March 2020 - Boost-users - lists.boost.org

Boost.Test failing to compare std::basic_string<char8_t> strings
by Leo Carreon 09 Mar '20

09 Mar '20

Hi Boost.Users, I’m using Boost.Test from Boost 1.72.0 and I’m getting a failure when I perform the following test: auto string_1 = u8”value”s; auto string_2 = u8”value”s; BOOST_TEST(string_1 == string_2); However, the problem goes away when I change the test line to: BOOST_TEST((string_1 == string_2)); What is the reason why Boost.Test is failing without the double parentheses? Kind regards, Leo

2 3

[date-time] USE_DATE_TIME_PRE_1_33_FACET_IO to be removed in next release
by Jeff Garland 06 Mar '20

06 Mar '20

In case you've not checked the calendar (see below) it's 2020 and a the c++ world has moved on from the issue this macro solved. boost date-time also needs to move forward into the decade with a modernized all header code base . If you're dependent on this macro for any reason I'd like to hear about it. Thanks, Jeff using namespace boost::gregorian; date today = day_clock::local_day(); cout << today.year() << endl; //2020 ;)

1 0

uBlas and Google Summer of Code 2020
by David Bellot 06 Mar '20

06 Mar '20

Hi Boost community, I've added new projects for the Google Summer of Code 2020 for the Boost uBlas library (matrix computations, tensors and linear algebra). If you're a student and wish to candidate, please contact me as soon as possible. The list of projects is here: https://github.com/boostorg/wiki/wiki/Google-Summer-of-Code%3A-2020 Cheers, David

2 2

[mp11][mpl] Recurse over type list using mp11's mp_rest
by deb 06 Mar '20

06 Mar '20

Hello, I am currently learning about mp11 from the docs . While mpl had an iterator functionality to get an alias over a type in mp_list it seemed mp11's mp_rest could be used for a similar task. So I coded up a simple example of recursing over a type list based on some type check , but am quite lost trying to make it work . So a possible pseudo code for the above might be template<typename list> struct recurse{ using rest_of_list=mp_rest<list>; using type = typename std::conditional< mp_empty<rest_of_list>::value, base type here on empty list, typename std::conditional< some "loop" break condition, type on true condition, typename recurse< rest_of_list >::type >::type >::type; } Is this even possible or there exists some other method to do the same in mp11. Thanks. -- Sent from: http://boost.2283326.n4.nabble.com/Boost-Users-f2553780.html

3 5

yabs: Yet Another Build System, seeking input/micro-team member
by degski 03 Mar '20

03 Mar '20

Hi All, Recently, I have started to get more and more dis-enchanted with Windows. The OS is being dumbed down every step of the way, while the internal complexity keeps going up. The latest and the most fun break-down to date of Windows was last week, the story is long, but in the end in came down to not being able to open the system-settings in the W10-style settings. The W7-style settings are also available, but the API's are overlapping but neither is a super-set of the other, so no way to fix the settings, because the settings themselves are broken (had something to do with display-drivers), way to go MS. In the meanwhile W10 is sorted again, but the whole lot is just getting increasingly fragile, while Fedora is getting increasingly solid (no bt-headphone stuttering, while connected to wifi and look at that brightness/crispness of that Full HD screen I have never seen before under W10!), the latter not upping complexity (2'001 packages to download and install to get from F30 to F31, but that's not complexity). I want to move to Fedora, and for now I'm dual-booting. This involves dumping VS in favor of VSCode (which is flat-packed for Fedora and compadres and is also available on OSX, so there is potentially a large audience), no acceptable other options (other than the one(s) VSCode is based on) are available AFAIKS, gnome-builder seems to have different goals from VS and it appears to be early days. VSCode seems (to me) to be like the best thing since sliced-bread, iff you are writing Python or Rust (Prolog works nicely as well), but C++ is only partially great, the language server works very nicely, code completion, good, but the build-system, proposed, is cumbersome. I would like VSCode to work more like VS, but without the complexity, the endless clicking, drop-down boxes etc etc. It comes down to writing json-files, which is good, but on a per project basis. Having to do this on a per project basis is a problem (that's going backwards from using project-templates in VS, I now in VS can create a new project in 3 seconds flat, and every detailed setting (clang and vc) is set correctly and how **I** want it, and I am running "Hello World!" (the default main in the template) in total 10 secs, this very handy to try things out). I don't want to go backwards. Therefore, I would like to do this on a repo-basis (the only one I have), with, iff required, another json in the project-folder (in the parent repo-folder), i.e.: /repos/ /repos/repo.json /repos/project-one/ /repos/project-one/project.json /repos/project-two/ /repos/project-two/project.json /repos/ ... I (want to) use vcpkg. I structure my projects always the same, all build systems (to my purpose) are basically an over-shoot on the one hand and overly fiddly on the other hand. The task at hand, build the .cpp files in the root of the project folder into one binary. The include files are either in the vcpkg/installed/.../include folder or in a sub-include-folder in the project. Libraries to link are either in vcpkg/installed/.../lib, or vcpkg/installed/.../debug/lib. We are not talking rocket-science here, I presume that many people have simple needs like this/me, or that if building (adopting this structure) becomes easy enough people would adopt such a structure. I would like to have one json to rule them all, and one json to either over-ride the defaults, or to add to the defaults, depending on whether the setting is defined in the repo.json. The repo.json might well be layered itself, to accommodate the various compilers, and/or configuration. For the moment nlohman's json seems perfectly adequate for the job, and it's certainly easy to use (and seems to just use std::maps in a clever way, so it's easy to follow what's going on). It's all about merging, fusing, overlaying those json's as needed (gonna be entertaining, making it smart), and pass that on to some generic API for building. So a new compiler would mean to just add a json-section, or a json-file, fill that with appropriate data (possibly nothing, iff the default-defaults are good), and it should work. The current way VSCode is setup, means that this imaginary build-system can be easily and readily included/used in VSCode (with more json). Yes, just like the working name indicates YABS, it must be as simple as possible, but no simpler. Any input is appreciated, but please refrain from arguing that BS-X does this better (and exists) or BS-Y has a large crowd of followers (and is used for a long period of time), that's water under the bridge, I will pursue this, just to see how far one can get pushing these initial ideas further. As usual my write-ups are a bit messy, I am learning I'm slightly dyslexic, sorry for the endless comma's. Thanks in advance for any ideas submitted. I would not mind partnering with somebody else, ya'll have a nice day, degski -- @systemdeg "We value your privacy, click here!" Sod off! - degski "Anyone who believes that exponential growth can go on forever in a finite world is either a madman or an economist" - Kenneth E. Boulding "Growth for the sake of growth is the ideology of the cancer cell" - Edward P. Abbey

2 1

How to trim headers into a strict header-only subset
by state_of_the_art 03 Mar '20

03 Mar '20

Hello, I have used Boost Library a few years but recently I hava a new need. I know that some modules need to be built separately before use.But for my personal need,I need a strict header-only subset so that I can place the trimed boost headers folder in any project and simply build that project. Therefore I tried to use bcp,but the bcp copies some modules which need separately build. For example,run "./bcp --report coroutine2 fiber dependencies.html" and I see this html,there are some lines: …………………… boost/context/continuation_ucontext.hpp -> libs/context/src/continuation.cpp -> libs/context/src -> boost/context/segmented_stack.hpp -> boost/coroutine2/segmented_stack.hpp -> boost/coroutine2boost/context/continuation_winfib.hpp -> libs/context/src/continuation.cpp -> libs/context/src -> boost/context/segmented_stack.hpp -> boost/coroutine2/segmented_stack.hpp -> boost/coroutine2boost/context/detail/apply.hpp -> boost/fiber/type.hpp -> boost/fiberboost/context/detail/config.hpp -> boost/fiber/condition_variable.hpp -> boost/fiberboost/context/detail/disable_overload.hpp -> boost/coroutine2/detail/disable_overload.hpp -> boost/coroutine2/detail -> boost/coroutine2 …………………… Coroutine2 can be header-only used.But I know that Boost.Context need to be built separately via webpage getting_started/unix-variants …… I don't mind the size after triming but I want to make sure that trimed subset of modules can be used strictly header-only.Is there a better way or if my need can hardly be met? Thank you for your kindly help! Logan Smith 2020-03-03

1 0

Shared memory between Visual Studio build and MinGW build
by Da Xu 03 Mar '20

03 Mar '20

Hi folks, got a strange one here. I am trying to get two programs to communicate with each other on Windows 10 via shared memory. Both of them are using Boost 1.72, compiled from the same set of headers. I am only using the shared memory portion of boost. However, one is compiled with Visual studio 2019 community, and the other one is compiled with MinGW 64, g++ version 9.2.0. I have to use MinGW, because I am writing a plugin for another program, which only takes dlls compiled with MinGW. The program compiled with MinGW creates the shared memory and allocates several objects. Then I started the second program that was compiled with Visual Studio 2019. It was able to find and connect to the managed segment. When I am viewing the segment info from Visual Studio's debugger, everything looks ok. However, when I call get_free_memory(). It spits back 18446744073709551104, which is just bizarre. Then, when I tried to find_or_construct on the managed segment. It just hangs, apparently stuck in winapi::sleep_tick() with the following tracedll!boost::interprocess::winapi::sleep_tick() Line 1288 C++dll!boost::interprocess::ipcdetail::thread_sleep_tick() Line 187 C++dll!boost::interprocess::spin_wait::yield() Line 128 C++dll!boost::interprocess::ipcdetail::try_based_lock<boost::interprocess::ipcdetail::spin_mutex>(boost::interprocess::ipcdetail::spin_mutex & m) Line 71 C++dll!boost::interprocess::ipcdetail::spin_mutex::lock() Line 81 C++dll!boost::interprocess::ipcdetail::spin_recursive_mutex::lock() Line 95 C++dll!boost::interprocess::interprocess_recursive_mutex::lock() Line 165 C++dll!boost::interprocess::scoped_lock<boost::interprocess::interprocess_recursive_mutex>::scoped_lock<boost::interprocess::interprocess_recursive_mutex>(boost::interprocess::interprocess_recursive_mutex & m) Line 81 C++dll!boost::interprocess::segment_manager<char,boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family,boost::interprocess::offset_ptr<void,__int64,unsigned __int64,0>,0>,boost::interprocess::iset_index>::priv_generic_named_construct<char>(unsigned char type, const char * name, unsigned __int64 num, bool try2find, bool dothrow, boost::interprocess::ipcdetail::in_place_interface & table, boost::interprocess::iset_index<boost::interprocess::ipcdetail::index_config<char,boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family,boost::interprocess::offset_ptr<void,__int64,unsigned __int64,0>,0>>> & index, boost::interprocess::ipcdetail::bool_<1> is_intrusive) Line 1076 C++dll!boost::interprocess::segment_manager<char,boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family,boost::interprocess::offset_ptr<void,__int64,unsigned __int64,0>,0>,boost::interprocess::iset_index>::priv_generic_construct(const char * name, unsigned __int64 num, bool try2find, bool dothrow, boost::interprocess::ipcdetail::in_place_interface & table) Line 760 C++dll!boost::interprocess::segment_manager<char,boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family,boost::interprocess::offset_ptr<void,__int64,unsigned __int64,0>,0>,boost::interprocess::iset_index>::generic_construct<SharedMem_Info>(const char * name, unsigned __int64 num, bool try2find, bool dothrow, boost::interprocess::ipcdetail::in_place_interface & table) Line 705 C++dll!boost::interprocess::ipcdetail::named_proxy<boost::interprocess::segment_manager<char,boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family,boost::interprocess::offset_ptr<void,__int64,unsigned __int64,0>,0>,boost::interprocess::iset_index>,SharedMem_Info,0>::operator()<char const (&)[17],int,int,int,int,bool,boost::interprocess::allocator<void,boost::interprocess::segment_manager<char,boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family,boost::interprocess::offset_ptr<void,__int64,unsigned __int64,0>,0>,boost::interprocess::iset_index>> &>(const char[17] & <args_0>, int && <args_1>, int && <args_2>, int && <args_3>, int && <args_4>, bool && <args_5>, boost::interprocess::allocator<void,boost::interprocess::segment_manager<char,boost::interprocess::rbtree_best_fit<boost::interprocess::mutex_family,boost::interprocess::offset_ptr<void,__int64,unsigned __int64,0>,0>,boost::interprocess::iset_index>> & <args_6>) Line 131 C++ It looks like it the segment manager was unable to get a mutex lock on the memory? But I don't believe I put a mutex lock on the segment manager from the other end. Everything works just fine if I am doing shared memory between two programs compiled with visual studio. So, it feels like somewhere some how memory alignment is off? Anybody ran into a similar issue before? Any help would be greatly appreciated! Thank you!!

4 5