May 2015 - Boost - lists.preview.boost.org

What variants should be build on Windows by default?
by Olaf van der Spek 11 May '15

11 May '15

Hi, What library variants should be build on Windows by default? Currently it builds: --build-type=<type> Build the specified pre-defined set of variations of the libraries. Note, that which variants get built depends on what each library supports. -- minimal -- (default) Builds a minimal set of variants. On Windows, these are static multithreaded libraries in debug and release modes, using shared runtime. On Linux, these are static and shared multithreaded libraries in release mode. Personally I use a static-release runtime often (for simple deployments). What variants do others use and what should be build by default? Greetings, -- Olaf

4 3

[array] Support for multidimensional arrays?
by Peter Dimov 10 May '15

10 May '15

Currently, creating a multidimensional boost/std::array is a bit unwieldy: array<array<array<int, 5>, 4>, 3> a; // int a[3][4][5] a[2][3][4] = 5; As a first step, we should consider enabling the element type T of array<T, N> to be an array type. array<int[4][5], 3> a; // int a[3][4][5] a[2][3][4] = 5; This almost works already, except for the array interface functions specified and implemented in terms of, say, std::equal or std::fill. It is an open question whether these standard algorithms should not be fixed themselves to work on array types, as swap_ranges has been, but for now, we can implement the array functions in a way that works. Opinions on that part? If we agree that this is a good idea, the next logical step would be to eliminate the special handling of the first extent and simply allow array<int[3][4][5]> a; // int a[3][4][5] a[2][3][4] = 5;

4 4

[boost.org website] Incorrect links on Boost website
by Michael Levine 10 May '15

10 May '15

I was browsing through the Boost website this morning and the links on the right sidebar caught my attention. It appears that there is a duplicate entry for "coding guidelines for constant integral expressions", under Development / Requirements and Guidelines. The first of these links points to the correct page. The second links to a page regarding library metadata. I'm not sure who, exactly, to contact for this, so I figured this list would be the most appropriate means to have this addressed. Best regards, Shmuel

2 1

[algorithm] Removal of preference to std implementation
by Rob Conde 10 May '15

10 May '15

Marshall/Others-in-the-know, In boost algorithm change set 4dac507b773cf4a82727703ddffa2b36c42ab814, the code which uses std version of algorithms when available was removed. I'm curious about the motivation for this change. I can guess at a few potential reasons, but I figure you can just share the real reason. This change potentially explains a performance regression in my code so I'm just trying to understand the motivation before I make any code changes. Thanks! Rob Conde

1 0

[testing] Add a tester with hidden visibility
by Andrey Semashev 09 May '15

09 May '15

Hi, Is it possible to add a tester that enables hidden visibility (- fvisibility=hidden -fvisibility-inlines-hidden) when building and testing Boost? I believe, a single tester on Linux+GCC would be enough to discover most issues related to symbol visibility. The reason I'm asking is that I believe Boost should support building with hidden visibility by default. This mode allows more optimizations and does not export unintended symbols from libraries. I've set hidden visibility for Boost.Log (in develop) and, as I understand, Robert is doing a similar change in Boost.Serialization. However, when building 1.58 with hidden visibility for my project I found out that not all libraries work well with that setting. What do you think about it?

4 20

Re: [boost] Windows XP support survey
by Thomas Trummer 07 May '15

07 May '15

The sad reality in a lot of industries is that we still have to support Windows XP for the foreseeable future. Though, I understand the desire to drop support for it. Believe me, everyone would be more than happy to just require Windows 7 as default and be done with it. However, it is hard to convince costumers to make the switch when there is only risk and no tangible benefit attached to it. In addition to this we do have the problem that we control complex (and expensive) hardware that consists of components that don’t have drivers for Windows 7 because the manufacturer discontinued them. For these people the end of support from Microsoft isn’t really an argument since the computers are not connected to the Internet and continue to work fine, not to mention that you can still buy extended support from Microsoft. The huge outcry from the C++ community when Microsoft wanted to drop support for XP in Visual Studio should be a clear indicator that a lot of people livelihoods still depend on it. Unfortunately since then the install base hasn’t really shrunken too much. So even Visual C++ 2015 will continue to support it. So, as long as a Visual C++ version supported by Boost can build for Windows XP and there isn’t a significant cost attached to it, Boost should also continue to support Windows XP.

3 2

Re: [boost] Windows XP support survey
by Andrzej Horoszczak - Billon 07 May '15

07 May '15

1 0

Re: [boost] Interest in a container which can hold multiple data types?
by James Armstrong 07 May '15

07 May '15

Sorry about the README file, I pretty much just uploaded what I had to github so people can grab the source. Admittedly, things there are lacking. I'll work on filling out the readme with some instructions hopefully today. The main difference between this and a container of boost:any is that the container of boost::any still is a container of one data type, just being that one data type is a very flexible one. This container natively supports insertion of different data types. So, you could do, int my_int = 5; double my_double0 = 9342.132; double my_double1 = 987.654; std::string my_string("This is a string."); std::vector<double> my_vec; my_vector.push_back(123.456); omni my_container; my_container.push_back(my_int); my_container.push_back(my_double); my_container.push_back(my_string); my_container.push_back(my_vector); and the container happily stores the data. You can then access the data through iterators //access all doubles for (auto itr = my_container.begin<double>(); itr != my_containert.end<double>(); ++itr) { std::cout << *itr << std::endl; } or through specifying the data type and index //get the 0th element string std::cout << my_container.at<string>(0) << std::endl; -- James Armstrong

9 27

[lambda] Addressing pull requests
by Edward Diener 07 May '15

07 May '15

The maintainer for lambda in the libs/maintainer.txt list is Jaako Jarvi, who is the original author. But I see others who have made changes to the library. So I am wondering if anyone is still actively maintaining the library. There are a number of pull requests for lambda, related to updating the code for type_traits Version2, by John Maddock and myself. Is someone handling these ?

5 6

Arguing about names in Boost libraries
by Edward Diener 06 May '15

06 May '15

What is wrong with the people commenting on libraries for Boost who are so concerned about the names being used ? We are talking only about C++ identifiers and yet there are now endless discussions where people disagree on names or argue that some name is not what they want or that they like some other name, with a list of names that they find adequate. This is irrelevant. There is something mentally wrong about all those peoples spending endless amounts of time around this issue. I notice this does not happen nowadays for just one library but for ever single library, or proposed library, which is mentioned on this mailing list. Is this the way all of you people actually work in your daily jobs, or on your own projects, spending hours and days arguing with others about C++ identifiers for every little thing ? Are you not aware of the absurdity of your endless remarks ? Grow up ! Surely there are better things to do than carry on endlessly about some name, that is should be x or y or z or some other inane preferable list. It's time for somebody to say this, so I will do it. A C++ identifier name, as long as it is vaguely about what the functionality entails, is perfectly adequate. It does not have to satisfy everyone in the universe. Period !

9 10

C++11 compilers and fenv
by Damian Vicino 06 May '15

06 May '15

Hi, I’m implementing the safe_float library (on c++11) proposed for the GSOC . I was planning to access FENV for detecting unsafe operations on the native floating point types and include some custom pre-post conditions for testing other floating types safety conditions. Conditions come from policies passed as template parameter to give users the possibility of chose their own set of safety conditions to be enforced. An example of the conditions that can be asked is enforcing overflows to infinite does not happen. An overflow to infinite can be detected by the FE overflow flag or because the addition of two finites numbers result in an infinite one. I just found today that clang++ does not support the pragma to use FENV yet. warning: pragma STDC FENV_ACCESS ON is not supported Googling a little I found out that neither does gcc. How are c++11 boost libraries dealing with these compiler issues? I got some options in mind: — Implement everything using custom detections, this will work in every compiler, but it will run slower for everyone too. — Finding a compiler with the support, implementing the library using the FENV, and assuming compilers will support it when the library is finished and reviewed. — Detect somehow the FENV pragma was not processed and have 2 implementations. No idea how to detect something like this. Had someone dealt with similar problem already? Is there a common approach? Best regards, Damian Vicino

4 5

[ boost ] [ Trie ]
by Cosmin Boaca 06 May '15

06 May '15

Hello, I am interested in working at the Boost.Trie project which was started during GSoC 2013 and I am searching for someone who can give me some advice about what should I change / add to the current implementation, because I'm not too experienced in developing library code. However, I do not want do this during GSoC because I won't be available during the summer. I want to contribute during my spare time. If anyone thinks that can help me with this, please reply to this email. Thank you, Cosmin

8 90

BenPope x86_64 Ubuntu issue at link time
by Vicente J. Botet Escriba 06 May '15

06 May '15

Hi, I'm getting this link error /usr/bin/ld: error: /home/ben/development/boost/test/build/results/boost/bin.v2/libs/thread/test/thread__constr__FArgs_p_lib.test/clang-linux-3.7~asan~c14_libc++/debug/threading-multi/threads/thread/constr/FArgs_pass.o: multiple definition of 'operator delete(void*)' /usr/bin/ld: /home/ben/development/llvm/trunk/install/release/bin/../lib/clang/3.7.0/lib/linux/libclang_rt.asan_cxx-x86_64.a(asan_new_delete.cc.o): previous definition here at http://www.boost.org/development/tests/develop/developer/output/BenPope%20x… I define it as follows #if defined _GLIBCXX_THROW void* operator new(std::size_t s) _GLIBCXX_THROW (std::bad_alloc) #elif defined BOOST_MSVC void* operator new(std::size_t s) #else void* operator new(std::size_t s) throw (std::bad_alloc) #endif { std::cout << __FILE__ << ":" << __LINE__ << std::endl; if (throw_one == 0) throw std::bad_alloc(); --throw_one; return std::malloc(s); } Is this a compiler issue, or is my definition of the operator new incorrect? Vicente

2 2

[move] unique_ptr converting constructor
by Peter Dimov 06 May '15

06 May '15

While looking at https://svn.boost.org/trac/boost/ticket/11259 I found that after implementing a converting shared_ptr constructor that takes a boost::movelib::unique_ptr, the straightforward code in the ticket still fails (with g++ in C++03 mode): namespace mystd { using boost::shared_ptr; using boost::make_shared; using boost::movelib::unique_ptr; using boost::movelib::make_unique; } mystd::unique_ptr<int> load_thing() { return mystd::make_unique<int>(1); } mystd::shared_ptr<int> get_thing() { return load_thing(); } So, out of curiosity, I looked at how unique_ptr itself implements its converting constructor. It turns out that it, too, doesn't work. mystd::unique_ptr<int const> get_thing_2() { return load_thing(); } Furthermore, the two possible ways to make the shared_ptr code work: mystd::shared_ptr<int> get_thing() { mystd::shared_ptr<int> r( load_thing() ); return r; } mystd::shared_ptr<int> get_thing() { mystd::unique_ptr<int> r( load_thing() ); return boost::move( r ); } fail for the unique_ptr<int const> case.

4 8

Re: [boost] Interest in a container which can hold multiple data types?
by Brian Wood 06 May '15

06 May '15

Joaquin M Lopez Munoz writes: > I wrote something on this some time ago, maybe worth having a look at: > > http://bannalia.blogspot.com/2014/05/fast-polymorphic-collections.html <http://bannalia.blogspot.com/2014/05/fast-polymorphic-collections.html> > http://bannalia.blogspot.com/2014/05/fast-polymorphic-collections-with.html <http://bannalia.blogspot.com/2014/05/fast-polymorphic-collections-with.html> I look forward to someone implementing these things. Any progress? -- Brian Ebenezer Enterprises - In G-d we trust. http://webEbenezer.net

2 1

Fwd: Boost.Python build error in 1.58 RC1, 2 and 3 (OK in Beta 1)
by Marshall Clow 06 May '15

06 May '15

---------- Forwarded message ---------- From: <mark.incley(a)synx.com> Date: Wed, Apr 8, 2015 at 1:29 AM Subject: Boost.Python build error in 1.58 RC1, 2 and 3 (OK in Beta 1) To: mclow.lists(a)gmail.com Hi Marshall, Please excuse the direct mail, but my mailing list posting wasn't picked up on by anyone. I believe I've found a problem in Boost.Python that causes build errors if the user has Python 3.0 or later on their system. This change appears to have been introduced after 1.58 Beta 1. Using: bjam address-model=32 runtime-link=shared toolset=msvc-12.0 --stagedir=./stage/x86 C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\INCLUDE\xtgmath.h(206) : warning C4273: 'round' : inconsistent dll linkage C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\INCLUDE\math.h(516) : see previous definition of 'round' libs\python\src\exec.cpp(91) : error C2664: 'FILE *_Py_fopen(PyObject *,const char *)' : cannot convert argument 1 from 'char *' to 'PyObject *' Types pointed to are unrelated; conversion requires reinterpret_cast, C-style cast or function-style cast call "C:\Users\MARK~1.INC\AppData\Local\Temp\b2_msvc_12.0_vcvarsall_x86.cmd" >nul cl /Zm800 -nologo @"bin.v2\libs\python\build\msvc-12.0\debug\link-static\threading-multi\exec.obj.rsp" ...failed compile-c-c++ bin.v2\libs\python\build\msvc-12.0\debug\link-static\threading-multi\exec.obj… The problem appears to be this Boost.Python code that was changed some time after 1.58 Beta 1 and appears in 1.58 RC1 within boost\libs\python\src\exec.cpp at around line 91: 1.58 beta 1: FILE *fs = fopen(f, "r"); 1.58 RCs: FILE *fs = _Py_fopen(f, "r"); ← f is a char *, hence the error shown above. It expects a PyObject * as the first parameter. This code occurs within a test for #if PY_VERSION_HEX >= 0x03000000 which may explain why no-one else has reported it – I have Python 3.3.0 on my HDD. Cheers, Mark. _____________________________________ Sent from http://boost.2283326.n4.nabble.com

13 31

Interest in a container which can hold multiple data types?
by James Armstrong 06 May '15

06 May '15

Hello forum, I was wanting to gauge interest in a container I developed which can natively handle multiple data types (as opposed to using a container of boost::any). I have the templated source hosted on my github account at https://github.com/armstrhu/omni There is some documentation on the current functions under omni/doc/html/index.html. main.cpp has some example codes on inserting, retrieving, and manipulating the data within the container. The one caveat to using the container is that the function destroy<T>() must be called for all data types inserted into the container, I currently have an assert in the destructor which will hit if destroy<T>() is not called for a data type. Also, this requires using a modified version of boost/core/typeinfo.hpp, so you will need to replace the original with the one I provide. The changes simply keep the boost functionality even when C++ typeinfo is available. I am trying to guage interest in this, and if there is interest, get some feedback on additional functionality to include. The iterators I have now work, but they are inefficient and that will be fixed eventually. Currently, I'm calling the container an omni, but don't really like that name and am looking for suggestions for something better :) Thanks -- James Armstrong

5 5

[Voronoi] Weighted Voronoi diagram
by Sybren A. Stüvel 06 May '15

06 May '15

Dear Boost mailing list, Currently I'm working on a Voronoi-based crowd simulation system. We model crowd agents as line segments, and feed them to the Boost Voronoi library. It would be very nice if we could give different line segments different weights. Does the Boost Voronoi library support different weights for its inputs? If not (which I'm afraid is the case, as it isn't mentioned in the documentation), how difficult would it be to add support for this? Kind regards, -- Sybren A. Stüvel http://stuvelfoto.nl/ http://stuvel.eu/

1 0

[threads] Unexpected behaviour of Thread loal storage
by Fu ji 06 May '15

06 May '15

Hello, I would like to ask if there is everything ok with my way of thinking. I have 3 solutions (projects) in MS Visual Studio. Let say: SolutionA (static lib) SolutionB (linked with Solution A) SolutionC (linked with Solution A) In Header file of solutionA I have delaration of TLS from boost static boost::thread_specific_ptr<int> TLS; And now in SolutionB I set tls (TLS.reset(uVariable)) - uVariable is integer 2 Call function from SolutionC and call TLS.get() but there is different value than 2. But when i return from function to SolutionB and call TLS.get there is everything ok (return 2). It's ok ? Why in solutionC there is no visibilty of value set in solutionB ? It's the same thread (with the same threadId). Best regards

2 18

[tuple] Changes living in develop
by Marcel Raad 06 May '15

06 May '15

Hi, there are three commits from last year in develop that haven't been merged to master yet. Could someone merge them please? 8bdf2a9 Make tuple constructor explicit 4fb2618 Add metadata file. b4f0590 Ported documentation to boostbook I always have to apply 8bdf2a9 manually to new Boost releases.

1 0

Re: [boost] [Boost-testing] Attempt to redeclare already registered project id '/check/predef'
by Tom Kent 06 May '15

06 May '15

On Tue, May 5, 2015 at 2:43 AM, Dmitry Moskalchuk <dm(a)crystax.net> wrote: > > Hello, > > Running tests of 'develop' branch, I've bumped with error today: > > error: at /home/cislave/.crystax/ci-boost/regression/android-19-x86-develop/test/develop/regression/x86/test/boost_bb/src/build/project.jam:1069 > error: Attempt to redeclare already registered project id '/check/predef'. > error: Original project: > error: Name: ../check/predef > error: Module: ../check/predef > error: Main id: /check/predef > error: File: ../check/predef.jam > error: Location: ../check/predef > error: New project: > error: Module: ../../predef/check/predef > error: File: ../../predef/check/predef.jam > error: Location: ../check/predef > > > It happens only when run regression tests for 'develop' branch ('master' > is OK). Is this known problem? > I'm getting the same error (which is stopping the regression run) on all my develop testers (windows and linux). Tom

2 1

[Master project] GitHub issues disabled
by Louis Dionne 06 May '15

06 May '15

Hi, The title pretty much says it all. Is there a reason for disabling GitHub issues on the master project? Also, I'd like to know when will the master project be tagged with the 1.58.0 release. Thanks, Louis

18 35

Re: [boost] Interest in a container which can hold multiple data types?
by James Armstrong 05 May '15

05 May '15

On 5/5/2015 7:49 PM, Joaquin M Lopez Munoz wrote: > TONGARI J <tongari95 <at> gmail.com> writes: > >> If the data sequence shows some affinity, I guess an ordered >> tuple_vector-based sequence can provide some performance benefit in >> traversal (with some crafted for_each method). >> >> That is, a sequence of [AAABBB] may be traversed faster than [ABABAB] for >> such a container, and I believe in the later case such that the types are >> uniformly distributed, vector<variant> will perform better. > I wrote something on this some time ago, maybe worth having a look at: > > http://bannalia.blogspot.com/2014/05/fast-polymorphic-collections.html > http://bannalia.blogspot.com/2014/05/fast-polymorphic-collections-with.html > Nice analysis. If we can identify different use cases, we may be able to come up with several implementations of heterogeneous containers which balance efficiency and flexibility. I think it would be an interesting project to work on and surely useful if the containers all the a similar 'feel' -- James Armstrong

1 0

[filesystem] copy_file compilation error
by Rob Conde 05 May '15

05 May '15

In boost 1.55 the following code compiled file on visual studio 2013 and g++ 4.4.7: #include "boost/filesystem/operations.hpp" void myFunc() { using namespace boost::filesystem; copy_option opt(copy_option::overwrite_if_exists); copy_file(path("p1"),path("p2"),opt); } in boost 1.58 I get the following error on g++ 4.4.7 (Visual Studio 2013 continues to work): test.cpp: In function 'void myFunc()': test.cpp:9: error: no matching function for call to 'copy_file(boost::filesystem::path, boost::filesystem::path, boost::filesystem::copy_option&)' boost/filesystem/operations.hpp:494: note: candidates are: void boost::filesystem::copy_file(const boost::filesystem::path&, const boost::filesystem::path&, boost::filesystem::copy_option::enum_type) boost/filesystem/operations.hpp:500: note: void boost::filesystem::copy_file(const boost::filesystem::path&, const boost::filesystem::path&) boost/filesystem/operations.hpp:505: note: void boost::filesystem::copy_file(const boost::filesystem::path&, const boost::filesystem::path&, boost::filesystem::copy_option::enum_type, boost::system::error_code&) boost/filesystem/operations.hpp:511: note: void boost::filesystem::copy_file(const boost::filesystem::path&, const boost::filesystem::path&, boost::system::error_code&) is this a bug in filesystem on gcc, a bug in gcc, a non-standard behavior in visual studio 2013, or an intended change? Rob Conde

2 2

Re: [boost] Description of pre-requisites for building Boost 1.58.0?
by Marko Asplund 05 May '15

05 May '15

On 2015-04-24 22:17:55 Ben Pope wrote: > This doesn't really answer the question, but as a practical note, you > can do: > sudo apt-get build-dep libboost* > Which will install the build dependencies for the version of Boost in > your repository, rather than the version you're trying to build, but it > should get you where you want to be most of the time. FYI, I submitted a documentation pull request for adding a short section about pre-requisite software installation: https://github.com/boostorg/boost/pull/78 marko

1 0