June 2013 - Boost - lists.preview.boost.org

[type_erasure] Compilation time
by Felipe Magno de Almeida 10 Jun '13

10 Jun '13

Hello Steven, Is there anyway to not include some definitions in header to delay instantiation for use with explicit template instantiations in specific translation units with type_erasure? I have a CORBA library which uses type_erasure for each interface defined in IDL. The problem is that if I include too many #include's for the precompiled headers which use type_erasure, then the compiler ends up consuming too much memory (>10GB), which makes it unfit for 32bit compilers which can use at most 3GB. Regards, -- Felipe Magno de Almeida

2 2

Re: [boost] [multiprecision] Are these types suitable as numeric types for unit tests?
by Daryle Walker 09 Jun '13

09 Jun '13

-----Original Message----- From: John Maddock Sent: Saturday, June 08, 2013 4:40 AM >> I'm getting warnings about integer overflow. The problem is at Line >> 59 of boost/test/tools/detail/print_helper.hpp. It sets an output >> stream's precision based on the digits given by std::numeric_limits. > >That's harmless but annoying: it's because std::numeric_limits<cpp_int>::digits is >INT_MAX. The stream precision isn't ever used by integer types anyway... I switched to one of the fixed-sized types (512 bits). >> This time it's an error, at Line 59 of >> boost/test/tools/floating_point_comparison.hpp. There no (close >> enough) matching function for "fpt_abs" from some >> "boost::multiprecision::detail::expression<..." type. (It was >> truncated by the display. I don't know how to get a clean print-out >> of errors from >> Code::Blocks.) I'm guessing the expression template setup is screwing >> things up. > >Yes, it's because Boost.Test can't be used with expression template enabled types. >It's a known issue, and I've added a patch for this: >https://svn.boost.org/trac/boost/ticket/8272, but Genediiy seems to be MIA at present :-( > >Do you particularly want to test with expression-template enabled types (I reaslise >there are cases when you do)? If not then you could use multiprecision types that >have expression templates turned off and that should fix things. I switched to the same type with expression-templates turned off. Daryle W.

1 0

[range] [extensions] Default values for range access adaptors?
by Viktor Sehr 09 Jun '13

09 Jun '13

I'd like to suggest the addition of a default value to the range access adaptors. Either by adding it as an extra, non-necessary parameter for the functions, or by duplicating the functions as "front_or_default(T default value)". For example it would be useful for finding objects which I am not sure exists in a range. std::vector<int> vec = boost::assign::list_of(2)(4)(6); auto first_odd = vec | filtered(is_odd) | front(); // undefined? auto first_odd = vec | filtered(is_odd) | front_or_default(-1); // everything is fine! Another option would be a function which returns the adress, or nullptr if the value does not exist? /Viktor

5 14

boost.range bug in transformed or sliced
by John Reid 09 Jun '13

09 Jun '13

Suppose you wish to transform and slice a random access range. Surprisingly the order in which you apply the operations makes a big difference to performance. I'm guessing this is not desired behaviour, correct me if I'm wrong. AFAICT when transforming happens before slicing the iterators involved in the internals of boost.range are being treated as forward iterators not random access iterators. An advance() operation is made on one of the iterators with a negative number and this advance is O(n) for forward iterators but O(1) for random access. The code below demonstrates (adapted from sliced example code). On an unrelated note the documentation for transformed does not mention that the function is part of the range return type. It is given as: boost::transformed_range<typeof(rng)> Code to demonstrate the transform then slice problem: #include <boost/range/adaptor/transformed.hpp> #include <boost/range/adaptor/sliced.hpp> #include <boost/range/algorithm/copy.hpp> #include <boost/assign.hpp> #include <iterator> #include <iostream> #include <vector> #include <functional> struct identity { typedef int result_type; result_type operator()( int i ) const { return i; } }; int main(int argc, const char* argv[]) { using namespace boost::adaptors; using namespace boost::assign; std::vector< int > input; input += 1,2,3,4,5,6,7,8,9; // slicing then transforming my iterator range std::cout << "Sliced then transformed: "; boost::copy( input | sliced( 2, 8 ) | transformed( identity() ), std::ostream_iterator< int >( std::cout, ",") ); std::cout << "\n"; // transforming then slicing my iterator range - takes a very long time.... std::cout << "Transformed then sliced: "; boost::copy( input | transformed( identity() ) | sliced( 2, 8 ), std::ostream_iterator< int >(std::cout, ",")); std::cout << "\n"; return 0; }

4 5

merge of interface modification?
by Oliver Kowalke 09 Jun '13

09 Jun '13

Do I get permission to merge a change in the interface of boost.coroutine into release?

2 3

Re: [boost] [Boost-users] [thread] synchronized_value: value and move semantics
by Klaim - Joël Lamotte 08 Jun '13

08 Jun '13

I just discovered synchronized_value from reading the 1.54.0 beta documentation. On Wed, Feb 20, 2013 at 7:02 PM, Vicente J. Botet Escriba < vicente.botet(a)wanadoo.fr> wrote: > Do you see something completely wrong with this addition? > No idea yet but I'm willing to test it in production code. See below for minor details. > Has some of you had a need for this? Could you share the context? > Yes, I think so. I have some cases (which I'm still working on but will be published as OSS soon) where I do something like that: struct ThingInfo { Id<ObjectInfo> id; std::string name; URI location; }; struct ThingContent { std::vector< Id<Foo> > foo_id_list; std::vector< Id<Bar> > bar_id_list; }; // this type MUST have thread-safe interface class Thing { public: explicit Thing( ThingInfo info ); ThingInfo info() const { boost::lock_guard<boost::mutex> lg( m_info_mutex ); return m_info; } ThingContent content() const { boost::lock_guard<boost::mutex> lg( m_content_mutex ); return m_content; } // + several modifying functions that use the work queue like that: void add( std::shared<Bar> bar ) { m_work_queue.push( [this, bar]{ m_bars.emplace_back( bar ); { boost::lock_guard<boost::mutex> lg( m_content_mutex ); m_content.bar_id_list.emplace_back( bar->id() ); } }); } void rename( std::string new_name ) { m_work_queue.push( [this, new_name]{ boost::lock_guard<boost::mutex> lg( m_content_mutex ); // SILENT ERROR!!! m_info.name = new_name; }); } private: WorkQueue m_work_queue; // tasks executed later std::vector< std::shared<Foo> > m_foos; // manipulated only by code in the work queue std::vector< std::shared<Bar> > m_bars; // idem ThingInfo m_info; // should be manipulated only with m_info_mutex locked ThingContent m_content; // should be manipulated only with m_content_mutex locked boost::mutex m_info_mutex; // protect m_info boost::mutex m_content_mutex; // protect m_content }; This is not real code but just how it looks like in my work-in-progress classes in my OSS project. >From reading the documentation, using synchronized_value I would then change the Thing implementation to: // this type MUST have thread-safe interface class Thing { public: explicit Thing( ThingInfo info ); ThingInfo info() const { return *m_info; } ThingContent content() const { *return m_content; } // + several modifying functions that use the work queue like that: void add( std::shared<Bar> bar ) { m_work_queue.push( [this, bar]{ m_bars.emplace_back( bar ); m_content.synchronize()->bar_id_list.emplace_back( bar->id() ); }); } void rename( std::string new_name ) { m_work_queue.push( [this, new_name]{ m_info.synchronize()->name = new_name; // OK: CAN'T USE THE WRONG MUTEX!! }); } private: WorkQueue m_work_queue; // tasks executed later std::vector< std::shared<Foo> > m_foos; // manipulated only by code in the work queue std::vector< std::shared<Bar> > m_bars; // idem boost::sychronized_value<ThingInfo> m_info; boost::sychronized_value<ThingContent> m_content; }; This version is to me: - more explicit on reading; - shorter; - avoid some mistakes like the one pointed in the comments - which might be hard to spot; The only thing that bother me is that synchronized_value is a long name and the "_value" part is, in my opinion, too much. Also, maybe using the pointer-semantic operators is not the best idea. I guess that if std::optional does, then it's ok. Once Boost 1.54.0 is released I will have the opportunity to try it by refactoring my code. I'll report if I find issues. Joel Lamotte

1 1

[Boost.Test] Error with "putenv" when using (1.53) release branch
by Daryle Walker 07 Jun '13

07 Jun '13

This happened to me a few months ago when trying out a Windows 8 compiling setup with Code::Blocks. I pointed to the release branch of Boost, which should be set to 1.53. I included Boost.Test, I think by #including boost/test/include/unit_test.hpp so everything comes in without linking in the library directly. There was some sort of error involving "putenv," not sure what anymore. I "fixed" it by switching to the trunk branch. Now I'm trying a new setup, with Eclipse. This time I'm adding Boost.Test by adding the files directly to the project list. (I'm experimenting with splitting my tests into 3 files.) I'm referencing the release branch again, which should still be on 1.53. I got "'putenv' was not declared in scope" error. It's at Line 95 of boost/test/utils/runtime/config.hpp. Did a hard error sneak by everyone, or is my setup incomplete? Both IDE setups use G++ on MinGW for the compiler. Is there some #define I'm missing? Should I switch to using the trunk again? Daryle W.

1 0

All talks for Meeting C++ 2013 are online!
by Jens Weller 06 Jun '13

06 Jun '13

Hi boost & C++ Community, to keep this short, all talks are online now, the schedule comes next week: http://meetingcpp.com/index.php/talks.html kind regards, Jens Weller

1 0

merge of bugfixes still possible
by Oliver Kowalke 06 Jun '13

06 Jun '13

Hi, I got a bug-report two days before - I'd like to merge the bugfix into the release. Do I get the permission? Oliver

2 1

[GIL] Can image::recreate reuse the memory?
by TONGARI 06 Jun '13

06 Jun '13

Hi, I'd expect image::recreate to reuse the memory much like std::vector::resize, but looked into the code shows that it creates a new one and swaps... So, is there a way to 'resize' the image without reallocating if the internal memory is sufficient? Thanks

2 17

server and client chat via boost::asio
by Nguyễn Trường Lâm 06 Jun '13

06 Jun '13

Hello everyone. I want to write a program using socket, a client and a server to communicate between each other. But i dont know how write a message to server and display in a IDE terminal. Thank you for your help

2 1

Migrating svn.boost.org
by DongInn Kim 06 Jun '13

06 Jun '13

Hi, This time I would like to setup a schedule to migrate svn.boost.org to the new server (wowbagger.crest.iu.edu) This migration may need some downtime on the subversion, trac, and svnmanager services because we do not want to have any inconsistencies in SVN and or trac database. I have already rsynched the 1) boost SVN repository including its hook script, 2) trac db, and 3) SVNManager and its database. Do you have any hidden processes to manage the subversion repository, trac, or SVNManager? I implemented some simple testings with some SVN activities on the new server and it seems to work fine. Can anyone volunteer to play with the migrated SVN repository, trac, and SVNManager on the new server? I can test the very basic stuff and I may miss some functionalities of the current services. If anyone who actually work on the boost project can check this out, it would be much appreciated. I would feel much safer. 1. The downtime that we may have while migrating svn.boost.org I expect that TTL value be changed to 1hour or so. While the new DNS entries are populated, the current server will NOT service svn.boost.org and the new server will instead. - Date: Wednesday June 5th, 2013 - Time: 11:00am-12:00pm Pacific US time 12:00pm-1:00pm Mountain US time 1:00pm-2:00pm Central US time 2:00pm-3:00pm Eastern US time 6:00pm-7:00pm GMT 2. The unavailable services hosted on svn.boost.org : - Subversion - trac - SVNManager 3. The DNS entries to update: - Make svn.boost.org and wiki.boost.org point to 129.79.13.33 Please let me know if you have any questions or concerns about this migration. Regards, -- - DongInn

7 15

Re: [boost] [circular_buffer] New version of documentation
by Paul A. Bristow 06 Jun '13

06 Jun '13

> -----Original Message----- > From: Boost [mailto:boost-bounces@lists.boost.org] On Behalf Of Marshall Clow > Sent: Thursday, May 23, 2013 8:01 PM > To: boost(a)lists.boost.org > Subject: Re: [boost] [circular_buffer] Volunteer(s) needed After more work than I bargained for (see below for excuses) I have produced a Quickbook, Doxygen, Autoindex version from the existing documentation for Boost.Circular_Buffer at https://svn.boost.org/svn/boost/trunk/libs/circular_buffer/doc/circular_buf… I have made few substantive changes to the text, mainly adding more hyperlinks, (and correcting some that were wrong of missing - using the inspect tool is a really, really good idea!) So that users can comment on this new version, I have put html and pdf versions in my Dropbox. The pdf version is at https://dl.dropboxusercontent.com/u/43940943/circular_buffer.pdf which can of course be viewed by any PDF reader (I have recently moved to using Foxit instead of Adobe Reader). The html Quickbook version is from folder (libs/circular_buffer/doc/html) the at https://dl.dropboxusercontent.com/u/43940943/html/index.html (Sadly some of the icon .png files are inaccessible from the Dropbox version, which mars its beauty but you'll get the idea). This library uses the full automatic indexing of C++ names and hyperlinking to and from the text. The header files have the classes and functions fully documented in Doxygen format, and this is used to create the C++ reference sections: https://dl.dropboxusercontent.com/u/43940943/html/boost_circular_buffer_c__… to that you can see what the classes do and the parameters, template parameters, returns etc for the functions. I believe that these details are essential for Really Useful documentation. The original Doxygen standalone (whose format some users prefer) is at https://dl.dropboxusercontent.com/u/43940943/doxygen/html/index.html So you can see the pros and cons of the same information displayed side by side. The PDF and HTML versions also has an index at the end (this could be improved as it only gives C++ functions etc at present). Enjoy! (?) Comments welcome - praise preferred ;-) Paul PS Implementation notes: To get all the links correct proved *very* tiresome. Much of the trouble was inconsistent layout of existing documentation and files. IMO we need to use the GITMOD project to impose much more order on the file structure. I fear that the modularization changes will invalidate many links in documentation, so much revision will be needed :-( We must be able to predict where index.html etc will be to be able to write links without the current hassle. The Boost inspect tool proved invaluable, and all links are now reported correct. (Lots were wrong - in Doxygen comments in the header files, and in the Quickbook stuff :-( ) Despite this, there are still some sections in red where the html layout is wrong :-( I hope to find all of these in a later revision. --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow(a)hetp.u-net.com

2 3

Ticket spam
by Hartmut Kaiser 06 Jun '13

06 Jun '13

This ticket got some (harmless) spam added: https://svn.boost.org/trac/boost/ticket/6126#comment:21. What can I do to remove that comment? Regards Hartmut --------------- http://boost-spirit.com http://stellar.cct.lsu.edu

2 1

[community] request for revision: generic programming
by Steve Lorimer 06 Jun '13

06 Jun '13

In the light of C++11, the page on generic programming techniques http://www.boost.org/community/generic_programming.html is somewhat outdated. Any chance of a revision?

2 1

BOOST_STRONG_TYPEDEF
by Eric Niebler 05 Jun '13

05 Jun '13

Someone on stackoverflow is having a problem with BOOST_STRONG_TYPEDEF getting defined twice. There's a good recommendation there about boost/strong_typedef.hpp merely including boost/serialization/strong_typedef.hpp. Regardless, it's strange that trunk and release have been out of sync in this regard for so long. Robert? http://stackoverflow.com/questions/16901250/why-does-boost-include-two-diff… -- Eric Niebler Boost.org http://www.boost.org

1 0

[DNS] Heads up: svn.boost.org moving to a new server; IP address changing
by Beman Dawes 05 Jun '13

05 Jun '13

At 2:00 pm Eastern US time (i.e 30 minutes from now) we will switch svn.boost.org to a new server with a new IP address. It would be best to hold off on any SVN commits until after the change over, and after your DNS server has the new IP address. If command line "ping svn.boost.org" is going to [129.79.245.252], you are still getting the old IP address. Once svn.boost.org is resolving to [129.79.13.13] it should be OK to resume normal Subversion activity. Thanks, --Beman

1 2

[1.54.0] Beta 1 Available
by Marshall Clow 05 Jun '13

05 Jun '13

Boost release 1.54.0 beta 1 is now available from SourceForge See http://sourceforge.net/projects/boost/files/boost/1.54.0.beta.1/ This release contains 3 new libraries: • Log: Logging library, from Andrey Semashev. • TTI: Type Traits Introspection library, from Edward Diener. • Type Erasure: Runtime polymorphism based on concepts. and numerous enhancements and bug fixes for existing libraries. For details of what's in the release, see http://www.boost.org/users/history/version_1_54_0.html. Note that the links to files on this web page are for the final release - use the SourceForge link above to get the beta files. Please download the beta, give it a try, and report any problems you encounter. Thanks, -- The Boost Release Team

3 2

[git] Library metadata
by Daniel James 04 Jun '13

04 Jun '13

Hello, Is anyone thinking about storing library metadata in modules? I'm thinking about this in context of the data required for the website, it'd be nice to automatically pull that from the library modules rather than having to maintain it myself, but I guess this might be something that others are thinking about (e.g. for the expected test failures, the library maintainer). So I was wondering if anyone has any plans? For anyone who isn't aware of it, the website data is stored in an xml file at: http://svn.boost.org/svn/boost/website/public_html/live/doc/libraries.xml Of course, it could be desirable to store modularized metadata in a different format. Daniel

3 4

[python] non-standard iteration / wrapping member function return value
by Matus Chochlik 04 Jun '13

04 Jun '13

Hi, I'm stuck at trying to make python bindings for a set of classes that use a non-c++-standard iteration protocol. I've made a minimal example of what I'm trying to do: the python-ized classes look in essence like this: //------------------------------------------------------------- class int_range { private: int i, max; public: int_range(int m) : i(0), max(m) { } bool empty(void) const { return i > max; } int front(void) const { return i; } void next(void) { ++i; } }; class range_maker { public: int_range make(void) const { return int_range(10); } }; //------------------------------------------------------------- the boost.python module that I've come up with so far: //------------------------------------------------------------- using namespace boost::python; class _py_int_range_iter { private: int_range _range; public: _py_int_range_iter(int_range range) : _range(range) { } _py_int_range_iter self(void) { return *this; } int next(void) { if(_range.empty()) objects::stop_iteration_error(); int result = _range.front(); _range.next(); return result; } }; class_<_py_int_range_iter>("_py_int_range_iter", init<const int_range&>()) .def("self", &_py_int_range_iter::self) .def("next", &_py_int_range_iter::next) ; class _py_int_range { private: int_range _range; public: _py_int_range(int_range range) : _range(range) { } _py_int_range_iter iter(void) { return _py_int_range_iter(_range); } }; class_<_py_int_range>("_py_int_range", init<const int_range&>()) .def("__iter__", &_py_int_range::iter) ; //------------------------------------------------------------- the problem is that when I register range_maker as follows... //------------------------------------------------------------- class_<range_maker>("range_maker") .def("make", &range_maker::make) ; //------------------------------------------------------------- ... it compiles but (obviously) fails in python at run-time because range_maker::make returns int_range and not _py_int_range. what do I need to do to make this work? I know that I could write _py_range_maker wrapping range_maker and let it return _py_int_range, but this approach does not scale very well. Is there some way to tell the def(...) function that I want to wrap the return value into an object of another type? Is there a return value policy for that or do I need to write my own? If so how? Thanks in advance Matus

2 2

Re: [boost] ATTN: Boost GSoC mentors and students: Continuous Integration for GSoC thoughts
by Niall Douglas 04 Jun '13

04 Jun '13

> > Personally speaking, I think this depth and degree of CI is > > *invaluable* to modern software development practice. > > I could not agree more. > > One thing you may want to add is coverage analysis with > https://coveralls.io/: > > If you build with gcc, pass "-fprofile-arcs -ftest-coverage" both to the compiler > and the linker. After running the tests, you can analyze the coverage with gcov > and send it to coveralls. > To collect and upload, I use the following script: > https://github.com/purpleKarrot/Karrot/blob/develop/test/coveralls.in > (Note that gcov 4.6 does not support all flags that I use in the script, I use it wiht > 4.8 only). That's really super useful of you Daniel. I wasn't aware that such a service existed for code coverage. Thank you. I don't suppose you'd know the answer to this question though: Travis CI's reporting is not great, it either reports success or failure and not *what* failed nor *why*. Specifically, if for example a build fails, I'd like to know which source files and what the compiler errors were. Similarly, if a unit test fails, I want to know which one and what the error was. Do you know if this can be done? Niall --- Opinions expressed here are my own and do not necessarily represent those of BlackBerry Inc.

1 0

Re: [boost] ATTN: Boost GSoC mentors and students: Continuous Integration for GSoC thoughts
by Niall Douglas 04 Jun '13

04 Jun '13

> Just curious: is Visual Studio support not included because not deemed > important? Or not supported by this particular lib? Or not supported by Travis CI > for some reason? (licensing?) As others have mentioned, Travis CI provides only Ubuntu 12.04 x64 instances and none other. Interestingly, you /could/ actually package up MSVC and run it under wine, but Travis CI's free service limits test runs to 20 mins and getting wine et al installed in that time is probably more trouble than it's worth. Besides, I have another very good reason not to: my primary development platform is Visual Studio. Why? It has the least C++11 support, but also the best debugger. Any C++11 feature supported by the Nov 2012 CTP experimental MSVC is easily supported in GCC or clang, whereas the converse is most certainly not true. A further factor is that the library is an async file i/o abstraction layer intended for Boost, and the only true async backend currently is Windows IOCP. Therefore, naturally development, and especially performance tuning, currently focuses there. Travis CI is really useful though for checking Visual Studio work: MSVC I am told does not use an AST internally, and therefore permits C++ no other compiler will allow. Very often I write template metaprogramming which MSVC is absolutely fine with but chokes either GCC or clang or both. Travis CI emails me to tell me when I've done that. This is very, very useful. Niall --- Opinions expressed here are my own and do not necessarily represent those of BlackBerry Inc.

1 0

[thread] call_once regression or it is planed so?
by Antony Polukhin 04 Jun '13

04 Jun '13

Hi, With 1.54.0_beta1 following code fails to compile: #include <boost/thread/once.hpp> void init_host(int); void init_host(); boost::once_flag instance_flag = BOOST_ONCE_INIT; int main() { boost::call_once(&init_host, instance_flag); } In 1.53 it worked fine. Is that meant to be so and no workaround is possible, or is it a bug? -- Best regards, Antony Polukhin

3 2

Asking for permission to merge small bugfix (Spirit)
by Hartmut Kaiser 04 Jun '13

04 Jun '13

I would like to merge a small bugfix to the release branch (commit set 84631, patch below). Regards Hartmut --------------- http://boost-spirit.com http://stellar.cct.lsu.edu Index: C:/CVS/boost/boost/spirit/home/lex/lexer/support_functions.hpp =================================================================== --- C:/CVS/boost/boost/spirit/home/lex/lexer/support_functions.hpp (revision 84630) +++ C:/CVS/boost/boost/spirit/home/lex/lexer/support_functions.hpp (revision 84631) @@ -49,7 +49,7 @@ { typedef typename typename remove_reference< - remove_const< + typename remove_const< typename mpl::at_c<typename Env::args_type, 4>::type >::type >::type

2 1

[1.54.0] Accidentally committed to release docs fix
by Antony Polukhin 04 Jun '13

04 Jun '13

Hi, Without permission, accidentally committed r84603 that fixes docs of Boost.Any. Is it OK to leave that commit or reverting it is preferable? Commit is small and fixes only a single line in docs. -- Best regards, Antony Polukhin

2 1