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

Issue with compilation using XCode 4.6/GCC
by Tim Burgess 18 Nov '13

18 Nov '13

Hi, I have a project for Windows/Mac that builds fine using VS2010, but generates errors when I build using Xcode 4.6 with the GCC compiler (required, as this project also uses wxWidgets). All of my modules compile cleanly, but I get the following "orphaned" error output that doesn't seem to belong to any module: SurfaceReader Group /users/tim/developer/libraries/boost_1_54_0/boost/type_traits/decay.hpp /users/tim/developer/libraries/boost_1_54_0/boost/type_traits/decay.hpp:28: '<anonymous enum>' is/uses anonymous type /users/tim/developer/libraries/boost_1_54_0/boost/type_traits/decay.hpp:28: Trying to instantiate 'template<class T> struct boost::remove_reference' /users/tim/developer/libraries/boost_1_54_0/boost/type_traits/decay.hpp:38: '<anonymous enum>' is/uses anonymous type /users/tim/developer/libraries/boost_1_54_0/boost/type_traits/decay.hpp:38: Trying to instantiate 'template<class T> struct boost::remove_reference' I realise that I haven't provided the command that Xcode is using to build the project - this isn't deliberate, but I can't seem to locate where Xcode provides this on-screen using the VoiceOver accessibility tool (I'm a totally blind developer). Can anybody offer any advice regarding decay.hpp errors (I'm not including it directly in any of my modules), or can direct to me to where I might find a log that would provide more detail as to the compilation options that Xcode is using (i.e. its GCC-equivalent command line)? Many thanks in advance. Best wishes. Tim Burgess

1 1

Can gil library in boost 1.49 load png palette images?
by Guangchao Wang 18 Nov '13

18 Nov '13

I'm tried to load a png which can be found in the attachment, I find the color of this png is PNG_COLOR_TYPE_PALETTE and its depth is 8, I reviewed the code of png_io_private.hpp and found that no such type of PNG is supported. Can you give me some help about how to load such PNG with boost.gil library? Thank you very much. Best, Guangchao

4 6

Did build of Boost succeed - lib vs. stage/lib?
by Lutz Gehlen 17 Nov '13

17 Nov '13

Hello all, I am trying to build the Boost libraries version 1.55.0 with Visual Studio 2010 Express under Windows XP. Following the instructions in the documentation, I have unpacked the source code distribution, opened a Visual Studio command line window, and executed bootstrap .\b2 I haven't noticed any errors, and I was told in the end that I had successfully built the Boost libraries. So far, so good. I then tried to build the example program in section 6 of the Getting started guide. Section 6.1 instructs to add "C:\Program Files\boost\boost_1_55_0\lib" (in my case E:\boost_1_55_0\lib) to the list of Additional Library Directories. However, I do not have a lib directory in E:\boost_1_55_0?! I tried to build the example program and then searched for the library file libboost_regex-vc100-mt-gd-1_55.lib that Visual Studio was failing to find. I found it in E:\boost_1_55_0\stage\lib instead of E:\boost_1_55_0\lib. Adding this directory to the list of Additional Library Directories allowed me to build and execute the example program successfully. However, I am wondering now whether my build worked correctly or not. I am trying to avoid running into subtle and hard to debug problems later. Does anyone know if or what went wrong with my build? What is the stage directory and is it safe to use it? Thank you very much for your help and best wishes Lutz

2 2

the link to the documentation for "explicit bool operator" is broken
by Adam Romanek 17 Nov '13

17 Nov '13

Dear Boosters, I've just discovered that there's a broken link to the documentation of "explicit bool operator" macro on http://www.boost.org/doc/libs/1_55_0/libs/utility/index.html WBR, Adam Romanek

2 1

Can I return std::unique_ptr in a boost::future?
by Aaron_Wright＠selinc.com 15 Nov '13

15 Nov '13

I'm trying to return a std::unique_ptr in a boost::future. I can't get it to compile. Is it possible? The code wraps a packaged_task in a functor to make it copyable enough to pass to a boost::asio::io_service, where the work will be done in another thread. The code works for other return types that are copyable. This is what I have for a small test case: std::unique_ptr< int > get_int_impl(int i) { return std::unique_ptr< int >(new int(i)); } std::unique_ptr< int > get_int(int i) { typedef boost::packaged_task< std::unique_ptr< int > > Task; boost::shared_ptr< Task > task = boost::make_shared< Task >(boost::bind(&get_int_impl, 42)); boost::unique_future< std::unique_ptr< int > > future = task->get_future(); // <--- Error from this line. m_service.post(boost::bind(&Task::operator(), task)); return future.get(); } BOOST_AUTO_TEST_CASE(test_get_int) { BOOST_CHECK_EQUAL(*get_int(42), 42); } The error: boost/thread/future.hpp(239): error C2248: 'std::unique_ptr<_Ty>::unique_ptr' : cannot access private member declared in class 'std::unique_ptr<_Ty>' --- Aaron Wright

4 6

Re: [Boost-users] [Math] BOOST_FPU_EXCEPTION_GUARD on Linux
by David Roberts 15 Nov '13

15 Nov '13

> As I recall this was a hardware issue on early AMD 64-bit CPU's, possibly a > GLIBC issue too. In fact if I recall it was this issue: > http://sourceware.org/bugzilla/show_bug.cgi?id=2445 which didn't get much > help from the glibc guys > > But basically if any of the FPU exception flags are set prior to a call to > pow or exp, you risk getting a garbage answer back. I'll add a link to the > above in the code. > > HTH, John. That Bugzilla link was extremely helpful - thanks! I've done a bit more research and I'm pretty sure this WAS a bug in glibc. It's fixed on recent versions of Linux though. Taking your test program from http://sourceware.org/bugzilla/show_bug.cgi?id=2445 and trying it on some different Linux versions and CPUs resulted in this: SLES 10: kernel 2.6.16.21-0.8-smp glibc 2.4 g++ 4.1.0 CPU Xeon E5-2630 FAILS CentOS 6.2: kernel 2.6.32-220.el6.x86_64 glibc 2.12 g++ 4.4.6 CPU Xeon X3450 FAILS Amazon Linux 2013.03: kernel 3.4.57-48.42.amzn1.x86_64 glibc 2.12 g++ 4.6.3 CPU Xeon E5507 FAILS OpenSuSE 12.1: kernel 3.1.10-1.16-ec2 glibc 2.14.1 g++ 4.6.2 CPU Xeon E5-2650 WORKS OpenSuSE 12.2: kernel 3.4.6-2.10-ec2 glibc 2.15 g++ 4.7.1 CPU Xeon E5-2650 WORKS Fedora 20 beta: kernel 3.11.6-300.fc20.x86_64 glibc 2.18 g++ 4.8.2 CPU Xeon E5645 WORKS This shows that the problem definitely occurs on Intel CPUs as well as AMD, and is much more likely to be caused by glibc than the compiler or the kernel. The most likely change to the glibc code that fixed the problem is this one from May 2011: http://repo.or.cz/w/glibc.git/commitdiff/8db736347c7aca3201f61e3f05b5f672bc… There was a place in powl() where 4500 decimal was used instead of 0x4500 hex. 4500 decimal is 0x1194 hex. This number was then tested against the floating point status word. The exception flags are in the low byte of this word (see http://www.nacad.ufrj.br/online/intel/Documentation/en_US/compiler_f/main_f… for the values) and as a result of the typo the presence of either FE_UNDERFLOW = 0x10 or FE_DIVBYZERO = 0x4 will cause the test to pass unintentionally. More experiments with your test program on CentOS 6.2 (glibc 2.12) seem to confirm this - if only the flags FE_INEXACT, FE_OVERFLOW and/or FE_INVALID are set they DON'T cause the weird result to occur. Given the bug was in the x86_64-specific powl() implementation and it seems to be fixed in glibc version 2.14, I think you could get away with tightening up the criteria on your guard from: #if ((defined(__linux__) && !defined(__UCLIBC__)) || defined(__QNX__) || defined(__IBMCPP__)) && !defined(BOOST_NO_FENV_H) to: #if ((defined(__linux__) && defined(__x86_64__) && !defined(__UCLIBC__) && (!defined(__GLIBC_PREREQ) || !__GLIBC_PREREQ(2,14))) || defined(__QNX__) || defined(__IBMCPP__)) && !defined(BOOST_NO_FENV_H) That still doesn't help people like me who have to work on current 64 bit enterprise Linux releases, but for the benefit of anyone else who stumbles across this having suffered the same performance problem, here's a hack that works OK if neither your code nor any other 3rd party library you're using cares about the floating point flags. In boost/math/tools/config.hpp change: struct fpu_guard { fpu_guard() { fegetexceptflag(&m_flags, FE_ALL_EXCEPT); feclearexcept(FE_ALL_EXCEPT); } ~fpu_guard() { fesetexceptflag(&m_flags, FE_ALL_EXCEPT); } private: fexcept_t m_flags; }; to: struct fpu_guard { fpu_guard() { if (fetestexcept(FE_UNDERFLOW | FE_DIVBYZERO) != 0) { feclearexcept(FE_ALL_EXCEPT); } } }; This is beneficial for performance because the expensive calls are feclearexcept() and fesetexceptflag(). The (more-or-less identical on Linux) fegetexceptflag() and fetestexcept() functions are much quicker. Not bothering to reset the cleared flags saves one expensive call - obviously this relies on nothing downstream caring about them. Only clearing the flags if one of the problematic ones is set often avoids the other expensive call. I've noticed that the FE_INEXACT flag is almost always set if you do any serious floating point calculation, but the other 4 flags are set much less frequently. I'm not suggesting that you put this hack into the official Boost codeline, but just thought it might be useful to anyone in my situation who finds this post via a search engine.

2 1

concept_check() fails for types from unnamed namespace on gcc 4.2.3
by Adam Romanek 15 Nov '13

15 Nov '13

Dear Boosters! I have a problem compiling some source code that uses Boost 1.54 and gcc 4.2.3 (it's a toolchain for a MIPS platform but that's irrelevant). Basically there is a problem with checking concepts on types that come from unnamed namespaces. See the attached code example and compiler error message. Note that the same code compiles without any problems on gcc 4.6.3. I've found a question on boost-dev list [1] and a ticket [2] related to this particular issue. Note that the ticket was closed with "worksforme" resolution, which I don't understand - the problem is still out there. Now since Boost's minimal requirements are gcc 3.3 [3] then should I consider it a bug that I should report or is it a known issue? In case the latter is true then what should I do? [1] http://boost.2283326.n4.nabble.com/concept-check-fails-with-types-declared-… [2] https://svn.boost.org/trac/boost/ticket/2037 [3] http://www.boost.org/users/history/version_1_55_0.html WBR, Adam Romanek

1 0

Can gil library in boost 1.49 load png palette images?
by Guangchao Wang 15 Nov '13

15 Nov '13

I'm tried to load a png which can be fund in the attachment, I find the color of this png is PNG_COLOR_TYPE_PALETTE and its depth is 8, I reviewed the code of png_io_private.hpp and found that no such type of PNG is supported. Can you give me some help about how to load such PNG with boost.gil library? Thank you very much. Best, Guangchao

1 0

boost::serialization warnings for template class
by Michele Vallisneri 15 Nov '13

15 Nov '13

Dear Boost experts, I'm serializing (intrusively) a template class Matrix<T> that includes a double-pointer buffer of data of type T. So my save/load pair looks like template<class Archive> void save(Archive & ar, const unsigned int version) const { ar & row_dim; ar & col_dim; for ( size_t i = 0; i < row_dim; i++ ) for ( size_t j = 0; j < col_dim; j++ ) { ar & value[i][j]; } } template<class Archive> void load(Archive & ar, const unsigned int version) { size_t rowdim, coldim; ar & rowdim; ar & coldim; resize( rowdim, coldim ); for ( size_t i = 0; i < rowdim; i++ ) for ( size_t j = 0; j < coldim; j++ ) ar & value[i][j]; } The type of value is T**. The serialization works OK, but I'm getting some warnings that lead me to suspect that Boost is trying to track the addresses of the value[i][j] elements, which is not needed, and presumably a problem for performance. g++ -c -m64 -O3 -Wall -I../../include -I/usr/depot/redhatel5_64/boost/1.48.0/include -o Linux-x86_64/obj/test_Matrix.o test_Matrix.cpp /usr/depot/redhatel5_64/boost/1.48.0/include/boost/mpl/print.hpp: In instantiation of 'boost::mpl::print<boost::serialization::BOOST_SERIALIZATION_STATIC_WARNING_LINE<98> >': /usr/depot/redhatel5_64/boost/1.48.0/include/boost/serialization/static_warning.hpp:92: instantiated from 'boost::serialization::static_warning_test<false, 98>' /usr/depot/redhatel5_64/boost/1.48.0/include/boost/archive/detail/check.hpp:98: instantiated from 'void boost::archive::detail::check_object_tracking() [with T = gcore::Matrix<int>]' /usr/depot/redhatel5_64/boost/1.48.0/include/boost/archive/detail/oserializer.hpp:313: instantiated from 'static void boost::archive::detail::save_non_pointer_type<Archive>::invoke(Archive&, T&) [with T = gcore::Matrix<int>, Archive = boost::archive::text_oarchive]' /usr/depot/redhatel5_64/boost/1.48.0/include/boost/archive/detail/oserializer.hpp:525: instantiated from 'void boost::archive::save(Archive&, T&) [with Archive = boost::archive::text_oarchive, T = gcore::Matrix<int>]' /usr/depot/redhatel5_64/boost/1.48.0/include/boost/archive/detail/common_oarchive.hpp:69: instantiated from 'void boost::archive::detail::common_oarchive<Archive>::save_override(T&, int) [with T = gcore::Matrix<int>, Archive = boost::archive::text_oarchive]' /usr/depot/redhatel5_64/boost/1.48.0/include/boost/archive/basic_text_oarchive.hpp:80: instantiated from 'void boost::archive::basic_text_oarchive<Archive>::save_override(T&, int) [with T = gcore::Matrix<int>, Archive = boost::archive::text_oarchive]' /usr/depot/redhatel5_64/boost/1.48.0/include/boost/archive/detail/interface_oarchive.hpp:63: instantiated from 'Archive& boost::archive::detail::interface_oarchive<Archive>::operator<<(T&) [with T = gcore::Matrix<int>, Archive = boost::archive::text_oarchive]' test_Matrix.cpp:43: instantiated from here /usr/depot/redhatel5_64/boost/1.48.0/include/boost/mpl/print.hpp:55: warning: comparison between signed and unsigned integer expressions Is my suspicion correct, and how do I change my code to avoid the warnings (and performance hits, if any)? Thanks! Michele

1 0

Boost 1.55 and warning in Boost.Circular
by gast128 14 Nov '13

14 Nov '13

Hello all, I just dl 1.55 and I got a warning: D:\Develop\SDK and Libraries\boost_1_55_0\boost/circular_buffer/base.hpp(199): warning C4172: returning address of local variable or temporary Code is crytpic due to all the templates. Is this a client error?

1 2

Boost.Log: Set formatter for named_scope attr.
by David Puffer 14 Nov '13

14 Nov '13

Hi, Being new to Boost.Log (and Boost in general) I have spent the last few days trying to implement Boost.Log - which Is working fairly well by now. However - I encountered one problem that I've not been able to solve by now, although it seems to me quite trivial: I have successfully registered the "named_scope" attribute and I am populating its value by a call to BOOST_LOG_FUNCTION. This works well and I can see the function-name in my log-output. However - I also want the line-number and the file-name, and discovered that there's a possibility to set the output-format of the "named_scope" attribute (which already incorporates file-name and line-number info) like so: #include <boost/log/expressions/formatters/named_scope.hpp> sink->set_formatter(expr::stream << expr::format_named_scope("Scopes", "%n")); As documented here: http://www.boost.org/doc/libs/1_54_0/libs/log/doc/html/log/detailed/express… I just don't seem to find any way to get to my sink - which is configured in a settings-file and evaluated by a call to the init_from_stream() function. Is there any way to query for all the sinks - or get a sink by name? How is one supposed to set a sink formatter for a sink which was generated by the library and not the user? Or is there a way to specify the output-format for named attributes via a settings-file as well? As far as I can see it's only possible to format the output of a log record as a whole - but not single attributes. Also: Is it correct, that the above function will only set the format for the "Scopes" attribute - which means I could call this function repeatedly for lets say the "TimeStamp" attribute as well? Best regards, David

2 2

[Log] Error with operator<< in basic_formatting_ostream
by Víctor Yo 14 Nov '13

14 Nov '13

Hi ewveryone, I'm working lately with Boost.Log (I'm a newbie) library and I want to write a logger with some parameters and formatting options. I followed the documentation, I created a backend and a frontend and I tried to format the output in the frontend with the set_formatter() function. I tried with the Lambda style and with the Boost.Format style but I'm getting an error that I can not locate. Iwas working on it for the last two whole days but I couldn't find a solution. It is related with the basic_formatting_ostream, but I don't understand what that means. I defined the backend and attached to it two streams (to a file and to cout) typedef boost::log::sinks::text_ostream_backend our_backend; typedef boost::log::sinks::synchronous_sink<our_backend> our_frontend; //............................................................................................................................................ boost::shared_ptr<our_backend> backend = boost::make_shared<our_backend>(); //Stream definitions boost::shared_ptr<std::ostream> stream_out(&std::cout, boost::log::empty_deleter()); boost::shared_ptr<std::ostream> stream_file(new std::ofstream("file_logger.log")); backend -> add_stream(stream_out); backend -> add_stream(stream_file); Created the frontend (and wrapping the backend to it) //Wrap the backend into the frontend and register in the core. //The frontend provides synchronization for the backend. boost::shared_ptr<our_frontend> frontend(new our_frontend(backend)); And now come the offending code which is supposed to format the output (note that the first commented block was my first try). None of them works and the error is the same in both cases: */*frontend -> set_formatter* * (* * boost::log::expressions::format("[%1%] %2%, File: %3%, Line: %4% <%5%>: %6% ")* * % boost::log::attributes::counter< unsigned int >(1)* * % boost::log::attributes::local_clock()* * % boost::log::expressions::attr< severity_level >("Severity")* * % boost::log::expressions::message* * );*/* //We are going to define the formatter in a different style to try to get formatted output frontend -> set_formatter ( boost::log::expressions::stream << boost::log::attributes::counter< unsigned int >(1) <<" ["<< boost::log::attributes::local_clock() <<"] <"<< boost::log::expressions::attr< severity_level >("Severity") <<"> "<< boost::log::expressions::message ); The error I get is the next one: */home/victor/Programacion/workspace/opice/build-parallel/../libs/include/boost/log/utility/formatting_ostream.hpp:664:5: error: no match for ‘operator<<’ in ‘(& strm)->boost::log::v2_mt_posix::basic_formatting_ostream<CharT, TraitsT, AllocatorT>::stream [with CharT = char, TraitsT = std::char_traits<char>, AllocatorT = std::allocator<char>, boost::log::v2_mt_posix::basic_formatting_ostream<CharT, TraitsT, AllocatorT>::ostream_type = std::basic_ostream<char>]() << value’* And I really don't know what is happening. I also included a lot of files (I'm sure I included too much files, indeed) to be sure it was not caused by some undefined operator or something similar. However I still have the same problem. The includes are: #include <iostream> #include <fstream> #include <string> #include <boost/thread/shared_mutex.hpp> #include <boost/log/core.hpp> #include <boost/log/common.hpp> #include <boost/make_shared.hpp> #include <boost/log/expressions.hpp> #include <boost/log/utility/setup/file.hpp> #include <boost/log/utility/setup/console.hpp> //puede ser innecesario #include <boost/log/utility/setup/common_attributes.hpp> //puede ser innecesario #include <boost/log/utility/empty_deleter.hpp> #include<boost/thread/mutex.hpp> #include <boost/log/attributes/timer.hpp> #include <boost/log/attributes/mutable_constant.hpp> #include <boost/log/sources/severity_logger.hpp> #include <boost/log/utility/formatting_ostream.hpp> #include <boost/log/sinks/text_ostream_backend.hpp> #include <boost/log/sinks/sync_frontend.hpp> #include <boost/log/support/date_time.hpp> As you can see, they are probably redundant or innecessary but I could rip off that files when I could get the program compiled. Now, I'm out of ideas so any help you could bring me should be very appreciated. Thank you in advance. Victor Hevia

2 4

[1.55] PDF Documentation build available.
by John Maddock 14 Nov '13

14 Nov '13

The final PDF documentation build for Boost-1.55 is now available at https://sourceforge.net/projects/boost/files/boost-docs/1.55.0/ As ever this only contains documentation for the subset of Boost libraries that use quickbook/boostbook/docbook based documentation. Enjoy! John.

1 0

Boost 1.54.0 Header Warnings: Shadowed Variables
by Tom Browder 13 Nov '13

13 Nov '13

I am trying to use Boost 1.54.0 to replace a previous version (1.47) and am using g++ 4.7 with most warnings turned on. I am getting many -Wshadow errors and wonder why they exist. Surely the developers test for such. Here is one example from xpressive/detail/utility/hash_peek_bitset.hpp: Both 'icase' and 'count' are reported to shadow members of 'this'. Note that the bool arg is named 'icase' but the compiler reports 'this' has a member named 'icase' (as can be plainly seen below). In this example it's fairly clear what is intended, but then again maybe not (and it doesn't give the user much confidence in the code). It isn't good practice IMHO to do such so that is why I like to use some prefix or suffix to disambiguate the two variables. // Make sure all sub-expressions being merged have the same case-sensitivity bool test_icase_(bool icase) { std::size_t count = this->bset_.count(); if(256 == count) { return false; // all set already, nothing to do } else if(0 != count && this->icase_ != icase) { this->set_all(); // icase mismatch! set all and bail return false; } this->icase_ = icase; return true; } I could file many bugs but it begs the question of why the warnings are allowed in the first place. Suggestions? Thanks and best regards, -Tom

8 17

[Boost.Iterator] Problem compiling zip_iterator with clang and C++11
by Jonathan Jones 13 Nov '13

13 Nov '13

Hi folks, The following code fails to compile on OSX 10.8.5 with Xcode 5.0.1 and Boost 1.49.0. This is not real world code, but it demonstrates the problem I'm facing. #include <boost/iterator/zip_iterator.hpp> template<typename T> void foo() { boost::zip_iterator<boost::tuple<T*, T*> > iter; std::fill(iter, iter, std::make_pair(T(),T())); } void bar() { foo<int>(); } Here are the commands I used to compile (if I drop "-stdlib=libc++" from the compile line, the code compiles just fine): setenv DEVELOPER_DIR /Applications/Xcode5.0.1.app/Contents/Developer xcrun -sdk macosx10.8 clang++ -c zip_iterator.cpp -I<boost-path>/include -std=c++11 -stdlib=libc++ and here is a heavily formatted snippet of the key error: /Applications/Xcode5.0.1.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/lib/c++/v1/algorithm:1824:1: note: candidate function [with _RandomAccessIterator = boost::zip_iterator<boost::tuples::tuple<int*, int*>>, _Tp = std::__1::pair<int, int>] not viable: no known conversion from typename iterator_traits<zip_iterator<tuple<int*, int*>>>::iterator_category (aka 'boost::detail::iterator_category_with_traversal<std::__1::input_iterator_tag, boost::random_access_traversal_tag>') to 'std::__1::random_access_iterator_tag' for 4th argument __fill(_RandomAccessIterator __first, _RandomAccessIterator __last, const _Tp& __value_, random_access_iterator_tag) Any help is appreciated. Thanks, Jon

1 0

[Math] BOOST_FPU_EXCEPTION_GUARD on Linux
by David Roberts 13 Nov '13

13 Nov '13

Hello, I have a program that makes considerable use of the gamma distribution from Boost.Math. Profiling with OProfile has shown that, on Linux, this program spends about 20% of its runtime calling the functions fesetexceptflag() and feclearexcept(). The code responsible for calling these is in the BOOST_FPU_EXCEPTION_GUARD macro which is defined in boost/math/tools/config.hpp. I have looked back through the source history to see when and why this macro was added, and it was in changeset 3186 from September 2006 with the comment "More linux fixes: finally got AMD64 machines to behave correctly" (see https://svn.boost.org/trac/boost/changeset/3186) In Boost 1.54 the conditions under which the guard macro is defined have been refined since it was originally added, but the guard class itself remains the same as in 2006. Given the performance penalty I am suffering as a result of constantly clearing and reseting the floating point exception flags, I have a couple of questions about this functionality: 1) Can anyone provide a specific test case that did not work without the BOOST_FPU_EXCEPTION_GUARD functionality but was fixed by its addition? It would be very useful to have such a test case because then I can check if this is still a problem on recent versions of Linux. 2) The commit comment for changeset 3186 refers to AMD64. Does this mean the original problem only occurred on AMD CPUs, or does this refer to the architecture, and hence it affected both Intel and AMD CPUs? Many thanks, David Roberts

2 1

[Graph] Static allocation of nodes and edges
by Sensei 13 Nov '13

13 Nov '13

Dear all, I need to optimize my code with regards to both memory and time. The problem is simple: pre-allocating the adjacency matrix. The graph may have a huge number of nodes N, but the number of edges K per node is limited. While N can be in the order of one hundred million, K is at most 10. The graph will be directed, so no symmetry is needed. Do you know if I can pre-allocate such an adjacency matrix? If not, I suppose I must use an adjacency list in order to minimize memory consumption, am I right? Thanks!

2 9

Help with lockfree queue
by TYSON Kieran 13 Nov '13

13 Nov '13

[@@ OPEN @@] Hello, I am using a global lockfree queue to transfer a message from different threads. The program compiles and runs without fault up until the program exits and the queues destructor is called. An unhandled exception is thrown on the destruction of the queue. Below is the code used to set up the queue. #define B1_OUTPUT_LENGTH_PER_CHANNEL (32000) #define B2_OUTPUT_LENGTH_PER_CHANNEL (32000) #define B3_OUTPUT_LENGTH_PER_CHANNEL (64000) #define NB_CHANNELS (8) typedef struct s_OutputMsg { float dataB1[(int)B1_OUTPUT_LENGTH_PER_CHANNEL*NB_CHANNELS]; float dataB2[(int)B2_OUTPUT_LENGTH_PER_CHANNEL*NB_CHANNELS]; float dataB3[(int)B3_OUTPUT_LENGTH_PER_CHANNEL*NB_CHANNELS]; int dataLength; } OutputMsg; boost::lockfree::queue<OutputMsg, boost::lockfree::fixed_sized<true>> dataMsgQueue(100); My first thoughts are that there are limitations on the size of the message because I have used simple integer queues in the past with no problem. Are there size limitations on the message of a lockfree queue? Thanks, Kieran This message contains OPEN information that is not sensitive and can be freely accessed by people both inside and outside of the Thales Group This email was classified by Tyson, Kieran on Wednesday, 13 November 2013 10:58:41 AM ------------------------------------------------------------------------- DISCLAIMER: This e-mail transmission and any documents, files and previous e-mail messages attached to it are private and confidential. They may contain proprietary or copyright material or information that is subject to legal professional privilege. They are for the use of the intended recipient only. Any unauthorised viewing, use, disclosure, copying, alteration, storage or distribution of, or reliance on, this message is strictly prohibited. No part may be reproduced, adapted or transmitted without the written permission of the owner. If you have received this transmission in error, or are not an authorised recipient, please immediately notify the sender by return email, delete this message and all copies from your e-mail system, and destroy any printed copies. Receipt by anyone other than the intended recipient should not be deemed a waiver of any privilege or protection. Thales Australia does not warrant or represent that this e-mail or any documents, files and previous e-mail messages attached are error or virus free. -------------------------------------------------------------------------

2 1

How to convert cmyk jpg file to rgb jpg file with BOOST.GIL?
by Guangchao Wang 13 Nov '13

13 Nov '13

I used following code: rgb8_image_t rgb8Img; jpeg_read_and_convert_image("CMYK.jpg", rgb8Img); //CMYK.jpg is CMYK image jpeg_write_view("RGB.jpg", const_view(rgb8Img)); to convert CMYK image to RGB image. The saved RGB.jpg is not correct. Could you give me some suggestion? Thanks.

2 2

[Locale] Getting correct system locale under Mac OS X
by Orhun Birsoy 11 Nov '13

11 Nov '13

I am using boost 1.55.0.rc2. I am running 10.8.5 and when I select "Norway (Norwegian Bokmal)" (nb_NO) as the region, LANG env. variable is undefined, making boost::locale::generator("") to generate a "C" locale. I can of-course get the locale identifier from NSLocale (or CFLocale) and initialize the boost::locale::generator with "nb_NO" which seems to be working, but with Mavericks things got even more interesting. It looks like, in 10.9 you can mix and match language with a country anyway you like i.e. I can select "Norway" as my region and "English" as my language and again LANG is empty but NSLocale returns "en_NO". But this time when I use boost::locale::generator("en_NO"), boost::locale::as::number uses "." (dot) as decimal separator, although NSLocale reports "," (comma). 1. Can boost::locale handle such combinations (en_NO, en_TR, etc)? (I am using ICU backend, others are explicitly turned off). a. for number formatting? b. for messages formatting (translations)? which mo will it pick for en_TR ? 2. How can I 'customize' locale generated by boost::locale to reflect the changes user might do in system locale settings (i.e. select en_US but change the decimal symbol)? Thanks. -- Orhun Birsoy

2 1

Thanks to everyone for all the work on 1.55.0
by Marshall Clow 11 Nov '13

11 Nov '13

I just wanted to take off my "release manager" hat and say, personally, thank you to all the people who tried out the RCs/Betas/etc, reported bugs, and suggested fixes. You were all a big help in making the release. -- Marshall Marshall Clow Idio Software <mailto:mclow.lists@gmail.com> A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait). -- Yu Suzuki

1 0

Boost 1.55.0 released!
by Marshall Clow 11 Nov '13

11 Nov '13

Release 1.55.0 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 one new library and numerous enhancements and bug fixes for existing libraries. === New Libraries === • Predef: This library defines a set of compiler, architecture, operating system, library, and other version numbers from the information it can gather of C, C++, Objective C, and Objective C++ predefined macros or those defined in generally available headers, from Rene Rivera. === Known bugs with Visual Studio 2013/Visual C++ 12 === Visual Studio 2013 was released quite late in the release process, so there exist several unresolved issues. These include: • Serialization can't compile because of a missing include. • Using has_member_function_callable_with from Boost.Container's allocator_traits causes a compile error (#9332). • For several libraries overloads with intializer lists are incorrectly reported as ambiguous. This is a Visual C++ bug and it isn't clear if there's a good workaround. • In libraries such as Unordered and MultiIndex, calling overloaded functions with initializer lists can result in a compile error, with Visual C++ claiming that the overloads are ambiguous. This is a Visual C++ bug and it isn't clear if there's a good workaround. This won't affect code that doesn't use initializer lists, or uses an initializer list that doesn't require an implicit conversion (i.e. an initializer list of the container's exact value type). • Thread: ex_scoped_thread compile fails (#9333). For details, including download links, see http://www.boost.org/users/news/version_1.55.0 You can also download directly from SourceForge: http://sourceforge.net/projects/boost/files/boost/1.55.0/ To install this release on your system, see http://www.boost.org/doc/libs/release/more/getting_started/index.html Thanks, --The Boost release team Vladimir Prus Rene Rivera Marshall Clow Eric Niebler Daniel James Beman Dawes

2 1

[Polygon] attaching data to polygon boundary segments
by Andreas Buhr 10 Nov '13

10 Nov '13

Dear Boost community, I would like to process 2D geometry for a finite element solver using boost.polygon. I programmed it, it's working fine. I use the polygon_data and polygon_set_data types from boost.polygon. Now I would like to attach information to the boundary segments. Then I'd like to do boolean operations. At the end I would like to have the boundary information merged. Is this possible with boost.polygon? A attached a little drawing of what behavior I would like to achieve, colors code the boundary markers. thanks a lot in advance, Best regards, Andreas

1 1

[BGL] how to use write_graphviz_dp with graph bundle properties
by Pablo Madoery 10 Nov '13

10 Nov '13

Hello. I know how to use write_graphviz_dp with vertices and edges bundle properties but i can't figure out how to print the graph_bundle property of a graph. i have the following code where the line with the error is commented: #include <boost/graph/properties.hpp> #include <boost/graph/graphviz.hpp> #include <boost/graph/adjacency_list.hpp> using namespace boost; using namespace std; struct VertexInfo { int id; }; struct EdgeInfo { int weight; }; struct GraphInfo { int duration; }; typedef adjacency_list<vecS, vecS, undirectedS, VertexInfo, EdgeInfo, GraphInfo> Graph; void printGraph(Graph graph, string file) { ofstream ofs(file.c_str()); dynamic_properties dp; dp.property("node_id", get(vertex_index, graph)); dp.property("label", get(&EdgeInfo::weight, graph)); ///dp.property("label", get(&graph[graph_bundle].duration)); ->error write_graphviz_dp(ofs, graph, dp); } int main() { Graph graph; Graph::vertex_descriptor a = add_vertex(graph); Graph::vertex_descriptor b = add_vertex(graph); Graph::vertex_descriptor c = add_vertex(graph); Graph::vertex_descriptor d = add_vertex(graph); Graph::edge_descriptor e1 = (add_edge(a, b, graph)).first; Graph::edge_descriptor e2 = (add_edge(b, c, graph)).first; Graph::edge_descriptor e3 = (add_edge(c, d, graph)).first; Graph::edge_descriptor e4 = (add_edge(d, a, graph)).first; graph[e1].weight = 1.1; graph[e2].weight = 1.5; graph[e3].weight = 5.2; graph[e4].weight = 1.2; graph[graph_bundle].duration = 10; printGraph(graph, "graph.dot"); }

2 3

[Boost.Python] Exposing protected virtual member functions
by Vitaly Budovski 09 Nov '13

09 Nov '13

Hi, I'd like to expose some protected virtual member functions to Python, but I've been unable to find a practical way to do so. I've created a wrapper making the protected functions public, but have not been able to make further progress due to either compile errors or Python runtime exceptions. Is there something I've overlooked in the documentation? Please see code below. Thanks, Vitaly #include <boost/python.hpp> #include <string> using namespace boost::python; class base { //protected: public: virtual std::string f() { return "base::f"; } public: virtual std::string g() = 0; }; class derived: public base { public: std::string g() { return "derived::g"; } }; struct base_wrapper: base, wrapper<base> { std::string f() { if(override f = this->get_override("f")) { return f(); } return base::f(); } std::string default_f() { return base::f(); } std::string g() { return this->get_override("g")(); } }; BOOST_PYTHON_MODULE(protected) { class_<base_wrapper, boost::noncopyable>("base") .def("f", &base::f, &base_wrapper::default_f) .def("g", pure_virtual(&base_wrapper::g)) ; class_<derived, bases<base>, boost::noncopyable>("derived") .def("g", &derived::g) ; }

1 0