- Boost - lists.boost.org

share_ptr serialization
by Robert Ramey 09 Jun '05

09 Jun '05

In the course if making a new shared_ptr seriailzation implementation, I've come upon a question. If a shared pointer has been created with a custome deleter, how is one to determine what kind of deleter is attached without opening the implementation of shared_ptr? That is, I see no way to correctly serialize/deserialize a shared_ptr with a custom deleter from the published interface of shared_ptr. Can anyone explain how to do this? Robert Ramey

3 2

[1.33.0] Timeline and remaining issues
by Douglas Gregor 09 Jun '05

09 Jun '05

I want to be rolling prerelease tarballs by next Friday, June 10th, and hope to release 1.33.0 shortly thereafter. To get there, I need your help. We've just brought the "Unresolved issues" page back online at: http://engineering.meta-comm.com/boost-regression/CVS-HEAD/developer/ issues.html This reflects all of the bugs that we need fixed or accounted for before we release Boost 1.33.0. The page itself has some flaws that we are aware of (and will work on), but please verify that your libraries don't have any entries on this page. Click on the name of the library to see the full set of results for the compilers important for this release. Focus on those compilers; ask me first before making changes to accommodate other compilers. We're...almost...there. Doug

7 11

Bug-Report.
by Juan Antonio Gálvez Gamboa 09 Jun '05

09 Jun '05

Greetings to everyone. It seems there´s a bug when using the astar_search.hpp in the dummie property map section. namespace detail { template <class VertexListGraph, class AStarHeuristic, class CostMap, class DistanceMap, class WeightMap, class IndexMap, class ColorMap, class Params> inline void astar_dispatch2 (VertexListGraph& g, typename graph_traits<VertexListGraph>::vertex_descriptor s, AStarHeuristic h, CostMap cost, DistanceMap distance, WeightMap weight, IndexMap index_map, ColorMap color, const Params& params) { dummy_property_map p_map; typedef typename property_traits<CostMap>::value_type C; astar_search (g, s, h, choose_param(get_param(params, graph_visitor), make_astar_visitor(null_visitor())), choose_param(get_param(params, vertex_predecessor), p_map), cost, distance, weight, index_map, color, choose_param(get_param(params, distance_compare_t()), std::less<C>()), choose_param(get_param(params, distance_combine_t()), closed_plus<C>()), choose_param(get_param(params, distance_inf_t()), std::numeric_limits<C>::max()), choose_param(get_param(params, distance_zero_t()), C())); } I´m using Visual C++ .NET 2003 and it seems it´s getting a confusion between: std::numeric_limits<C>::max()) and a macro called max, defined somewhere, whose body would be: #define max(((a)>(b))?(a):(b)) I got the definition of this macro pointing with the mouse cursor over the "max" word on std::numeric_limits<C>::max()) .Before trying to use the astar_search() function, in the copilements always a warning telling "not enough parameters for the macro max" has appeared (that´s logic if it´s interpreted as a macro -where are paseed the two parameters?-) but the questions is... it´s necessary to be interpreted as the "max" numeric_limit. I´m not sure if this is an atar_search bug but, in case it´s not and ,so that, I have a "max" macro defined, any suggestions where should I fin it for modifying it´s name to MAX or directly deleted from the header file?. Thanks for your time... ______________________________________________ Renovamos el Correo Yahoo! Nuevos servicios, más seguridad http://correo.yahoo.es

2 1

Re: Who's Using Boost?
by Jeffrey Williams 09 Jun '05

09 Jun '05

Hi boosters, We use boost too, here is my "report". Application: Rule in Hell Company: Megahard Software Technologies Inc. Description of Application: Massively Multiplayer Online Role Playing Game (still in beta) Boost libs used: bind function any tuples ref shared_ptr type_traits noncopyable By *FAR* the combination of bind, function,shared_ptr and ref is what we use most heavily. As time goes on we tend to replace our custom code with boost libs that perform more or less the same functionality (where it makes sense) url: http://www.ruleinhell.com

1 0

[parameter] missing docs?
by Joaquín Mª López Muñoz 09 Jun '05

09 Jun '05

Are the docs for Boost.Parameter already into the CVS? I cannot seem to find them. Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

[filesystem] error codes revisited
by Beman Dawes 09 Jun '05

09 Jun '05

Peter Dimov suggested that Boost.Filesystem's error reporting codes use the POSIX <cerrno>/<errno.h> codes as the portable way to encode errors. That made a lot of sense for POSIX systems, and since POSIX is now an ISO standard (ISO/IEC 9945:2002), other portions of Boost.Filesystem are also as of the i18n branch being formally defined by reference to the ISO POSIX standard. Checking how the various standards interrelate as far as errno.h, cerrno, and errno definitions are concerned, the C++ and POSIX standards both defer to the C standard, although POSIX does add additional codes. No problems there. As far as actual header implementations go, there was a concern that the Windows errno.h implementations didn't provide the needed definitions (since C only mandates EDOM, EILSEQ, and ERANGE). That turned out to be a non-problem. The Windows implementations only provide a subset of the POSIX macros, but with few exceptions they do provide all the macros Boost.Filesystem users will need. One exception was that Borland fails to define one of the needed macros. There were also two cases where additional non-POSIX macros were needed, but the standards do permit extensions, so again no problem. That has been worked around by providing a <boost/filesystem/cerrno.hpp> header which sucks in <cerrno> and adds any needed definitions. I had some concerns about Boost.Filesystem headers gratuitously including <cerrno> (or <boost/filesystem/cerrno.hpp>) because that could cause macro name clashes. But it turned out the Boost.Filesystem headers have no reason to include <cerrno> or <boost/filesystem/cerrno.hpp>. So no problem. The CVS filesystem i18n branch now has the POSIX aligned error mechanisms in place. Seems to work well and be a very nice improvement to the library. Thanks to Peter for the suggestion! --Beman

1 0

[boost][date_time] errors in time_facet.hpp (cvs HEAD)
by Jan Hermelink 09 Jun '05

09 Jun '05

I think there are several errors in time_facet.hpp from cvs HEAD: The lines using std::basic_string::find() evaluate the return code wrongly, e.g. line 227: if (format.find(seconds_with_fractional_seconds_format)) { should be if (format.find(seconds_with_fractional_seconds_format) != string_type::npos) { I think. In addition I think "%%" should always evaluate to "%" and leave the following char unexpanded. Current e.g. "%%s" evaluates to "%59.000000", not to "%s". This is the outcome of using boost::algorithm::replace_all(). Otherwise thank you for your great library! Jan

2 1

Serialization - no & operator for primitives
by Marcin Tustin 09 Jun '05

09 Jun '05

I'm using a recent (a few days old) cvs version of boost, and the serialization libraries, and Visual Studio 7.1 is producing the error below. This is only happening to classes which I am exporting (like BOOST_CLASS_EXPORT_GUID(D2::Message,"D2::Message");). The reason I am exporting them is that they are in many cases a) subclasses of each other b) accessed through a pointer. Any help would be deeply appreciated. My code is: template<class Archive > static void serialize(Archive & ar, D2::Message& g, const unsigned int version) { ar & g.getType(); // an int - error complains about this line. ar & g.getSender(); // a string ar & g.getMessage(); // a string } Error message: c:\D2\API\NewClient\EnhancedMessageHandler.h(132) : error C2679: binary '&' : no operator found which takes a right-hand operand of type 'uint32' (or there is no acceptable conversion) C:\boost\boost-cvs-20050606\boost\boost\serialization\serialization.hpp( 140) : see reference to function template instantiation 'void boost::serialization::serialize<Archive>(Archive &,D2::Message &,const unsigned int)' being compiled with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,IplJrsMediaTimePoint>::tail,IplJrsMediaTimePoint>: :head ] C:\boost\boost-cvs-20050606\boost\boost\archive\detail\iserializer.hpp(1 61) : see reference to function template instantiation 'void boost::serialization::serialize_adl<Archive,T>(Archive &,T &,const unsigned int)' being compiled with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,IplJrsMediaTimePoint>::tail,IplJrsMediaTimePoint>: :head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\serialization\extended_type_info _typeid.hpp(71) : while compiling class-template member function 'void boost::archive::detail::iserializer<Archive,T>::load_object_data(boost:: archive::detail::basic_iarchive &,void *,const unsigned int) const' with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\archive\detail\iserializer.hpp(1 72) : see reference to class template instantiation 'boost::archive::detail::iserializer<Archive,T>' being compiled with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\archive\detail\iserializer.hpp(1 71) : while compiling class-template member function 'const boost::archive::detail::basic_iserializer &boost::archive::detail::pointer_iserializer<T,Archive>::get_basic_seria lizer(void) const' with [ T=D2::Message, Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head ] C:\boost\boost-cvs-20050606\boost\boost\archive\detail\iserializer.hpp(5 01) : see reference to class template instantiation 'boost::archive::detail::pointer_iserializer<T,Archive>' being compiled with [ T=D2::Message, Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head ] C:\boost\boost-cvs-20050606\boost\boost\serialization\export.hpp(78) : see reference to function template instantiation 'const boost::archive::detail::basic_pointer_iserializer &boost::archive::detail::instantiate_pointer_iserializer<Archive,T>(Arch ive *,T *)' being compiled with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\serialization\export.hpp(74) : while compiling class-template member function 'void boost::archive::detail::export_impl::archive<Archive,T>::i::invoke(void) ' with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\serialization\export.hpp(105) : see reference to class template instantiation 'boost::archive::detail::export_impl::archive<Archive,T>::i' being compiled with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\serialization\export.hpp(89) : while compiling class-template member function 'void boost::archive::detail::export_impl::archive<Archive,T>::instantiate(voi d)' with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\serialization\export.hpp(116) : see reference to class template instantiation 'boost::archive::detail::export_impl::archive<Archive,T>' being compiled with [ Archive=boost::archive::detail::export_impl::for_each_archive<boost::arc hive::detail::export_impl::for_each_archive<boost::archive::detail::know n_archive_types::type,D2::Message>::tail,D2::Message>::head, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\serialization\export.hpp(115) : while compiling class-template member function 'void boost::archive::detail::export_impl::for_each_archive<ASeq,T>::instantia te(void)' with [ ASeq=boost::archive::detail::export_impl::for_each_archive<boost::archiv e::detail::known_archive_types::type,D2::Message>::tail, T=D2::Message ] C:\boost\boost-cvs-20050606\boost\boost\serialization\export.hpp(122) : see reference to class template instantiation 'boost::archive::detail::export_impl::for_each_archive<ASeq,T>' being compiled with [ ASeq=boost::archive::detail::export_impl::for_each_archive<boost::archiv e::detail::known_archive_types::type,D2::Message>::tail, T=D2::Message ] Marcin Tustin IT Innovation Centre 2 Venture Road Chilworth Science Park Southampton, SO16 7NP, UK tel: +44 23 8076 0834 fax: +44 23 8076 0833 mailto:mt@it-innovation.soton.ac.uk http://www.it-innovation.soton.ac.uk

2 1

[testing] [math] [gcc] Problems with long double math functions
by Caleb Epstein 09 Jun '05

09 Jun '05

These test failures all seem to stem from problems with std::exp (long double) and perhaps some other math functions using long double on gcc. Does this smell like a gcc bug, or is there something wrong with the tests? gcc-4_0-darwin: http://tinyurl.com/8tts6 http://tinyurl.com/atzbx gcc-windows: http://tinyurl.com/dwmdk http://tinyurl.com/9to33 gcc-3_4_3-sunos: http://tinyurl.com/8zvpz http://tinyurl.com/bx6mt -- Caleb Epstein caleb dot epstein at gmail dot com

2 1

[program_options, type_traits] is_reference and VC 7.0
by Vladimir Prus 09 Jun '05

09 Jun '05

Hi, I get program_options failures on VC 7.0 that result from the following sequences of events: variable_map::as<string> is called any_cast<string&> is called the following code in any.hpp gives static assertion: typedef BOOST_DEDUCED_TYPENAME remove_reference<ValueType>::type nonref; #ifdef BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION // If 'nonref' is still reference type, it means the user has not // specialized 'remove_reference'. // Please use BOOST_BROKEN_COMPILER_TYPE_TRAITS_SPECIALIZATION macro // to generate specialization of remove_reference for your class // See type traits library documentation for details BOOST_STATIC_ASSERT(!is_reference<nonref>::value); The problem is that on VC 7.0 specializations for remove_reference must be defined explicitly, and it's not done for string and wstring. I see three choices: 1. Declare specialization for string and wstring in program_options 2. Declare them somewhere in type_traits 3. Mark 7.0 as unsupported. 4. Don't use any_cast to reference type I don't like (4) because I've introduced cast to reference myself explicitly for program_options sake. (1) can lead to conflicts with user code. So it leaves (2) and (3). Any opinions which one is better? - Volodya

5 11