". 4.) I wrote this email. I have asked the attendee to delete the tweet (and repost with a clearer message if they wish). I don't want to share the document from LIAW. Jeff put a lot of work into it and I don't want to steal his thunder (I realize this email inevitably does that). I'm sure he will follow up with a post explaining what ideas we came up with soon. I'm very sorry for the confusion, and I'm sure the attendee is as well. This was just an innocent mixup. ------------------------------------------------------ Bryce Adelstein Lelbach aka wash ISO C++ Committee Member CppCon and C++Now Program Chair Thrust Maintainer, HPX Developer CUDA Convert and Reformed AVX Junkie Sleep is for the weak ------------------------------------------------------

4 3

[interprocess]Request to support wchar_t* file name on Windows
by 孙朝阳 15 May '18

15 May '18

Hi, I'm using boost.interprocess.file_mapping on Windows, and I need to use it to open files with localized file path, so I want file_mapping can support wchar_t file name on Windows. Does any reason prevent file_mapping to support wchar_t file name? If no, would someone like to add wchar_t supporting for file_mapping? I also created a pull request for this: https://github.com/boostorg/interprocess/pull/28 Thanks, Jackson Sun

1 0

boost libs compiled for gcc toolset with boost 1.67 and CMake 3.11 won't find libs
by Demetri Kopsas 12 May '18

12 May '18

Hi, I discovered that the latest version of boost does not work with CMake when using the GCC toolset. I submitted this bug: https://gitlab.kitware.com/cmake/cmake/issues/17974#note_408614 Using MSVC as the toolset is fine on windows, but using mingw (gcc compiler) cannot find the libs unless I take out ‘x32’ or ‘x64’ from the name after compilation.

2 1

[build] recent issues w. VS 2017
by Raffi Enficiaud 12 May '18

12 May '18

Dear all, I am experiencing an issue with VS 2017 (15.6.6) while compiling boost.test on my agent. Traces below. I do not recall having applied any change to this agent apart from installing C++/CLI support for CMake people. Is this problem known? Thanks, Raffi build 01-May-2018 11:43:49 D:\bamboo_build_dir\SW-BCB161-VC2017WIN\library\libs\test\test>..\..\..\b2 --abbreviate-paths toolset=msvc-14.1 cxxflags="/std:c++latest" build 01-May-2018 11:43:50 Performing configuration checks build 01-May-2018 11:43:50 build 01-May-2018 11:43:50 - default address-model : 32-bit (cached) build 01-May-2018 11:43:50 - default architecture : x86 (cached) build 01-May-2018 11:43:50 - symlinks supported : no (cached) build 01-May-2018 11:43:50 - junctions supported : yes (cached) build 01-May-2018 11:43:50 - hardlinks supported : yes (cached) build 01-May-2018 11:43:51 - BOOST_COMP_GNUC >= 4.3.0 : no build 01-May-2018 11:43:54 - Boost.Config Feature Check: cxx11_auto_declarations : no build 01-May-2018 11:43:54 - Boost.Config Feature Check: cxx11_function_template_default_args : no build 01-May-2018 11:43:54 - Boost.Config Feature Check: cxx11_lambdas : no build 01-May-2018 11:43:55 - Boost.Config Feature Check: cxx11_hdr_initializer_list : no build 01-May-2018 11:43:55 - Boost.Config Feature Check: cxx11_template_aliases : no build 01-May-2018 11:43:55 - Boost.Config Feature Check: cxx11_decltype : no build 01-May-2018 11:43:55 - Boost.Config Feature Check: cxx11_hdr_random : no build 01-May-2018 11:43:55 - Boost.Config Feature Check: cxx11_hdr_tuple : no build 01-May-2018 11:43:55 - Boost.Config Feature Check: cxx11_variadic_macros : no build 01-May-2018 11:43:55 - Boost.Config Feature Check: cxx11_trailing_result_types : no build 01-May-2018 11:43:56 - Boost.Config Feature Check: cxx11_nullptr : no build 01-May-2018 11:43:56 - Boost.Config Feature Check: cxx11_unified_initialization_syntax : no build 01-May-2018 11:43:56 - Boost.Config Feature Check: cxx11_variadic_templates : no build 01-May-2018 11:43:57 - Boost.Config Feature Check: cxx11_decltype : no build 01-May-2018 11:43:57 - Boost.Config Feature Check: cxx11_hdr_random : no build 01-May-2018 11:43:57 - Boost.Config Feature Check: cxx11_hdr_tuple : no build 01-May-2018 11:43:57 - Boost.Config Feature Check: cxx11_hdr_initializer_list : no build 01-May-2018 11:43:57 - Boost.Config Feature Check: cxx11_variadic_macros : no build 01-May-2018 11:43:57 - Boost.Config Feature Check: cxx11_trailing_result_types : no build 01-May-2018 11:43:57 - Boost.Config Feature Check: cxx11_template_aliases : no build 01-May-2018 11:43:57 ...patience... build 01-May-2018 11:43:57 ...found 3946 targets... build 01-May-2018 11:43:57 ...updating 553 targets... build 01-May-2018 11:43:58 compile-c-c++ ..\..\..\bin.v2\libs\test\test\string_cast-test.test\msvc-14.1\dbg\thrd-mlt\utils-ts\string_cast-test.obj build 01-May-2018 11:43:58 string_cast-test.cpp build 01-May-2018 11:43:58 D:\bamboo_build_dir\SW-BCB161-VC2017WIN\library\boost/config/detail/select_stdlib_config.hpp(18): fatal error C1083: Cannot open include file: 'cstddef': No such file or directory build 01-May-2018 11:43:58 build 01-May-2018 11:43:58 call "..\..\..\bin.v2\standalone\msvc\msvc-14.1\msvc-setup.bat" >nul build 01-May-2018 11:43:58 cl /Zm800 -nologo @"..\..\..\bin.v2\libs\test\test\string_cast-test.test\msvc-14.1\dbg\thrd-mlt\utils-ts\string_cast-test.obj.rsp" build 01-May-2018 11:43:58 build 01-May-2018 11:43:58 ...failed compile-c-c++ ..\..\..\bin.v2\libs\test\test\string_cast-test.test\msvc-14.1\dbg\thrd-mlt\utils-ts\string_cast-test.obj... build 01-May-2018 11:43:58 compile-c-c++ ..\..\..\bin.v2\libs\test\build\msvc-14.1\dbg\lnk-sttc\thrd-mlt\compiler_log_formatter.obj

3 5

new b2 property or new boost library?
by Oliver Kowalke 11 May '18

11 May '18

Hi, boost.fiber provides some rudimentary NUMA support. NUMA specific OS APIs require the native thread handle (native_handle_type std::thread::native_handle()). But unfortunately the returned type does not match in all cases the OS handle type. For instance MinGW-W64 using the POSIX thread impl returns long long instead of void*. A solution would be to add an additional b2 property like `numa=on` that compiles the NUMA related part of boost.fiber if set (the user is responsible to compile boost with the appropriate properties). Or the NUMA related code is move from boost.fiber into a complete new boost library. What would be you advice? Oliver

1 0

[type_traits] revisiting is_virtual_base_of
by John Maddock 10 May '18

10 May '18

There is a problem with is_virtual_base_of in that it doesn't work especially well - most of the time it works OK, but can be made to fail catastrophically in a few situations, see for example: https://svn.boost.org/trac10/ticket/11323 Now... it turns out that with C++14 and later we now have is_detected which can test the validity of almost any expression... and I'm wondering if there's a cunning way to press this into action to create a more accurate is_virtual_base_of, but I'm having trouble coming up with anything. Any bright ideas? Note that an explicit static_cast can't be tested for, though we can use is_constructible if required. Thanks in advance, John. --- This email has been checked for viruses by AVG. http://www.avg.com

2 3

build failure in develop: chrono, type_traits?
by James E. King, III 08 May '18

08 May '18

The only change in my pull request was to remove a deprecated, unrelated header in uuid. This happened in the builds last night on develop: In file included from ./boost/config.hpp:61:0, from ./boost/chrono/config.hpp:15, from ./boost/chrono/detail/inlined/process_cpu_clocks.hpp:15, from libs\chrono\src\process_cpu_clocks.cpp:17: ./boost/type_traits/detail/common_arithmetic_type.hpp:209:27: error: duplicate 'const' BOOST_STATIC_CONSTANT(const int, selector = sizeof(select(cond() ? T() : U()))); ^ "x86_64-w64-mingw32-g++.exe" -std=c++03 -m64 -O0 -fno-inline -Wall -pedantic -g -Wextra -Wno-long-long -DBOOST_ALL_NO_LIB=1 -DBOOST_All_STATIC_LINK=1 -DBOOST_SYSTEM_STATIC_LINK=1 -I"." -c -o "bin.v2\libs\chrono\build\gcc-mingw64\debug\cxxstd-03-iso\link-static\process_cpu_clocks.o" "libs\chrono\src\process_cpu_clocks.cpp" ...failed gcc.compile.c++ bin.v2\libs\chrono\build\gcc-mingw64\debug\cxxstd-03-iso\link-static\process_cpu_clocks.o... Failed on cygwin and mingw builds: https://ci.appveyor.com/project/jeking3/uuid-gaamf/build/1.0.116-develop/jo… Also failed in c++03 builds on unix: https://travis-ci.org/boostorg/uuid/jobs/376161574#L2158 gcc.compile.c++ ../../bin.v2/libs/chrono/build/gcc-7/release/cxxstd-03-iso/link-static/thread_clock.o In file included from ../../boost/config.hpp:61:0, from ../../boost/chrono/config.hpp:15, from ../../boost/chrono/detail/inlined/thread_clock.hpp:14, from ../../libs/chrono/src/thread_clock.cpp:18: ../../boost/type_traits/detail/common_arithmetic_type.hpp:209:27: error: duplicate ‘const’ BOOST_STATIC_CONSTANT(const int, selector = sizeof(select(cond() ? T() : U()))); ^ "g++-7" -std=c++03 -m64 -O3 -finline-functions -Wno-inline -Wall -pedantic -Wextra -Wno-long-long -DBOOST_ALL_NO_LIB=1 -DBOOST_All_STATIC_LINK=1 -DBOOST_SYSTEM_STATIC_LINK=1 -DNDEBUG -I"../.." -c -o "../../bin.v2/libs/chrono/build/gcc-7/release/cxxstd-03-iso/link-static/thread_clock.o" "../../libs/chrono/src/thread_clock.cpp" ...failed gcc.compile.c++ ../../bin.v2/libs/chrono/build/gcc-7/release/cxxstd-03-iso/link-static/thread_clock.o... Thanks, Jim

2 1

[filesystem] and C++17 std::filesystem
by Egor Pugin 07 May '18

07 May '18

Hi! gcc8.1 and VS15.7 (2017) are now shipped with std::filesystem. I'd like to raise a discussion about pure c++ interoperability with boost(::filesystem). With boost-1.67 it's not possible to do such things: #if __has_include(<filesystem>) #include <filesystem> namespace fs = std::filesystem; #else #include <boost/filesystem.hpp> namespace fs = boost::filesystem; #endif Because of boost::filesystem::copy_option::overwrite_if_exists != std::filesystem::copy_options::overwrite_existing, std::fs::path() / boost::fs::unique_path(), directory iterators etc. Next, if we take boost::process, for example, we see that it currently operates only on boost::filesystem. If my application uses c++17 fs, I have to write something like this: fs::path f(fs::path in) { bp::search_path(in.u8string()).wstring(); } So, mixing pure c++ and boost becomes more complex. Any ideas how to deal better with such issues on the user side? What are directions of proper fixing this in boost::process (in case I'll try to prepare a PR)? What are boost policies in adopting standard features in general? -- Egor Pugin

2 1

Boost Graph Library(biconnected_components) issue
by Meghana.M Reddy 06 May '18

06 May '18

Hi, I am working on the Boost Graph Library, currently trying to create an interface of the biconnected_components method to a Python code. I used the example code available here <https://www.boost.org/doc/libs/1_67_0/libs/graph/example/biconnected_compon…> for the method and modified it for my purpose. The template of the method is here <https://www.boost.org/doc/libs/1_67_0/libs/graph/doc/biconnected_components…>. The final code seems fine but I get an error while compiling it. And the errors are related to adjacency.hpp which I have not changed so I'm confused as to what the exact problem is. Hoping someone can help me out. My code: class BoostGraph { private: typedef typename boost::adjacency_list< OutEdgeListS, VertexListS, DirectedS, boost::property<boost::vertex_index_t, v_index>, EdgeProperty, boost::no_property, EdgeListS> adjacency_list; typedef typename boost::property_map<adjacency_list, boost::edge_index_t>::type components; public: adjacency_list graph; std::vector<v_index> bc_tree() { components components_bc = get(boost::edge_index, graph); std::size_t num_comps = biconnected_components(graph, components_bc); std::cerr << "Found " << num_comps << " biconnected components.\n"; std::vector<vertex_descriptor> art_points; articulation_points(graph, std::back_inserter(art_points)); std::cerr << "Found " << art_points.size() << " articulation points.\n"; std::vector<v_index> to_return; return to_return; } The error I'm getting: build/cythonized/sage/graphs/base/boost_graph.cpp:3730:26: required from here /home/meghana/Documents/sage_gsoc/sage_may/sage/local/include/boost/graph/detail/adjacency_list.hpp:2696:29: error: forming reference to void typedef value_type& reference; ^ /home/meghana/Documents/sage_gsoc/sage_may/sage/local/include/boost/graph/detail/adjacency_list.hpp:2697:35: error: forming reference to void typedef const value_type& const_reference; ^ /home/meghana/Documents/sage_gsoc/sage_may/sage/local/include/boost/graph/detail/adjacency_list.hpp:2701:61: error: forming reference to void typename Graph::vertex_descriptor,Property,Tag> type; ^ /home/meghana/Documents/sage_gsoc/sage_may/sage/local/include/boost/graph/detail/adjacency_list.hpp:2704:68: error: forming reference to void typename Graph::vertex_descriptor,const Property, Tag> const_type; ^ error: command 'gcc' failed with exit status 1 Makefile:33: recipe for target 'sage' failed make: *** [sage] Error 1 Thank you, Meghana

1 0

Fw: Mysterious Boost.System-related errors when running Beast tests
by Peter Dimov 05 May '18

05 May '18

The story starts with https://travis-ci.org/vinniefalco/beast/jobs/375049193 which gives the error dyld: Symbol not found: __ZN5boost6system6detail19system_category_ncxEv Referenced from: /Users/travis/build/vinniefalco/boost-root/bin.v2/libs/beast/test/beast/core/clang-darwin-4.2.1/debug/threadapi-pthread/threading-multi/fat-tests Expected in: /usr/local/lib/libboost_system.dylib boost::system::detail::system_category_ncx is a new function (in develop), and careful reading of the above reveals that the compiler is trying to link to a preinstalled libboost_system.dylib, in /usr/local/lib. Looking at the log, Boost.System is not even being built. This is odd; the Beast root Jamfile contains <library>/boost/system//boost_system (at https://github.com/boostorg/beast/blob/develop/Jamfile#L78) but even if it didn't, it also links to Filesystem, which _is_ being built, and Filesystem also depends on System in its build/Jamfile: https://github.com/boostorg/filesystem/blob/develop/build/Jamfile.v2#L29 So Boost.System should be built, but it isn't. Its Jamfile hasn't been touched in 10 years: https://github.com/boostorg/system/blob/develop/build/Jamfile.v2 In addition, there's not even a `mklink-or-dir boost/system` in the log, even though the link must be created because the new headers are being used (this could be a separate issue). The Linux/g++ job, https://travis-ci.org/vinniefalco/beast/jobs/375049187, also doesn't seem to build Boost.System, but it works for some reason. When I try `b2 libs/beast/test` on Windows, with msvc-14.1, Boost.System is being built, in the very beginning: ...patience... ...patience... ...patience... ...found 6992 targets... ...updating 1090 targets... link.hardlink ..\..\boost\utility\string_view.hpp Hardlink created for ..\..\boost\utility\string_view.hpp <<===>> ..\utility\include\boost\utility\string_view.hpp compile-c-c++ ..\..\bin.v2\libs\beast\test\beast\core\type_traits.test\msvc-14.1\debug\threadapi-win32\threading-multi\type_traits.obj type_traits.cpp compile-c-c++ ..\..\bin.v2\libs\beast\test\beast\core\type_traits.test\msvc-14.1\debug\threadapi-win32\threading-multi\main.obj main.cpp compile-c-c++ ..\..\bin.v2\libs\system\build\msvc-14.1\debug\threading-multi\error_code.obj error_code.cpp msvc.link.dll ..\..\bin.v2\libs\system\build\msvc-14.1\debug\threading-multi\boost_system-vc141-mt-gd-x32-1_66.dll (The 1.66 version instead of 1.68 was caused by a typo in the superproject which is now fixed.) When I try the same on Ubuntu Trusty, the result is the same as on Travis; no `mklink-or-dir boost/system`, no building of Boost.System, yet everything compiles and works, and I don't have a preinstalled Boost on my VM. It makes no sense at all. Does anyone have any ideas what might be happening here?

1 1

[release][website] Help wanted
by Daniel James 01 May '18

01 May '18

Hello all, For the last couple of releases, I've been handling the releases mostly on my own, but I'm starting a new job and I'm going to have a lot less time, or maybe even no time, for Boost. So we really need more people helping with release management and updating the website. The main release manager responsibilities are described here: https://svn.boost.org/trac10/wiki/ReleasePractices/ManagerCheckList It's probably best if there are a few people involved. I don't think all of them will need admin permissions. For updating the website, you mainly need to occasionally edit html/quickbook files, accept pull requests to the release notes, and use the php scripts to updates the site at release time (I should be able to help a little there). thanks, Daniel

6 5