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

Fwd: [serialization] is the weak_ptr parent kept inside an iarchive
by ypotin 06 May '20

06 May '20

Hello, I was running some tests involving std::weak_ptr using boost serialization, and I realized that when I read a std::weak_ptr from an archive that doesn't contain the "parent" std::share_ptr, the behavior of the read pointer is "hard to understand". Here is an example : #include <cassert> #include <sstream> #include <memory> #include <boost/archive/binary_iarchive.hpp> #include <boost/archive/binary_oarchive.hpp> #include <boost/serialization/export.hpp> #include <boost/serialization/shared_ptr.hpp> #include <boost/serialization/weak_ptr.hpp> class MyInt { public: MyInt() : value(0) { } int value; template<class Archive> void serialize(Archive &ar, unsigned int) { ar & value; } }; int main() { std::stringstream data; { std::shared_ptr<MyInt> shared = std::make_shared<MyInt>(); assert(shared.use_count() == 1); shared->value = 12; /* Write a weak_ptr */ { std::weak_ptr<MyInt> weak(shared); assert(shared.use_count() == 1); boost::archive::binary_oarchive ar(data); ar << weak; assert(shared.use_count() == 1); } assert(shared.use_count() == 1); /* Read a weak_ptr */ std::weak_ptr<MyInt> weak; { boost::archive::binary_iarchive ar(data); ar >> weak; assert(shared.use_count() == 1); assert(weak.use_count() == 1); { std::shared_ptr<MyInt> loaded = weak.lock(); assert(loaded); /* It looks good */ assert(loaded->value == 12); /* Let's see... */ assert(loaded != shared); shared->value = 42; assert(loaded->value == 12); /* So another shared_ptr exists somewhere... */ assert(shared.use_count() == 1); assert(loaded.use_count() == 2); assert(weak.use_count() == 2); } /* But where ? */ assert(shared.use_count() == 1); assert(weak.use_count() == 1); } /* Let's see */ assert(shared.use_count() == 1); assert(weak.use_count() == 0 /* weak.expired() */); /* Hmm, was it inside the archive itself ? */ } /* Let's try again, we need the input stream to be back to the beginning */ data.seekg(0); std::weak_ptr<MyInt> weak; { boost::archive::binary_iarchive ar(data); assert(weak.use_count() == 0); ar >> weak; assert(weak.use_count() == 1); /* It looks fine */ } assert(weak.use_count() == 0); /* aw, on a second thought, it really doesn't */ return 0; } I am building and running this using g++-9.3.0 with boost-1.72.0-1. Clearly, the weak_ptr gets itself a shared_ptr "somewhere" as it can be used after being read. I was wondering "where it is stored and how to access it, if possible, without some `weak.lock()` call ?" It looks like the shared_ptr gets destroyed at the same time as the archive, so I guess it is cleaned up thanks to the delete_created_pointers method when the iarchive is destroyed : does delete_created_pointers actually contains also shared_ptr ? As far as I can see in the code serialization/weak_ptr.hpp, we actually save then load a std::shared_ptr, so I was expecting the weak_ptr to expire as soon as the call returns : since it is not, there *is* another reference. Now, I know I must save the shared_ptr in the archive myself to have my own reference and to get all weak_ptr running properly. But I came across this use case and I can't take it out of my mind, so if you have an explanation, I'm all ears. Thank you

1 0

Boost 1.73, beast and -fconcepts
by David P. Riedel 04 May '20

04 May '20

Hi It seems that beast still does not work with gcc 9.3 and -fconcepts. If I try to build the beast test programs using -fconcepts, I still get a lot of errors such as those below. I thought this was fixed in 1.73. ???? Thanks --------------------- In file included from ../../../boost/beast/core/basic_stream.hpp:1457, from ../../../boost/beast/core.hpp:16, from doc/core_snippets.cpp:19: ../../../boost/beast/core/impl/basic_stream.hpp:781:1: error: no declaration matches ‘auto boost::beast::basic_stream< <template-parameter-1-1>, <template-parameter-1-2>, <template-parameter-1-3> >::async_connect(const endpoint_type&, ConnectHandler&&)’ 781 | basic_stream<Protocol, Executor, RatePolicy>:: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ In file included from ../../../boost/beast/core.hpp:16, from doc/core_snippets.cpp:19: ../../../boost/beast/core/basic_stream.hpp:1202:5: note: candidates are: ‘template<class Protocol, class Executor, class RatePolicy> template<class Iterator, class ConnectCondition, class IteratorConnectHandler> requires completion_token_for<IteratorConnectHandler, void()> typename boost::asio::async_result<typename std::decay<RangeConnectHandler>::type, void(boost::system::error_code, Iterator)>::return_type boost::beast::basic_stream< <template-parameter-1-1>, <template-parameter-1-2>, <template-parameter-1-3> >::async_connect(Iterator, Iterator, ConnectCondition, IteratorConnectHandler&&)’ 1202 | async_connect( | ^~~~~~~~~~~~~ ../../../boost/beast/core/basic_stream.hpp:1141:5: note: ‘template<class Protocol, class Executor, class RatePolicy> template<class Iterator, class IteratorConnectHandler> requires completion_token_for<IteratorConnectHandler, void()> typename boost::asio::async_result<typename std::decay<RangeConnectHandler>::type, void(boost::system::error_code, Iterator)>::return_type boost::beast::basic_stream< <template-parameter-1-1>, <template-parameter-1-2>, <template-parameter-1-3> >::async_connect(Iterator, Iterator, IteratorCon

2 2

[review] [LEAF] LEAF Review Starts May 22
by Michael Caisse 01 May '20

01 May '20

The Boost formal review for LEAF (Lightweight Error Augmentation Framework) will take place May 22, 2020 through May 31st. LEAF is brought to us by Emil Dotchevski, the author of Boost.Exception. Similar to Boost.Exception, this library allows arbitrary error objects to be returned; however, unlike Boost.Exception it does not require dynamic memory. The library can be used with or without exception handling. Error handling is such an import part of every software system. LEAF provides a unique solution to the problem. The excellent documentation can be found here: <https://zajo.github.io/leaf/> and includes a tutorial, examples, reference, design rationale, and comparisons to other error handling approaches/libraries. The source is available at github. The review branch will be used during the review period. <https://github.com/zajo/leaf/tree/review> The purpose of this announcement is to give people some time to start looking at the documentation and source code for the library. Feel free to begin asking questions. Most of all, get involved! Everybody has some experience with handling errors. Tell the Boost community what you think of LEAF. Thank you for your participation! michael -- Michael Caisse Ciere Consulting ciere.com

1 0

Three-way comparison operator
by Leo Carreon 30 Apr '20

30 Apr '20

Hi Boosters, Does anyone know if Boost.Test has been updated to provide support for the three-way comparison operator? I’m asking because I’ve encountered a situation where my user-defined class provides the three-way comparison operator but when I perform comparisons using Boost.Test it complains that my class has no ‘operator==’ which is true because it provides ‘operator<=>’. Please note that I’m using Boost 1.73.0 and g++ 10.0.1. Is there any plan to add the three-way comparison operator support? Kind regards, Leo

4 4

Windows boost::filesystem canonical on NTFS network drives mounted in a directory
by gga 29 Apr '20

29 Apr '20

When on Windows, if you use boost::filesystem::canonical() to resolve a NTFS network drive mounted in a directory, the result is what I see as a useless result. For example: C:\ntfs_mount\other\dirs\file boost::filesystem returns: C:\?VolumeXXX-XXX-XXX-XXX\ other\dirs\file This volume notation is useless as it cannot be passed to any file function. Also, the path is divided by a space to other\dirs. It would be better if canonical does not mess with the network drive at all and returns the directory as is, like boost::filesystem::absolute() does. -- Gonzalo Garramuño

1 0

[release] Boost 1.73.0
by Marshall Clow 28 Apr '20

28 Apr '20

Release 1.73.0 of the Boost C++ Libraries is now available. These open-source libraries work well with the C++ Standard Library, and are usable across a broad spectrum of applications. The Boost license encourages both commercial and non-commercial use. This release contains 2 new libraries and numerous enhancements and bug fixes for existing libraries. For details, including download links, see < http://www.boost.org/users/news/version_1.73.0> You can also download directly from: < https://dl.bintray.com/boostorg/release/1.73.0/> To install this release on your system, see < http://www.boost.org/doc/libs/release/more/getting_started/index.html> Thanks to everyone who participated in this release. Thanks, -- The Boost Release Team Marshall Clow, Micheal Caisse and Glen Fernandes

1 0

[release] Boost 1.73.0 Release Candidate 1
by Marshall Clow 26 Apr '20

26 Apr '20

The first release candidates for the 1.73.0 release are now available at: <https://dl.bintray.com/boostorg/release/1.73.0/source/> The SHA256 checksums are as follows: d2e7501bb04fe7abc09aa93f013ae997604286a882da1bd36ddd96ea1163ea71 boost_1_73_0_rc1.7z 4eb3b8d442b426dc35346235c8733b5ae35ba431690e38c6a8263dce9fcbb402 boost_1_73_0_rc1.tar.bz2 9995e192e68528793755692917f9eb6422f3052a53c5e13ba278a228af6c7acf boost_1_73_0_rc1.tar.gz 0909a79524f857ef54570ceef8f397cc0629202532cc997785479c7c08bbc2a4 boost_1_73_0_rc1.zip As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. -- The Release managers

3 7

Using coroutines in ASIO
by Pedro Pinto 25 Apr '20

25 Apr '20

Hi there, I am trying to compile the asio co-routine example but as far as I can tell boost is looking for an <experimental/coroutine> header that does not exist on my system. I am using boost 1.71 and gcc on ubuntu 20.04. I tried both stock gcc (v9 on ubuntu) which does not have a coroutine header and gcc 10 which does have it but not under the experimental directory. Here is the output with gcc 10: /usr/lib/ccache/g++ -I../../src -g -Wall -pedantic -Wextra -Werror -DASIO_HAS_CO_AWAIT -DBOOST_ASIO_HAS_CO_AWAIT -fcoroutines --std=c++2a -fdiagnostics-color -MD -MT CMakeFiles/mcbridge.dir /src/main.cpp.o -MF CMakeFiles/mcbridge.dir/src/main.cpp.o.d -o CMakeFiles/mcbridge.dir/src/main.cpp.o -c ../../src/main.cpp In file included from /usr/include/boost/asio/co_spawn.hpp:22, from ../../src/main.cpp:4: /usr/include/boost/asio/awaitable.hpp:22:10: fatal error: experimental/coroutine: No such file or directory 22 | #include <experimental/coroutine> Any ideas on how to compile this example?

2 1

10 slots for the Google Summer of Code 2020
by David Bellot 22 Apr '20

22 Apr '20

Hi Boost community, this year Google allocated 10 slots for the Google Summer of Code. It's again a great achievement thanks to the hard work of our team of dedicated mentors. We are now selecting the best students among all the proposals we receive. We received 79 proposals this year. 10 of them will be selected. This year, a lot of proposals if not all are targeting a finalization of the work which has been done in the previous GSoCs and we should see a lot of code being pushed in production. It's a very exciting time for all of us. The mentors are now hard at work to choose the students. We'll keep you informed when it's official. Best regards, David

1 0

Re: [Boost-users] [random_device] App crash on boost::ramdom::random_device
by 박형근(HYUNGGUN PARK) 22 Apr '20

22 Apr '20

> On Fri, Apr 17, 2020 at 10:29 AM 박형근(HYUNGGUN PARK) via Boost-users > <boost-users(a)lists.boost.org> wrote: > > > > My app crash on customer's site. It crashes everytime the app is > starting. > > It happens on Chinese Windows7 (64bit) OS but it runs well on my PC > > with > Korean Windows7 OS and English Windows7 OS. > > My app is built on VisualStudio 2010 (x86 build, C++) using > > websocketpp > library and boost 1.55. > > The call-stack from the dump file shows my app(TMSClient100) called > websocketpp::endpoint, then it called boost::random::random_device > then ThrowException. Here's the call-stack info: > > > > 14 001cf6ac 7113872d e06d7363 00000001 00000003 > > KERNELBASE!RaiseException+0x58 > > 15 001cf6e4 713b7c5a 001cf704 713d9350 b621b77b > > msvcr100!_CxxThrowException+0x48 > > [f:\dd\vctools\crt_bld\self_x86\crt\prebuild\eh\throw.cpp @ 157] > > 16 001cf814 713b7da3 001cf840 b621b9a7 002a19bc > > TMSClient100!boost::random::random_device::impl::error+0x15a > > 17 001cf8b8 713b7f7b 001cf8dc b621b94b 00000000 > > TMSClient100!boost::random::random_device::impl::impl+0x143 > > Check the exception message. Is it failing because it: > 1. Could not find a provider name > 2. Could not acquire CSP context > > If #1, this might be solved by upgrading from Boost 1.55 to a newer > version. This fix in particular: > https://github.com/boostorg/random/commit/dc2fdeac443fb56c87a03a0887c9 > e7c8 > e1d3296d > went into Boost 1.56. > > Glen Glen, Thank you for your reply. Thanks to your suggestion, upgrade from Boost 1.55 to 1.56 solved this problem. Thank you very much. Harry

1 0