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

Fwd: Re: [Boost-users] Re: Re: Re: Re: Boost serializationanddependencynightmares...
by kauai 11 May '05

11 May '05

Hi Robert et all, Ok details. Now that I am able to compile/link/run using that macro Jeff provided I can move on. Quick summary, for legacy reasons we have a structure called ByteArray passed around everywhere. It is simply a public class like this: class ByteArray { public: unsigned char *pArray; int length; int offset; } On our receiving side we have a piece of code like this: // The function argument(s) have to be like this unfortunately. Also pArray is serialized by the sender with code like this but using binary_oarchive OrderedCollection* OrderedCollection::GetOrderedCollection(ByteArray *pArray) // this is a static method, pArray points to valid heap memory { OrderedCollection *pCollection = NULL; std::string astring ((char*) (pArray->pData), pArray->length + pArray->offset); std::istringstream iss(astring); boost::archive::binary_iarchive ia(iss); // exception here in Windows not Linux, below is the reason and call stack is << BOOST_SERIALIZATION_NVP(pCollection); return pCollection; } Again this works in Linux, and I've run it through Valgrind (a linux tool) with no errors or leaks. If there is a better way to construct a binary input archive I'd like to know, im new to stl classes. I get a std::bad_alloc exception, here is the call stack: cmnTest2.exe!boost::archive::basic_binary_iprimitive<boost::archive::binary_iarchive,std::basic_istream<char,std::char_traits<char> > >::load(std::basic_string<char,std::char_traits<char>,std::allocator<char> > & s={...}) Line 93 C++ > cmnTest2.exe!boost::archive::load_access::load_primitive<boost::archive::binary_iarchive,std::basic_string<char,std::char_traits<char>,std::allocator<char> > >(boost::archive::binary_iarchive & ar={...}, std::basic_string<char,std::char_traits<char>,std::allocator<char> > & t={...}) Line 96 C++ cmnTest2.exe!boost::archive::detail::load_non_pointer_type<boost::archive::binary_iarchive,std::basic_string<char,std::char_traits<char>,std::allocator<char> > >::load_primitive::invoke(boost::archive::binary_iarchive & ar={...}, std::basic_string<char,std::char_traits<char>,std::allocator<char> > & t={...}) Line 308 + 0xd C++ cmnTest2.exe!boost::archive::detail::load_non_pointer_type<boost::archive::binary_iarchive,std::basic_string<char,std::char_traits<char>,std::allocator<char> > >::invoke(boost::archive::binary_iarchive & ar={...}, std::basic_string<char,std::char_traits<char>,std::allocator<char> > & t={...}) Line 366 + 0xd C++ cmnTest2.exe!boost::archive::load<boost::archive::binary_iarchive,std::basic_string<char,std::char_traits<char>,std::allocator<char> > >(boost::archive::binary_iarchive & ar={...}, std::basic_string<char,std::char_traits<char>,std::allocator<char> > & t={...}) Line 529 + 0xd C++ cmnTest2.exe!boost::archive::basic_binary_iarchive<boost::archive::binary_iarchive>::load_override<std::basic_string<char,std::char_traits<char>,std::allocator<char> > >() Line 66 + 0x15 C++ cmnTest2.exe!boost::archive::binary_iarchive_impl<boost::archive::binary_iarchive>::load_override<std::basic_string<char,std::char_traits<char>,std::allocator<char> > >() Line 46 C++ cmnTest2.exe!boost::archive::detail::interface_iarchive<boost::archive::binary_iarchive>::operator>><std::basic_string<char,std::char_traits<char>,std::allocator<char> > >(std::basic_string<char,std::char_traits<char>,std::allocator<char> > & t={...}) Line 84 C++ cmnTest2.exe!boost::archive::detail::common_iarchive<boost::archive::binary_iarchive>::init() Line 66 C++ cmnTest2.exe!boost::archive::binary_iarchive_impl<boost::archive::binary_iarchive>::binary_iarchive_impl<boost::archive::binary_iarchive>(std::basic_istream<char,std::char_traits<char> > & is={...}, unsigned int flags=0) Line 60 C++ cmnTest2.exe!boost::archive::binary_iarchive::binary_iarchive(std::basic_istream<char,std::char_traits<char> > & is={...}, unsigned int flags=0) Line 78 + 0x1e C++ cmnTest2.exe!OrderedCollection::GetOrderedCollection(ByteArray * pArray=0x0034c620) Line 111 + 0x14 C++ Thanks, Kauai --- kauai <kauai_diver(a)yahoo.com> wrote: > Date: Wed, 11 May 2005 10:15:34 -0700 (PDT) > From: kauai <kauai_diver(a)yahoo.com> > Subject: Re: [Boost-users] Re: Re: Re: Re: Boost > serializationanddependencynightmares... > To: boost-users(a)lists.boost.org > > > Hi Robert, Jeff, et all, > > Thank you Jeff, using that macro helped, I can now build > with the Microsoft compiler. Now getting some sort of > exception in my serialization code. > > I am looking into that today, have to setup my visual > studio and debugger :) And will get back to you all a > bit > later... > > Thanks, > K > > > --- Robert Ramey <ramey(a)rrsd.com> wrote: > > For what its worth, changes have been made for 1.33 > which > > work better with > > argument dependent lookup as well as two-phase lookup. > > > > I envisioned that the EXPORT marcros be in the class > > header after the class > > declaration. > > > > For export to work properly, the archive includes have > to > > before the class > > heades with EXPORT in them > > > > All serialization traites macros like EXPORT must be > > invoke from the global > > namespace. > > > > Robert Ramey > > > > > > Jeff Flinn wrote: > > > "kauai" <kauai_diver(a)yahoo.com> wrote in message > > > > > > news:20050510182406.59092.qmail@web60712.mail.yahoo.com... > > >> Hi Robert, > > >> > > >> The EXPORT macros were initially inside a namespace, > > >> interesting that it compiled and linked with gcc. > > >> > > >> After fixing this I still have the same problem with > > >> Microsofts compiler. It complains about the same > > thing as > > >> I state below. Is there some include file that I > may > > be > > >> missing? I can send you some code if that helps. > > > > > > I use: > > > > > > #define BOOST_NO_ARGUMENT_DEPENDENT_LOOKUP > > > > > > previous to including any boost serialization > headers. > > I'm not sure > > > it'll solve your issues, but it did with mine with > > VC7.1. > > > > > > Jeff Flinn > > > > > > > > _______________________________________________ > > Boost-users mailing list > > Boost-users(a)lists.boost.org > > http://lists.boost.org/mailman/listinfo.cgi/boost-users > > > > > > __________________________________ > Yahoo! Mail Mobile > Take Yahoo! Mail with you! Check email on your mobile > phone. > http://mobile.yahoo.com/learn/mail > Yahoo! Mail Stay connected, organized, and protected. Take the tour: http://tour.mail.yahoo.com/mailtour.html

1 0

Question
by Sliwa, Przemyslaw (London) 11 May '05

11 May '05

All, I have a question with regards more to the C++ than Boost. I would like to write a template which depending on the argument type does different things. For example the pseudo code: template <class T, class C> foo(T& nT, C& nC) { // common code; if(T==C) // specific code; ..... // again common code; } Can the template meta programming help me in this case? Sorry, I have no experice with this and I am just guessing. Do you have any suggestions? Thanks for help, Pshemek -------------------------------------------------------- If you are not an intended recipient of this e-mail, please notify the sender, delete it and do not read, act upon, print, disclose, copy, retain or redistribute it. Click here for important additional terms relating to this e-mail. http://www.ml.com/email_terms/ --------------------------------------------------------

2 1

Re: [Boost-users] Re: Re: Re: Re: Boost serializationanddependencynightmares...
by kauai 11 May '05

11 May '05

Hi Robert, Jeff, et all, Thank you Jeff, using that macro helped, I can now build with the Microsoft compiler. Now getting some sort of exception in my serialization code. I am looking into that today, have to setup my visual studio and debugger :) And will get back to you all a bit later... Thanks, K --- Robert Ramey <ramey(a)rrsd.com> wrote: > For what its worth, changes have been made for 1.33 which > work better with > argument dependent lookup as well as two-phase lookup. > > I envisioned that the EXPORT marcros be in the class > header after the class > declaration. > > For export to work properly, the archive includes have to > before the class > heades with EXPORT in them > > All serialization traites macros like EXPORT must be > invoke from the global > namespace. > > Robert Ramey > > > Jeff Flinn wrote: > > "kauai" <kauai_diver(a)yahoo.com> wrote in message > > > news:20050510182406.59092.qmail@web60712.mail.yahoo.com... > >> Hi Robert, > >> > >> The EXPORT macros were initially inside a namespace, > >> interesting that it compiled and linked with gcc. > >> > >> After fixing this I still have the same problem with > >> Microsofts compiler. It complains about the same > thing as > >> I state below. Is there some include file that I may > be > >> missing? I can send you some code if that helps. > > > > I use: > > > > #define BOOST_NO_ARGUMENT_DEPENDENT_LOOKUP > > > > previous to including any boost serialization headers. > I'm not sure > > it'll solve your issues, but it did with mine with > VC7.1. > > > > Jeff Flinn > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > __________________________________ Yahoo! Mail Mobile Take Yahoo! Mail with you! Check email on your mobile phone. http://mobile.yahoo.com/learn/mail

1 0

Boost Logging Library 1.3 troubles
by Ormond, Randy (Mission Systems) 11 May '05

11 May '05

Hi, If I'm following the threads correctly, it looks like John Torjo's logging library may become an official part of Boost someday. It certainly seems to be perfect for my needs. But I'm having all sorts of trouble on Red Hat with gcc 3.2.2-5. I had to modify some source to avoid compile errors, which has me worried given my lack of expertise. In log.hpp, I had to change (sorry, I can't paste in code or compiler output, this is typed in) template<bool is_compile_time, bool is_enabled> struct logger_keeper_is_enabled... to template<bool is_compile_time, bool is_enabledInput> struct logger_keeper_is_enabled... to avoid shadowing is_enabled in ts_posix.hpp, I had to change the mutex ctr from mutex() : m_mutex(0) {... to mutex() {... This seems OK because m_mutex is initialized in the ctr body. After making those changes, I tried the example in the basic_usage.html. It hangs on the first call to BOOST_DEFINE_LOG. I traced it with gdb to find_log_by_name() at log_manager.cpp:294. From there, it goes to the logger_info ctr. It hangs in lwm_pthreads.hpp:72 in the scoped_lock ctr, apparently trying to get a lock. Of course my mods to the source and general lack of insight have me worried. I expect difficulties as an "early adopter" but any help would be greatly appreciated. BTW, is anyone else trying to use this yet? Thanks, Randy

1 0

RE: [Boost-users] Re: Serialize templates classes > automatique exportdeclaration
by Olivier Vanden Bos 11 May '05

11 May '05

I am not sure of my fact, but can we do it dynamically , as the macro call extended_type_info::self_register() and void extended_type_info::key_register(const char *key_). I ve read in the list some stuff about registering templates like this... template<class> struct pack; template<class T> struct pack<void (T)> { typedef T type; }; BOOST_CLASS_EXPORT( pack<void (DerivedStatus< Algo<int, double > >)>::type ); Can we mix that together to achieve this goal ? In fact, I don't care if we register the type each time we call, I know it s not optimal , but at this point I can live with it... void Add() { B<int>* objectToAdd= new B<int>(0) BoostRegister(B<int>); array.push_back(objectToAdd); } - VDB - "LNK1210: exceeded internal ILK size limit" ________________________________ From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Robert Ramey Sent: 11 mai 2005 02:22 To: boost-users(a)lists.boost.org Subject: [Boost-users] Re: Serialize templates classes > automatique exportdeclaration I don't know how to avoid this either. If you figure out how to do this let me know Robert Ramey And when i archive the array in another .cpp the exception is raised, because the class B<int> as not been exported. I would like to avoid explicitly declare BOOST_CLASS_EXPORT_GUID(B<int>, "B_int") For each instanciation of B<> somewhere in the glo bal space.

2 1

FOREACH with maps
by Dave Steffen 11 May '05

11 May '05

Another question about the to-be-released FOREACH macro, which I find I am using heavily... (BTW - Thanks Eric for your quick responses, and for the cool macro!) The issue is declaring the 'loop variable' for maps. For example, vector<int> v; ... FOREACH (int a, v) { ... } // this compiles int b; FOREACH (b, v) {...} // this also compiles However, for maps, the second variation (declaring outside the macro) doesn't work: map<int,int> m; ... FOREACH (map<int,int>::value_type i, m) {...} // compiles map<int,int>::value_type i; FOREACH (i, m) {...} // COMPILE ERROR Specifically, the compiler says error: non-static const member ` const int std::pair<const int,int>::first', can't use default assignment operator (from GCC 3.3.1). Despite my best efforts, I can't figure out what this error means, or what it's talking about. Anybody else run into this? Any thoughts? ---------------------------------------------------------------------- Dave Steffen, Ph.D. "The only justification for our concepts and Software Engineer IV systems of concepts is that they serve to Numerica Corporation represent the complex of our experiences; ph (970) 419-8343 x27 beyond this they have no legitimacy." fax (970) 223-6797 -- Albert Einstein dgsteffen(a)numerica.us

7 9

Unable to create libboost-thread-vc71-mt-sgd-1_31.lib
by Shetty, Diwakar 11 May '05

11 May '05

Hi all, We included the header files of boost in VC++ 7.1 During linking the program automatically tries to link against the static library libboost-thread-vc71-mt-sgd-1_31.lib. However, "bjam install" does not create this library. Though it creates the dll version of the thread library (i.e. thread*.dll and thread*.lib) We tried various combinations such as following to get the static library built. 1) bjam --with-thread -sBUILD="debug <runtime-link>static <threading>single" install 2) bjam --with-thread -sBUILD="debug static single" install 3) bjam -sBUILD="debug static single" install But none of the above worked. We also tried to enable "BOOST_ALL_NO_LIB" macro so that we can specify the dynamic version of the library manually. It did not work. We are using the dynamic version of runtime library. The question I have is, how to build the library "libboost-thread-vc71-mt-sgd-1_31.lib" Any help would be highly appreciated. I have tried every possible thing and seem to be going nowhere with this issue. Thanks, Diwakar

2 1

Fwd: Re: [Boost-users] Re: Boost serialization and dependency nightmares...
by kauai 11 May '05

11 May '05

In addition I resurrected this post to the newsgroup. It is exactly the same problem I am having: boost 1.32, with VC7.1 I'm getting compiler errors in two unrelated translation units both dealing with std::set<std::string> data members. I've gone through and ensured that i've #include'd the <set.hpp> and <string.hpp> headers from boost/serialization. Any ideas on what might be causing this problem? This has worked fine previously with boost 1.31 and serialization20 version of the library. Why would the 3 argument overload not be found? Below are the (beginnings of the ) related compiler errors. Thanks, Jeff ======================== C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(55) : error C2780: 'void boost::archive::load(Archive &,T &)' : expects 2 arguments - 3 provided C:\boost\boost_1_32_0\boost\archive\detail\iserializer.hpp(514) : see declaration of 'boost::archive::load' C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(54) : while compiling class-template member function 'void boost::serialization::free_loader<Archive,T>::invoke(Archive &,T &,const unsigned int)' with [ Archive=boost::archive::text_iarchive, T=std::set<std::string> ] .... ======================== C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(55) : error C2780: 'void boost::archive::load(Archive &,T &)' : expects 2 arguments - 3 provided C:\boost\boost_1_32_0\boost\archive\detail\iserializer.hpp(514) : see declaration of 'boost::archive::load' C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(54) : while compiling class-template member function 'void boost::serialization::free_loader<Archive,T>::invoke(Archive &,T &,const unsigned int)' with [ Archive=boost::archive::binary_iarchive, T=std::set<std::string> ] .... --- kauai <kauai_diver(a)yahoo.com> wrote: > Date: Mon, 9 May 2005 14:07:43 -0700 (PDT) > From: kauai <kauai_diver(a)yahoo.com> > Subject: Re: [Boost-users] Re: Boost serialization and > dependency nightmares... > To: boost-users(a)lists.boost.org > > Robert, > > Thank you. My project requires both linux and Win32 > support. When I use this macro in win32 environment I > get > a compile time error. It generates a fairly long error > which I can send you but here is the first few lines > which > are possibly the most important. > > Again it is fully compiling/linking/running in Linux. > > > cl /nologo /W3 /MDd /Od /EHsc /RTC1 /GS /Gy /Zc:forScope > /GR /Gd /TP /D_DEBUG /D > WIN32 /D_CONSOLE /D_WINDOWS /DUNICODE /DQT_DLL > /DQT_THREAD_SUPPORT > /IW:\vob\abvsw\boost\1_32_0\win32\include\boost-1_32 > /Iroot\abcmn\ /c build-win32-debug\...\xyz.cpp /Fobui > ld-win32-debug\...\xyz.obj xyz.cpp > W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_32\boost\ > serialization\split_free.hpp(55) : error C2780: 'void > boost::archive::load(Archi > ve &,T &)' : expects 2 arguments - 3 provided > W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_3 > 2\boost\archive\detail\iserializer.hpp(514) : see > declaration of 'boost::archive > ::load' > W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_3 > 2\boost\serialization\split_free.hpp(54) : while > compiling > class-template member > function 'void > boost::serialization::free_loader<Archive,T>::invoke(Archive > &,T > &,const unsigned int)' > > When I #ifdef the BOOST_CLASS_EXPORT(ClassName) macros I > compile/link/and run. Of course base class serialzation > is > disable for Win32 at the moment because of this. > > Thanks for any help, > -K > > --- Robert Ramey <ramey(a)rrsd.com> wrote: > > This is addressed with the alternative > BOOST_CLASS_EXPORT > > method. its > > described in the serialization documentation. > > > > Robert Ramey > > > > kauai wrote: > > > Hi all, > > > > > > I have a container class that has a vector which > > contains > > > several other classes. My container class has a > > serialize > > > function that looks like this: > > > > > > friend class boost::serialization::access; > > > template<class Archive> void serialize (Archive &ar, > > const > > > unsigned int version) > > > { > > > // save/load base class information > > > ar & > > BOOST_SERIALIZATION_BASE_OBJECT_NVP(MyBaseClass); > > > ar.register_type(static_cast<SomeClass1*>(NULL)); > > > ar.register_type(static_cast<SomeClass2*>(NULL)); > > > ... > > > ar & BOOST_SERIALIZATION_NVP(m_Vector); // > > std::vector > > > } > > > > > > Got everything working but now an issue has come up: > > > Someone wants to add a new class but it is not part > of > > this > > > package that is doing the serialzation. I want to > > avoid > > > what I call "dependency hell", is there a way for > that > > new > > > class to get registered someplace other than in this > > > function? > > > > > > Thanks for any help, > > > Kauai > > > > > > > > > > > > > > > Discover Yahoo! > > > Stay in touch with email, IM, photo sharing and more. > > Check it out! > > > http://discover.yahoo.com/stayintouch.html > > > > > > > > _______________________________________________ > > Boost-users mailing list > > Boost-users(a)lists.boost.org > > http://lists.boost.org/mailman/listinfo.cgi/boost-users > > > > > > > __________________________________ > Do you Yahoo!? > Yahoo! Mail - You care about security. So do we. > http://promotions.yahoo.com/new_mail > Discover Yahoo! Have fun online with music videos, cool games, IM and more. Check it out! http://discover.yahoo.com/online.html

3 4

Serialize templates classes > automatique export declaration
by Olivier Vanden Bos 11 May '05

11 May '05

I think there have been a lot of discussion about this, It might been related to the template serialization traits in the documentation , but the explanation was a little to subtle for me, \ In fact, I cannot really understand how to avoid The unregistered exception when I use a template like this. // template declaration pseudo code Class A { Virtual ~A(){}; template<class Archive> void serialize(Archive & ar, const unsigned int version){}; }; Template <T> Class B : public A { T Object; B(T& _object) { Object=_object; } template<class Archive> void serialize(Archive & ar, const unsigned int version) { Ar & Object; }; } //Main Code Vector<A*> array; Template<T> void Add(T& val) { B<T>* objectToAdd= new B<T>(val) array.push_back(objectToAdd); } //User code Add((int)0); And when i archive the array in another .cpp the exception is raised, because the class B<int> as not been exported. I would like to avoid explicitly declare BOOST_CLASS_EXPORT_GUID(B<int>, "B_int") For each instanciation of B<> somewhere in the global space. In the Add() or in the template declaration that s ok, but I cannot suppose what types that the user will use. Thanks in advance.. - VDB - "LNK2019: unresolved external symbol" Olivier Vanden Bos MSN#TheVdeub(a)hotmail.com AI Programmer UBISOFT Montreal

2 1

Fwd: Re: [Boost-users] Re: Re: Re: Boost serialization and dependencynightmares...
by kauai 11 May '05

11 May '05

Hi Robert, I've been looking into this further and found some interesting behavior. Don't know if this will help or not, but here goes: * In linux I have my xyz.cpp file, several includes, the archive include, then the export, like this: #include <boost/serialization/export.hpp> // should always be after archive includes After that I have: BOOST_CLASS_EXPORT(Foo::Bar::MyClass); using namespace Foo::Bar; // the rest of the file. This code plus the rest will compile, link, and run properly in Linux using gcc 3.2.2. (I can get more version info if you need, kernel, glibc, etc.) * In Windows Server 20003 with cl version 13.10.3077 I can not compile. I get the same error message I stated earlier. Now the wierd part: If I remove the #include <boost/serialization/export.hpp> I cannot compile anymore with gcc, as expected. There error I get is this: ISO C++ forbids declaration of `BOOST_CLASS_EXPORT' with no type In Windows not only does it compile, but it links and runs. 'runs' meaning there is an .exe file but my serialization test code still fails ;) I dont know if these annoying details may help but I just thought I'd share. -K --- kauai <kauai_diver(a)yahoo.com> wrote: > Date: Tue, 10 May 2005 11:24:06 -0700 (PDT) > From: kauai <kauai_diver(a)yahoo.com> > Subject: Re: [Boost-users] Re: Re: Re: Boost > serialization and dependencynightmares... > To: boost-users(a)lists.boost.org > > Hi Robert, > > The EXPORT macros were initially inside a namespace, > interesting that it compiled and linked with gcc. > > After fixing this I still have the same problem with > Microsofts compiler. It complains about the same thing > as > I state below. Is there some include file that I may be > missing? I can send you some code if that helps. > > Also where is the best place to call these macros? At > the > moment they are at the top in the appropriate .cpp files. > > Thanks, > -K > > > --- Robert Ramey <ramey(a)rrsd.com> wrote: > > > Just a guess, but are the EXPORT macros being used in > > side a namespace? > > They should be outside of any namespace. > > > > Robert Ramey > > > > kauai wrote: > > > In addition I resurrected this post to the newsgroup. > > > It > > > is exactly the same problem I am having: > > > > > > boost 1.32, with VC7.1 I'm getting compiler errors in > > two > > > unrelated translation units both dealing with > > > std::set<std::string> data members. I've gone through > > and > > > ensured that i've #include'd the <set.hpp> and > > <string.hpp> > > > headers from boost/serialization. Any ideas on what > > might > > > be causing this problem? This has worked fine > > previously > > > with boost 1.31 and serialization20 version of the > > library. > > > Why would the 3 argument overload not be found? > > > > > > Below are the (beginnings of the ) related compiler > > errors. > > > > > > Thanks, Jeff > > > > > > ======================== > > > > > > > > > C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(55) > > >> error C2780: 'void boost::archive::load(Archive &,T > > &)' : > > > expects 2 arguments - 3 provided > > > > > > > > > C:\boost\boost_1_32_0\boost\archive\detail\iserializer.hpp(514) > > >> see declaration of 'boost::archive::load' > > > > > > > > > C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(54) > > >> while compiling class-template member function 'void > > > > > > boost::serialization::free_loader<Archive,T>::invoke(Archive > > > &,T &,const unsigned int)' > > > with > > > [ > > > Archive=boost::archive::text_iarchive, > > > T=std::set<std::string> > > > ] > > > .... > > > > > > ======================== > > > > > > > > > C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(55) > > >> error C2780: 'void boost::archive::load(Archive &,T > > &)' : > > > expects 2 arguments - 3 provided > > > > > > > > > C:\boost\boost_1_32_0\boost\archive\detail\iserializer.hpp(514) > > >> see declaration of 'boost::archive::load' > > > > > > > > > C:\boost\boost_1_32_0\boost\serialization\split_free.hpp(54) > > >> while compiling class-template member function 'void > > > > > > boost::serialization::free_loader<Archive,T>::invoke(Archive > > > &,T &,const unsigned int)' > > > with > > > [ > > > Archive=boost::archive::binary_iarchive, > > > T=std::set<std::string> > > > ] > > > > > > .... > > > > > > > > > --- kauai <kauai_diver(a)yahoo.com> wrote: > > >> Date: Mon, 9 May 2005 14:07:43 -0700 (PDT) > > >> From: kauai <kauai_diver(a)yahoo.com> > > >> Subject: Re: [Boost-users] Re: Boost serialization > and > > >> dependency nightmares... > > >> To: boost-users(a)lists.boost.org > > >> > > >> Robert, > > >> > > >> Thank you. My project requires both linux and Win32 > > >> support. When I use this macro in win32 environment > I > > >> get > > >> a compile time error. It generates a fairly long > > error > > >> which I can send you but here is the first few lines > > >> which > > >> are possibly the most important. > > >> > > >> Again it is fully compiling/linking/running in > Linux. > > >> > > >> > > >> cl /nologo /W3 /MDd /Od /EHsc /RTC1 /GS /Gy > > /Zc:forScope > > >> /GR /Gd /TP /D_DEBUG /D > > >> WIN32 /D_CONSOLE /D_WINDOWS /DUNICODE /DQT_DLL > > >> /DQT_THREAD_SUPPORT > > >> /IW:\vob\abvsw\boost\1_32_0\win32\include\boost-1_32 > > >> /Iroot\abcmn\ /c build-win32-debug\...\xyz.cpp > /Fobui > > >> ld-win32-debug\...\xyz.obj xyz.cpp > > >> > > > W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_32\boost\ > > >> serialization\split_free.hpp(55) : error C2780: > 'void > > >> boost::archive::load(Archi > > >> ve &,T &)' : expects 2 arguments - 3 provided > > >> > > W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_3 > > >> 2\boost\archive\detail\iserializer.hpp(514) : see > > >> declaration of 'boost::archive > > >>>> load' > > >> > > W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_3 > > >> 2\boost\serialization\split_free.hpp(54) : while > > >> compiling > > >> class-template member > > >> function 'void > > >> > > > > > > boost::serialization::free_loader<Archive,T>::invoke(Archive > > >> &,T > > >> &,const unsigned int)' > > >> > > >> When I #ifdef the BOOST_CLASS_EXPORT(ClassName) > macros > > I > > >> compile/link/and run. Of course base class > > serialzation > > >> is > > >> disable for Win32 at the moment because of this. > > >> > > >> Thanks for any help, > > >> -K > > >> > > >> --- Robert Ramey <ramey(a)rrsd.com> wrote: > > >>> This is addressed with the alternative > > >> BOOST_CLASS_EXPORT > > >>> method. its > > >>> described in the serialization documentation. > > >>> > > >>> Robert Ramey > > >>> > > >>> kauai wrote: > > >>>> Hi all, > > >>>> > > >>>> I have a container class that has a vector which > > >>> contains > > >>>> several other classes. My container class has a > > >>> serialize > > >>>> function that looks like this: > > >>>> > > >>>> friend class boost::serialization::access; > > >>>> template<class Archive> void serialize (Archive > &ar, > > >>> const > > >>>> unsigned int version) > > >>>> { > > >>>> // save/load base class information > > >>>> ar & > > >>> BOOST_SERIALIZATION_BASE_OBJECT_NVP(MyBaseClass); > > >>>> > ar.register_type(static_cast<SomeClass1*>(NULL)); > > >>>> > ar.register_type(static_cast<SomeClass2*>(NULL)); > > >>>> ... > > >>>> ar & BOOST_SERIALIZATION_NVP(m_Vector); // > > >>> std::vector > > >>>> } > > >>>> > > >>>> Got everything working but now an issue has come > up: > > >>>> Someone wants to add a new class but it is not > part > > of this > > >>>> package that is doing the serialzation. I want to > > >>> avoid > > >>>> what I call "dependency hell", is there a way for > > that new > > >>>> class to get registered someplace other than in > this > > >>>> function? > > >>>> > > >>>> Thanks for any help, > > >>>> Kauai > > >>>> > > >>>> > > >>>> > > >>>> > > >>>> Discover Yahoo! > > >>>> Stay in touch with email, IM, photo sharing and > > more. > > >>> Check it out! > > >>>> http://discover.yahoo.com/stayintouch.html > > >>> > > >>> > > >>> > > >>> _______________________________________________ > > > === message truncated === > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection > around > http://mail.yahoo.com > Discover Yahoo! Find restaurants, movies, travel and more fun for the weekend. Check it out! http://discover.yahoo.com/weekend.html

2 1

creating static functions with metaprogramming
by Felipe Magno de Almeida 11 May '05

11 May '05

I would like to know if it is possible to generate a number of static functions that redirect to some other function inside the class, taking an object as void * and is casted to the object (that it knows which is), and be possible to take the address of such functions... well, i dont have a good english, so I will reexplain: For the ones that already are familiar with Win32 GUI: I must create a static function to be the DialogProc of a window, the problem is... I'm creating some property sheets to be used with one class, and I would like to make it automatic the redirection from the static function to some other method inside the class, that isnt static... and I would like to create a class, probably with CRTP to say to it to create the functions that would redirect to me, saying how much functions would that be, and make it creates the codes that gives to windows the static functions addresses and create the functions that would redirect.... Then I just needed to create my functions that they redirect to me... making it a LOT easier to me, and would make my code a lot more concise... Sorry for the off-topic... But I think you would be the best ones to direct this question.... Thanks for the attention, Felipe Magno de Almeida -- Felipe Magno de Almeida UIN: 2113442 email: felipe.m.almeida+spam at gmail com felipe.almeida at ic unicamp br, felipe.m.almeida at gmail com, felipe at synergy com I am a C, modern C++, MFC, ODBC, Windows Services, MAPI developer from synergy, and Computer Science student from State University of Campinas(UNICAMP). To know more about: Unicamp: http://www.ic.unicamp.br Synergy: http://www.synergy.com.br current work: http://www.mintercept.com "There is no dark side of the moon really. Matter of fact it's all dark."

2 2

regex lib - using replace
by Lucian Ion 10 May '05

10 May '05

Hi, I need one regular expression and one format string to do the following 1) if text starts with s1, replace s1 with r1 2) else, if text starts with s2, replace s2 with r2 3) else, if text starts with s3, replace s2 with r3 4) else, prepend r3 to text While I managed to work out the first three using an expression like (^s1)|(^s2)|(^s3) and the format as 1?r1:2?r2:3?r3: I have no idea how to include the fourth case into my expression. Any suggestion? Thank you in advance, Lucian

2 1

Re: regex lib - using replace
by Lucian Ion 10 May '05

10 May '05

sorry, I mean 4) else, prepend r4 to text (that is, has nothing to do with the other replacement strings) ----- Original Message ----- From: "Lucian Ion" <lucianion(a)rapidsolution.ro> To: <boost-users(a)lists.boost.org> Sent: Tuesday, May 10, 2005 8:50 PM Subject: regex lib - using replace > Hi, > I need one regular expression and one format string to do the following > 1) if text starts with s1, replace s1 with r1 > 2) else, if text starts with s2, replace s2 with r2 > 3) else, if text starts with s3, replace s2 with r3 > 4) else, prepend r3 to text > > While I managed to work out the first three using an expression like > (^s1)|(^s2)|(^s3) and the format as 1?r1:2?r2:3?r3: > I have no idea how to include the fourth case into my expression. Any > suggestion? > > Thank you in advance, > Lucian >

1 0

Boos.Bind: Binding to a placeholder address
by Adi Shavit 10 May '05

10 May '05

Hi, Say have function I want to bind that expects an int*, but my placeholder _1 is of type int. How can I tell the bind function to bind "&_1"? I guess it would be something like: bind(myFun, address_of(_1)); Is this even possible? Thanks, Adi

3 3

Standard Library Extension TR2 Call for Proposals
by Beman Dawes 10 May '05

10 May '05

The C++ Standards Committee has begun work on a second library technical report. A Call for Proposals is available at http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2005/n1810.html --Beman

1 0

Building Boost on MacOS 10.4 (Tiger)
by Dave Snowdon 10 May '05

10 May '05

Hello Has anyone successfully built boost on an Mac running MacOS 10.4? If so I'd appreciate some pointers on what worked for you. I've tried both boost 1.32 and a "fresh from CVS" copy with gcc 3.3 and gcc 4.0 on a Powerbook G4 with 512Mb RAM with a fresh install of MacOS 10.4. The Mac has 1.15Gb free on the startup partition and 1.36Gb free on the partition where I'm compiling boost. So far no attempt has met with success. All builds were attempted from the Terminal app with no other applications runnng. Compiling boost 1.32 with gcc 4.0 (the default compiler on MacOS 10.4) fails (unrecognised compiler version) because at the time 1.32 was released gcc 4.0 was not available. Compiling boost 1.32 with gcc 3.3 (bjam "-sTOOLS=darwin" "-sGCC=gcc-3.3" "-sGXX=gcc-3.3" stage) gives some rather strange results. Link actions seems to fail such as: ...failed darwin-Link-DyLib-action bin/boost/libs/date_time/build/libboost_date_time.dylib/darwin/debug/shared-linkable-true/libboost_date_time-d-1_32.dylib... Almost all the calls to ld gives undefined symbol errors similar to (I'm not sure if this is due to previously failed link actions): darwin-Link-DyLib-action bin/boost/libs/date_time/build/libboost_date_time.dylib/darwin/debug/shared-linkable-true/libboost_date_time-d-1_32.dylib ld: Undefined symbols: __ZNKSbIwSt11char_traitsIwESaIwEE3endEv __ZNKSbIwSt11char_traitsIwESaIwEE5beginEv __ZNKSs2atEm __ZNKSs3endEv __ZNKSs5beginEv __ZNKSs6lengthEv __ZNKSs7compareERKSs __ZNKSt11logic_error4whatEv __ZNKSt18basic_stringstreamIcSt11char_traitsIcESaIcEE3strEv __ZNKSt18basic_stringstreamIwSt11char_traitsIwESaIwEE3strEv __ZNKSt9basic_iosIcSt11char_traitsIcEE4failEv __ZNKSt9basic_iosIwSt11char_traitsIwEE4failEv __ZNKSt9type_infoeqERKS_ __ZNSaIcEC1Ev __ZNSaIcED1Ev After about 15 minutes or so the Mac grinds to a halt and starts popping up dialog boxes warning that the startup disk is full and then later that "Your Mac OS X startup disk has no more space available for application memory. To avoid problems with your computer quit any applications you may not be using" Using gcc 4.0 (bjam "-sTOOLS=darwin" stage) on the the CVS copy gives many errors similar to this: ...failed darwin-C++-action bin/boost/libs/wave/build/libboost_wave.a/darwin/release/cpp_re.o... ...skipped <@boost!libs!wave!build/libboost_wave.a/darwin/release>libboost_wave-1_33.a for lack of <@boost!libs!wave!build/libboost_wave.a/darwin/release>instantiate_cpp_exprgrammar.o... ...skipped <@boost!lib>libboost_wave-1_33.a for lack of <@boost!libs!wave!build/libboost_wave.a/darwin/release>libboost_wave-1_33.a... Previously I've been using boost 1.32 on a machine running Red Hat Linux 8 with gcc 3.2 and have not experienced any problems getting it to build. Any advice appreciated.... thanks Dave

2 1

Need Help
by pankaj jain 10 May '05

10 May '05

Hi , I am new user of boost.python. Please help me in solving this problem. problem 1) When I am running bjam from home/boost/libs/python/example/tutorial folder its running fine but building my hello.so in home/boost/bin/boost/libs/python/example/tutorial/hello.so/gcc/.. folder.is there any way to chagne the path for hello.so problem 2): when I am running bjam from any other folder ( home/boost/sample) its giving me this error :: /home/jamrules no such directory .../.../allyourbase.jam 1451 see definition of rules "root-path: being called. Please let me know any solution. Thanks for all your help Pankaj Joaqu�n M� L�pez Mu�oz <joaquin(a)tid.es> wrote: Dave Snowdon ha escrito: > I'd like to second what Stefan said. Before I started using boost (only > a couple of months back) I'd seen references to it in many articles in > the C++ User Journal not to mention a few C++ related books. Maybe that's a significant difference wrt to the rest of the C++ community: Many programmers are not subscribed to CUJ and do not buy intermediate/advanced books (entry-level references do not usually list Boost.) I'd like a website that holds the attention of the casual comer. A newbie with just a passing knowledge of what Boost is might be scared by the contents of boost.org and never return back; a more friendly site, with forums and stuff, is a more agreeable place to hang around and eventually decide to try Boost for real. Joaqu�n M L�pez Mu�oz Telef�nica, Investigaci�n y Desarrollo > After > starting on a new C++ project I realised I needed some of the libraries > mentioned and went straight for them. So it was not so much a case of > "do I need boost?" but "where do I find the bits I need now, and how do > I use them?" > > Dave > > Stefan Strasser wrote: > > > I don't think people come by accident to the boost page(and they > > definitely don't to the community page) and need to be convinced that > > they need boost. > > they look for a specific library/class and they've been told, "boost > > has that". later on they look what also is in boost. > > at least that is how it worked for me. > > > > so a community page(especially user comments in documentation) helps a > > new user to get into it. it is imho not the function of a community > > page to raise interest. (and I think it can't). > > > > and I think the best thing you can do to bring a new user into the > > community thing is to have documentation and user comments on the same > > page(see php). because what a user definitely does is look at the > > documentation when he's found the library he's looking for. > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users --------------------------------- Yahoo! Mail Stay connected, organized, and protected. Take the tour

2 1

order of pointer serialisation
by martin f krafft 10 May '05

10 May '05

Consider a doubly-linked list. How do you serialise this? The natural way would be to provide a serialize function such as: template <class Archive> void serialize(Archive& ar, unsigned int const) { ar & prev; ar & next; } This, however, causes "pointer conflict" exception to be thrown, because the next pointers are serialised before the actual list node is serialised. The following is an abridged scneario illustrating the same dilemma: class Foo { friend class boost::serialization::access; template <typename AT> void serialize(AT& ar, unsigned int const) {} }; int main(void) { Foo x; Foo *y = &x; boost::archive::text_oarchive oa(std::cout); oa & y; // serialise the pointer before the actual instance oa & x; // throws boost::archive::archive_exception(pointer_conflict) return 0; } I could come up with many more examples in which it is now granted that an instance has been serialised by the time that a pointer to that very instance needs to be serialised. How does boost::serialization expect this challenge to be handled? Note that I understand very well what pointer_conflict is all about. I do wonder whether this protection is always necessary. Worse yet: there seems to be no way to disable it; catching the pointer_conflict skips over the serialisation of the instance. Is the only way to deal with this to ensure that instances are always saved first? -- martin; (greetings from the heart of the sun.) \____ echo mailto: !#^."<*>"|tr "<*> mailto:" net@madduck invalid/expired pgp subkeys? use subkeys.pgp.net as keyserver! spamtraps: madduck.bogus(a)madduck.net micro$oft encrypts your windows nt password when stored on a windows ce device. but if you look carefully at their encryption algorithm, they simply xor the password with "susageP", Pegasus spelled backwards. Pegasus is the code name of windows ce. this is so pathetic it's staggering.

1 0

uBLAS with MS vc 7.1
by Tux Han 10 May '05

10 May '05

I wrote a simple program using uBLAS under vc 7.1, but got so many warnings, how can I correct it? The code is: ---------------------- #include <boost/numeric/ublas/matrix.hpp> namespace ublas = boost::numeric::ublas; int main(int argc, _TCHAR* argv[]) { ublas::identity_matrix<double> A(4); ublas::identity_matrix<double> B(4); prod(A, B); return 0; } ------------------ But I got the following warning: d:\Project\Libraries\boost_1_32_0\boost\numeric\ublas\traits.hpp(941) : warning C4267: 'argument' : conversion from 'size_t' to 'const unsigned int', possible loss of data d:\Project\Libraries\boost_1_32_0\boost\numeric\ublas\matrix_expression.hpp(4462) : see reference to class template instantiation 'boost::numeric::ublas::promote_traits<X,Y>' being compiled with [ X=boost::numeric::ublas::identity_matrix<double>::size_type, Y=boost::numeric::ublas::identity_matrix<double>::size_type ] d:\test\boosttest\boosttest\boosttest.cpp(14) : see reference to class template instantiation 'boost::numeric::ublas::matrix_matrix_binary<E1,E2,F>' being compiled with [ E1=boost::numeric::ublas::identity_matrix<double>, E2=boost::numeric::ublas::identity_matrix<double>, F=boost::numeric::ublas::matrix_matrix_prod<double,double,boost::numeric::ublas::matrix_matrix_binary_traits<double,boost::numeric::ublas::identity_matrix<double>,double,boost::numeric::ublas::identity_matrix<double>>::promote_type> ] d:\Project\Libraries\boost_1_32_0\boost\numeric\ublas\traits.hpp(941) : warning C4244: 'argument' : conversion from '__w64 int' to 'const int', possible loss of data d:\Project\Libraries\boost_1_32_0\boost\numeric\ublas\matrix_expression.hpp(4463) : see reference to class template instantiation 'boost::numeric::ublas::promote_traits<X,Y>' being compiled with [ X=boost::numeric::ublas::identity_matrix<double>::difference_type, Y=boost::numeric::ublas::identity_matrix<double>::difference_type ] Linking... How to correct it? thank you very much:) __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

1 0

Re: Building Boost on MacOS 10.4 (Tiger)
by Jean-François Brouillet 10 May '05

10 May '05

On 9 May 2005, at 18:08, boost-users-request(a)lists.boost.org wrote: If you don't mind modifying boost/config/compiler/gcc.hpp ... Here's how I patched it: (very last lines in the file) // last known and checked version is 4.0: #if (__GNUC__ > 4) || ((__GNUC__ == 3) && (__GNUC_MINOR__ > 4)) # if defined(BOOST_ASSERT_CONFIG) # error "Unknown compiler version - please run the configure tests and report the results" # else # warning "Unknown compiler version - please run the configure tests and report the results" # endif #endif BTW: Although my build half-succeeded (or half failed depending on your point of view :-) I'm more and more of the opinion that all this bjam rigmarole is strictly *not* needed! When I got fed up with the miles and miles of warnings/errors, I just decided to drop the "boost" subdir into Xcode and use it "as is". Since boost is supposed to be a 100% header file solution, the rationale for "building it" somehow eludes me ... Anyway, I'm glad to report that I'm up and running with boost 1.32 on OS X 10.4 Tiger with Xcode 2.0 (next step: do the same for CW 9.5 ... :) -- Jean-François Brouillet verec(a)mac.com

2 1

[BGL] Iterate through neighbour vertices
by Rui Carvalho 09 May '05

09 May '05

Hi, I have a directed graph (vecS, vecS, bidirectionalS) and I'm trying to iterate through the neighbour vertices of a vertex u. When using for (tie(ai0,ai0_end) = adjacent_vertices(u,g); ai0 != ai0_end; ++ai0) the code goes only through the out vertices of u. Is there a way to iterate through all the neighbouring vertices in one loop (i.e. through both the out and in vertices of u)? Thanks, Rui

2 3

Re: [Boost-users] Re: Boost serialization and dependency nightmares...
by kauai 09 May '05

09 May '05

Robert, Thank you. My project requires both linux and Win32 support. When I use this macro in win32 environment I get a compile time error. It generates a fairly long error which I can send you but here is the first few lines which are possibly the most important. Again it is fully compiling/linking/running in Linux. cl /nologo /W3 /MDd /Od /EHsc /RTC1 /GS /Gy /Zc:forScope /GR /Gd /TP /D_DEBUG /D WIN32 /D_CONSOLE /D_WINDOWS /DUNICODE /DQT_DLL /DQT_THREAD_SUPPORT /IW:\vob\abvsw\boost\1_32_0\win32\include\boost-1_32 /Iroot\abcmn\ /c build-win32-debug\...\xyz.cpp /Fobui ld-win32-debug\...\xyz.obj xyz.cpp W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_32\boost\ serialization\split_free.hpp(55) : error C2780: 'void boost::archive::load(Archi ve &,T &)' : expects 2 arguments - 3 provided W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_3 2\boost\archive\detail\iserializer.hpp(514) : see declaration of 'boost::archive ::load' W:\vob\abvsw\boost\1_32_0\win32\include\boost-1_3 2\boost\serialization\split_free.hpp(54) : while compiling class-template member function 'void boost::serialization::free_loader<Archive,T>::invoke(Archive &,T &,const unsigned int)' When I #ifdef the BOOST_CLASS_EXPORT(ClassName) macros I compile/link/and run. Of course base class serialzation is disable for Win32 at the moment because of this. Thanks for any help, -K --- Robert Ramey <ramey(a)rrsd.com> wrote: > This is addressed with the alternative BOOST_CLASS_EXPORT > method. its > described in the serialization documentation. > > Robert Ramey > > kauai wrote: > > Hi all, > > > > I have a container class that has a vector which > contains > > several other classes. My container class has a > serialize > > function that looks like this: > > > > friend class boost::serialization::access; > > template<class Archive> void serialize (Archive &ar, > const > > unsigned int version) > > { > > // save/load base class information > > ar & > BOOST_SERIALIZATION_BASE_OBJECT_NVP(MyBaseClass); > > ar.register_type(static_cast<SomeClass1*>(NULL)); > > ar.register_type(static_cast<SomeClass2*>(NULL)); > > ... > > ar & BOOST_SERIALIZATION_NVP(m_Vector); // > std::vector > > } > > > > Got everything working but now an issue has come up: > > Someone wants to add a new class but it is not part of > this > > package that is doing the serialzation. I want to > avoid > > what I call "dependency hell", is there a way for that > new > > class to get registered someplace other than in this > > function? > > > > Thanks for any help, > > Kauai > > > > > > > > > > Discover Yahoo! > > Stay in touch with email, IM, photo sharing and more. > Check it out! > > http://discover.yahoo.com/stayintouch.html > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > __________________________________ Do you Yahoo!? Yahoo! Mail - You care about security. So do we. http://promotions.yahoo.com/new_mail

1 0

Re: [Boost-users] Re: Re: The Boost Community
by JOAQUIN LOPEZ MU?Z 09 May '05

09 May '05

----- Mensaje original ----- De: Andy Little <andy(a)servocomm.freeserve.co.uk> Fecha: Lunes, Mayo 9, 2005 8:20 pm Asunto: [Boost-users] Re: Re: The Boost Community > > "Joaquín Mª López Muñoz" <joaquin(a)tid.es> wrote > > I'd like a website that holds the attention of the casual comer. A > newbiewith just a passing knowledge of what Boost is might be > scared by > the contents of boost.org and never return back; a more friendly site, > with forums and stuff, is a more agreeable place to hang around and > eventually decide to try Boost for real. > > that has to be http://bike-shed.boost.org :-) > Well, I'm sorry this is perceived as bike shedding :( To be more specific, my wishes boil down to having a site with forums, anonymous posting to the forums and some easy way for users (not necessarily Boost authors) to upload their contributions in the form of Boost articles. Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 09 May '05

09 May '05

Hmm... Ok Dave made some pretty good points below. Thinking... The perspective I am coming from is this. I have been connected to the boost world now for two years. I am a relative newbie. But I am thinking back to the things that can "immediately hook" a new person. I think the points you raised towards mine are all valid. I think the disconnect is we are thinking about different use cases if you will. I am coming from a "ok so a non-boost person hits the page, how do we best speak to that person" point of view. I think your points come more from a practical day to day perspective. I respect that. I can speak FROM the "newbie" perspective because it wasn't that long ago for me. And I can tell you that it takes an awful lot of patience and persistence to become someone who "sees what boost does" and can actually start using it in projects. The work done on the web site an the libraries is of excellent quality. For that I applaud. But for the person coming in cold, the bigger picture is difficult to get initially. I believe the issues I am raising can be best spoken to by: 1- Driving and encouraging more use of the Wiki 2- I would place a link to the users boost list and the wiki somewhere in the upper left of the page 3- Having a quick run down of each library also linked in the same area (this was suggested before) 4- Having a runtime environment for boost (like CGYWin does) (I have done something like this internal to our company- Dave- you may recall me showing you this at oopsla) I believe these suggestions would help the newbie. Of course these have to be balanced with the day to day needs of the day to day people. Brian :) > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of David Abrahams > Sent: Friday, May 06, 2005 6:05 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: The Boost Community > > "Brian Braatz" <brianb(a)rmtg.com> writes: > > > > > [Brian Braatz Writes:] > > Did some more peeking. Since I use outlook for everything (or the main > > page's archive searching) I was not aware of this view > > http://lists.boost.org/boost/2005/03/date.php > > > > This speaks to what I was raising before. However it is not "in my > > face" (I never noticed this view existed on the site before today). > > Maybe because the main Boost web archive was only recently changed to > use new software. > > > I would like to amend my suggestion to be the following: > > > > 1- Formal link to the list on boost.org main page > > There are lots of Boost mailing lists. If we drive everything we > think is important up to the top level it will become even harder to > use than it already is. What's wrong with > www.boost.org/more/mailing_lists.htm, which is prominently linked from > the main page? > > > 2- Boost wiki links also a "first class citizen" link on the first > > page I.e. it goes from the BOTTOM of the page to the TOP- and it > > goes from being "Unofficial" to being one if not the FIRST thing on > > the list > > > > My suggestion would be to list the left pane like so: > > > > Boost Wiki > > Boost Mailing list (takes you to the lists view above) > > I don't know about that; that view doesn't tell you how to post or > subscribe to the list. I don't think the "Boost Mailing List" link > ought to lead people down a tunnel that only gives them a view of the > archive. > > > And then everything else > > > > Addtionally- I would add putting a link to the Wiki in the mail > > (previously discussed) and at the top of the "archive view" link > > above. (responding to my own post) > > Not sure about that one either. We need some criteria to decide what > gets driven to the top level everywhere. > > -- > Dave Abrahams > Boost Consulting > www.boost-consulting.com > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

6 6

Re:
by pankaj jain 09 May '05

09 May '05

Hi , I am new user of boost.python. Please help me in solving this problem. problem 1) When I am running bjam from home/boost/libs/python/example/tutorial folder its running fine but building my hello.so in home/boost/bin/boost/libs/python/example/tutorial/hello.so/gcc/.. folder.is there any way to chagne the path for hello.so problem 2): when I am running bjam from any other folder ( home/boost/sample) its giving me this error :: /home/jamrules no such directory .../.../allyourbase.jam 1451 see definition of rules "root-path: being called. Please let me know any solution. Thanks for all your help Pankaj Doug Gregor <dgregor(a)cs.indiana.edu> wrote: On May 9, 2005, at 5:54 AM, Rui Carvalho wrote: > Hi, > > I have a directed graph (vecS, vecS, bidirectionalS) and I'm trying to > iterate through the neighbour vertices of a vertex u. When using > > for (tie(ai0,ai0_end) = adjacent_vertices(u,g); ai0 != ai0_end; ++ai0) > > the code goes only through the out vertices of u. Is there a way to > iterate > through all the neighbouring vertices in one loop (i.e. through both > the out > and in vertices of u)? I don't know of a good way to do this with the BGL. You'll need to either build some kind of smart iterator adaptor that traverses out_edges then in_edges, or have two separate loops. Doug _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users --------------------------------- Do you Yahoo!? Take Yahoo! Mail with you! Get it on your mobile phone.

1 0

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 08 May '05

08 May '05

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Beman Dawes > > FWIW, I really like http://community.boost.org > [Brian Braatz Writes:] Yes, it grows on you after awhile. What I like about it, is the undertone of community.

1 0

RE: [Boost-users] Re: The Boost Community
by Foster, Gareth 08 May '05

08 May '05

Wouldn't something simply and descriptive like http://boostcommunity.org/ or http://community.boost.org/ make more sense? I know everyone is thinking marketing etc, but I think it will be possible to come up with something along those practical lines, that still sounds slick. Thesaurus entry for community anyone? Gaz > -----Original Message----- > From: Brian Braatz [mailto:brianb@rmtg.com] > Sent: 05 May 2005 20:55 > To: boost-users(a)lists.boost.org > Subject: RE: [Boost-users] Re: The Boost Community > > www.igotboost.org is available > > www.gotboosted.org and .com is available > > www.getboosted.org is available > > www.getmeboosted.org and .com is available > > www.boostme.net is available. > > www.boostme.org is available. > > > > -----Original Message----- > > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > > bounces(a)lists.boost.org] On Behalf Of Vladimir Krasovsky > > Sent: Thursday, May 05, 2005 1:13 AM > > To: David Abrahams > > Subject: Re: [Boost-users] Re: The Boost Community > > > > Hello David, > > > > Thursday, May 5, 2005, 12:17:14 AM, you wrote: > > > > > I actually love that. > > > > > www.gotboost.org > > > > By the way this domain is already used by someone. Can it be a > > problem? > > > > > > -- > > Best regards, > > Vladimir mailto:vkrasovsky@yandex.ru > > > > > > _______________________________________________ > > Boost-users mailing list > > Boost-users(a)lists.boost.org > > http://lists.boost.org/mailman/listinfo.cgi/boost-users > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

4 3

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 08 May '05

08 May '05

HAHAHAHAHA I was just trying anything I could think of.. (name wise) Though "www.pimpmycode" does have a interesting ring to it... What we need is a discovery channel show, where we talk boring old app code and "pimp it out".... Geeks all over the world would be tuning in... > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of David Abrahams > Sent: Friday, May 06, 2005 6:06 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: The Boost Community > > "Brian Braatz" <brianb(a)rmtg.com> writes: > > > boostmycode.com is available. > > boostmycode.net is available. > > boostmycode.org is available. > > These are kinda like "pimp my ride." Fun, but... appropriate? > > -- > Dave Abrahams > Boost Consulting > www.boost-consulting.com > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

iostreams build problem (current cvs)
by Chris Hoeppler 08 May '05

08 May '05

Hi, using a current boost download from cvs, I encountered the following build problem: the iostreams build seems to somehow pick up the names of the source files of the serialization lib. I am appending the head of the bjam output. Any idea what's going on here? Thanks Chris chris@fred:~/boost_050505> bjam stage ****************************************************** Building Boost.Regex with Unicode/ICU support disabled. HINT: define the environment variable ICU_PATH to point to the root directy of your ICU installation if you have one. Couldn't find utypes.h in ****************************************************** ...patience... don't know how to make <@boost!libs!iostreams!build>../src/basic_archive.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_iarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_oarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_iserializer.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_oserializer.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_pointer_iserializer.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_pointer_oserializer.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_serializer_map.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_text_iprimitive.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_text_oprimitive.cpp don't know how to make <@boost!libs!iostreams!build>../src/basic_xml_archive.cpp don't know how to make <@boost!libs!iostreams!build>../src/binary_iarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/binary_oarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/extended_type_info.cpp don't know how to make <@boost!libs!iostreams!build>../src/extended_type_info_no_rtti.cpp don't know how to make <@boost!libs!iostreams!build>../src/extended_type_info_typeid.cpp don't know how to make <@boost!libs!iostreams!build>../src/polymorphic_iarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/polymorphic_oarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/stl_port.cpp don't know how to make <@boost!libs!iostreams!build>../src/text_iarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/text_oarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/void_cast.cpp don't know how to make <@boost!libs!iostreams!build>../src/xml_grammar.cpp don't know how to make <@boost!libs!iostreams!build>../src/xml_iarchive.cpp don't know how to make <@boost!libs!iostreams!build>../src/xml_oarchive.cpp ...patience... ...found 7779 targets... ...updating 1725 targets... ...can't find 25 targets... ...can't make 220 targets... MkDir1 stage MkDir1 stage/lib MkDir1 bin MkDir1 bin/boost MkDir1 bin/boost/libs MkDir1 bin/boost/libs/test MkDir1 bin/boost/libs/test/build MkDir1 bin/boost/libs/test/build/libboost_prg_exec_monitor.so MkDir1 bin/boost/libs/test/build/libboost_prg_exec_monitor.so/gcc MkDir1 bin/boost/libs/test/build/libboost_prg_exec_monitor.so/gcc/debug MkDir1 bin/boost/libs/test/build/libboost_prg_exec_monitor.so/gcc/debug/shared-linkable -true gcc-C++-action bin/boost/libs/test/build/libboost_prg_exec_monitor.so/gcc/debug/shared-linkable -true/execution_monitor.o <snip> -- +++ Lassen Sie Ihren Gedanken freien Lauf... z.B. per FreeSMS +++ GMX bietet bis zu 100 FreeSMS/Monat: http://www.gmx.net/de/go/mail

3 4

Boost serialization and dependency nightmares...
by kauai 08 May '05

08 May '05

Hi all, I have a container class that has a vector which contains several other classes. My container class has a serialize function that looks like this: friend class boost::serialization::access; template<class Archive> void serialize (Archive &ar, const unsigned int version) { // save/load base class information ar & BOOST_SERIALIZATION_BASE_OBJECT_NVP(MyBaseClass); ar.register_type(static_cast<SomeClass1*>(NULL)); ar.register_type(static_cast<SomeClass2*>(NULL)); ... ar & BOOST_SERIALIZATION_NVP(m_Vector); // std::vector } Got everything working but now an issue has come up: Someone wants to add a new class but it is not part of this package that is doing the serialzation. I want to avoid what I call "dependency hell", is there a way for that new class to get registered someplace other than in this function? Thanks for any help, Kauai Discover Yahoo! Stay in touch with email, IM, photo sharing and more. Check it out! http://discover.yahoo.com/stayintouch.html

2 1

subclassing shared_ptr?
by Jean-François Brouillet 08 May '05

08 May '05

- I'm porting to C++ a huge java library. - Java constructor semantics is close to that of C++ but not identical. - boost::shared_ptr does nearly exactly what I want except for 1 little thing peculiar to the project I'm working on. The short story is that, for the C++ version, I need an "after ctor" method to be called before any other member function is allowed to be called. The logical next step was to override operator-> which led me to a long route ending now with boost::shared_ptr. I should be able to solve my problem by _patching_ my local copy of shared_ptr.hpp as follows: ... template<class Y> explicit shared_ptr(Y * p): px(p), pn(p, checked_deleter<Y>()) // Y must be complete { p->constructor() ; // single patch for my purpose detail::sp_enable_shared_from_this( pn, p, p ); } ... But obviously, I'd rather not do that. So I came up with: template <typename T> struct Bridge : public boost::shared_ptr<T> { template <typename Y> explicit Bridge(Y * y) : boost::shared_ptr<Y>(y) { if (y) { y->constructor() ; } } } ; But then, the guarantee given in the boost shared_ptr documenation was broken: Quote: shared_ptr<T> can be implicitly converted to shared_ptr<U> whenever T* can be implicitly converted to U*. In particular, shared_ptr<T> is implicitly convertible to shared_ptr<T const>, to shared_ptr<U> where U is an accessible base of T, and to shared_ptr<void>. So, I added a copy constructor: template <typename T> struct Bridge : public boost::shared_ptr<T> { template <typename Y> explicit Bridge(Y * y) : boost::shared_ptr<Y>(y) { if (y) { y->javaConstructor() ; } } template<typename Y> Bridge(Bridge<Y> const & r) : boost::shared_ptr<Y>(r) { } } ; But then I get: /Users/verec/Tools/boost_1_32_0/osx/main.cpp: In constructor `Bridge<T>::Bridge(const Bridge<Y>&) [with Y = DisplayStruct, T = DeviceStruct]': /Users/verec/Tools/boost_1_32_0/osx/main.cpp:243: instantiated from here /Users/verec/Tools/boost_1_32_0/osx/main.cpp:21: error: type 'class boost::shared_ptr<DisplayStruct>' is not a direct base of 'Bridge<DeviceStruct>' Here's the relevant part of the code: #include <iostream> #include <vector> #include <string> #include <sstream> #include <boost/shared_ptr.hpp> struct ObjectStruct { ObjectStruct() { } ; virtual void constructor() { } virtual ~ObjectStruct() { } } ; typedef Bridge<ObjectStruct> Object ; struct DeviceStruct : public ObjectStruct { typedef Bridge<DeviceStruct> Device ; DeviceStruct() { std::cout << "Device standard c++ constructor." << std::endl ; } virtual void deviceMethod() { std::cout << "DeviceStruct::deviceMethod." << std::endl ; } virtual void constructor() { std::cout << "java's Device::constructor." << std::endl ; } static Device current ; static Device getCurrent() { return current ; } } ; typedef Bridge<DeviceStruct> Device ; struct DisplayStruct : public DeviceStruct { typedef Bridge<DisplayStruct> Display ; DisplayStruct() { std::cout << "Display standard c++ constructor." << std::endl ; } DisplayStruct() { std::cout << "Display standard c++ constructor." << std::endl ; } virtual void constructor() { DeviceStruct::constructor () ; std::cout << "java's Display::constructor." << std::endl ; } virtual void deviceMethod() { DeviceStruct::deviceMethod () ; std::cout << "DisplayStruct::deviceMethod." << std::endl ; } virtual void displayMethod() { std::cout << "displayMethod." << std::endl ; } } ; typedef Bridge<DisplayStruct> Display ; void callDevice(Device d) { d->deviceMethod() ; } int main (int argc, char * const argv[]) { Display display(new Display) ; Device device(new DeviceStruct) ; callDevice(device) ; callDevice(display) ; return 0 ; } What am I missing ? Many Thanks. -- Jean-François Brouillet verec(a)mac.com

2 2

Regex matching of an empty string
by Edward Diener 07 May '05

07 May '05

If the string to match is an empty string, "", and the regular expression is such that it matches the empty string, such as "^$", should a match be found ?

3 2

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 07 May '05

07 May '05

www.igotboost.org is available www.gotboosted.org and .com is available www.getboosted.org is available www.getmeboosted.org and .com is available www.boostme.net is available. www.boostme.org is available. > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Vladimir Krasovsky > Sent: Thursday, May 05, 2005 1:13 AM > To: David Abrahams > Subject: Re: [Boost-users] Re: The Boost Community > > Hello David, > > Thursday, May 5, 2005, 12:17:14 AM, you wrote: > > > I actually love that. > > > www.gotboost.org > > By the way this domain is already used by someone. Can it be a > problem? > > > -- > Best regards, > Vladimir mailto:vkrasovsky@yandex.ru > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

3 2

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 07 May '05

07 May '05

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Jon Kalb > We could change the configuration of the footer for messages sent to this > list to include the wiki's URL. > > > _______________________________________________ > > Boost-users mailing list > > Boost-users(a)lists.boost.org > > http://lists.boost.org/mailman/listinfo.cgi/boost-users > > Could be something like: > > ______________________________________________________________ > Boost-users(a)lists.boost.org http://boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl > > I'm not crazy about adding more text to the footer and I'm not certain > that > anyone actually ever *reads* the footer, but perhaps it would raise the > visibility of the wiki. (I wish the URL wasn't so long.) > > -- > Jon Kalb > Kalb(a)LibertySoft.com > [Brian Braatz Writes:] This is a good idea. My simple suggestion is to do this only use tinyurl.

4 4

Boost.Jam - long file names in .a files lookup fix
by Snaury 06 May '05

06 May '05

Hello, boost-users, There is a bug in file `filent.c' in function `file_archscan', which causes long .o files in archives invisible to jam, and thus jam always recompiles such .o files, when using, for instance, mingw. The problem is in microsoft's coff definition of stringtables in archives, in which each name must be ended by '\0' versus (linux?) coff, where '\n' is used instead. The following patch is trivial, and I hope someone with cvs access can apply it. Btw, the same 'bug' is in current perforce's jam. diff -druN boost-jam-3.1.10/filent.c boost-jam-3.1.10-df/filent.c --- boost-jam-3.1.10/filent.c Tue Jun 1 09:45:14 2004 +++ boost-jam-3.1.10-df/filent.c Tue May 3 16:07:51 2005 @@ -251,7 +251,8 @@ */ name = string_table + atoi( ar_hdr.ar_name + 1 ); - endname = name + strlen( name ); + for( endname = name; *endname && *endname != '\n'; ++endname ) + /* nothing */; } else {

2 1

Boost and QNX 6.3.0
by Alexandre Sénéchal 06 May '05

06 May '05

Hi, Anyone knows how to make Boost.Threads libray compile and link on QNX 6.3.0 ? I tried some things, including : ./bjam "-sTOOLS=gcc" "-sGXX=QCC -V3.3.1,gcc_ntox86" "-sBUILD=debug release <runtime-link>dynamic <threading>multi" ... and it ended up telling me : cc: unknown option -pthread. I know there is no configuration whatsoever already defined for QNX, as stated by #error "Unknown Platform -- Please configure .... " How do I "configure" and what do I need to configure ? Thanks, Alexandre Senechal

2 2

Boost.Serialization Problems
by Brett Gmoser 06 May '05

06 May '05

Hello. I'm trying to use the boost serialization library. I'm using Visual C++ 7.1 under Windows 2000. The Boost.Serialization library seems excellent, and its ease of use is great. However I'm having two problems with it: First Problem. Here's the code: cdx::shrec_script script(""); std::ofstream ofs("compiled_shrec.txt"); boost::archive::text_oarchive oa(ofs); oa << script; // close archive ofs.close(); I get an error from boost static_assert, static_assert.hpp (below). Line 437 of regexconv.hpp refers to the "oa << script;" statement. If I copy the script archive to a new const object: const cdx::shrec_script temp_script(script); And try to serialize the temp_script object, the error goes away, and everything (seems to) work okay. I don't really have a problem with copying it, although it is inconvenient, in order to get this to work. However, the real problem is when I try to de-serialize the class: cdx::shrec_script script(""); std::ifstream ifs("compiled_shrec.txt"); boost::archive::text_iarchive oa(ifs); ia << script; // close archive ifs.close(); I get the same error (below). This is the big problem, because obviously the object needs to be non-const to be altered with the new data. I tried doing the same trick as above to get it to work (copy to a const object), but I still get the STATIC_ASSERTION_FAILURE error. c:\Boost\include\boost-1_33\boost\archive\detail\oserializer.hpp(551) : error C2027: use of undefined type 'boost::STATIC_ASSERTION_FAILURE<x>' with [ x=false ] c:\Boost\include\boost-1_33\boost\archive\detail\oserializer.hpp(551) : see reference to class template instantiation 'boost::STATIC_ASSERTION_FAILURE<x>' being compiled with [ x=false ] c:\Boost\include\boost-1_33\boost\archive\basic_text_oarchive.hpp(78) : see reference to function template instantiation 'void boost::archive::save<Archive,T>(Archive &,T &)' being compiled with [ Archive=boost::archive::text_oarchive, T=cdx::shrec_script ] c:\Boost\include\boost-1_33\boost\archive\detail\interface_oarchive.hpp(76) : see reference to function template instantiation 'void boost::archive::basic_text_oarchive<Archive>::save_override<T>(T &,int)' being compiled with [ Archive=boost::archive::text_oarchive, T=cdx::shrec_script ] regexconv.hpp(437) : see reference to function template instantiation 'Archive &boost::archive::detail::interface_oarchive<Archive>::operator <<<cdx::shrec_script>(T &)' being compiled with [ Archive=boost::archive::text_oarchive, T=cdx::shrec_script ] Second Problem. I'm getting linker errors when trying to compile Boost.Serialization with multi threaded runtimes in VC71. Linker errors are pasted at the bottom of this message. Somebody had told me that boost-1.31 and up headers now provide the proper #pragma for linking with the correct static library automagically, but I tried adding the #pragma's myself anyway. I'm not sure exactly what library to link with (there are lots of libbost_serialization*.lib files), so I tried each one. Each produced different results with different linker errors ranging from multiply defined symbols to unresolved externals. One did compile, however: #pragma comment(lib, "libboost_serialization-vc71-s.lib") But the program crashes as soon as it is run with the error (an uncaught exception): This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information. Press any key to continue I'm assuming this may be related to the fact that the 's' in libboost_serialization-vc71-s.lib means that the lib is compiled for single threads, and this is a multi threaded program. But of course, I'm probably wrong. Any light anybody can shed on either of these issues would be greatly appreciated! Thanks! Brett aftli(a)optonline.net Compiling... boost_test.cpp Linking... Creating library Release/boost_test.lib and object Release/boost_test.exp boost_test.obj : error LNK2001: unresolved external symbol "private: static char const * const boost::serialization::detail::extended_type_info_typeid_0::type_info_key" (?type_info_key@extended_type_info_typeid_0@detail@serialization@boost@@0PBD B) boost_test.obj : error LNK2019: unresolved external symbol "protected: virtual __thiscall boost::archive::detail::basic_iarchive::~basic_iarchive(void)" (??1basic_iarchive@detail@archive@boost@@MAE@XZ) referenced in function "public: virtual __thiscall boost::archive::detail::common_iarchive<class boost::archive::text_iarchive>::~common_iarchive<class boost::archive::text_iarchive>(void)" (??1?$common_iarchive@Vtext_iarchive@archive@boost@@@detail@archive@boost@@U AE@XZ) boost_test.obj : error LNK2019: unresolved external symbol "protected: virtual __thiscall boost::archive::detail::basic_oarchive::~basic_oarchive(void)" (??1basic_oarchive@detail@archive@boost@@MAE@XZ) referenced in function "public: virtual __thiscall boost::archive::detail::common_oarchive<class boost::archive::text_oarchive>::~common_oarchive<class boost::archive::text_oarchive>(void)" (??1?$common_oarchive@Vtext_oarchive@archive@boost@@@detail@archive@boost@@U AE@XZ) boost_test.obj : error LNK2019: unresolved external symbol "protected: void __thiscall boost::serialization::extended_type_info::self_register(void)" (?self_register@extended_type_info@serialization@boost@@IAEXXZ) referenced in function "private: __thiscall boost::serialization::extended_type_info_typeid<class gps_position const >::extended_type_info_typeid<class gps_position const >(void)" (??0?$extended_type_info_typeid@$$CBVgps_position@@@serialization@boost@@AAE @XZ) boost_test.obj : error LNK2019: unresolved external symbol "public: void __thiscall boost::archive::text_oarchive_impl<class boost::archive::text_oarchive>::save(class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > const &)" (?save@?$text_oarchive_impl@Vtext_oarchive@archive@boost@@@archive@boost@@QA EXABV?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@@Z) referenced in function "public: static void __cdecl boost::archive::save_access::save_primitive<class boost::archive::text_oarchive,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > >(class boost::archive::text_oarchive &,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > const &)" (??$save_primitive@Vtext_oarchive@archive@boost@@V?$basic_string@DU?$char_tr aits@D@std@@V?$allocator@D@2@@std@@@save_access@archive@boost@@SAXAAVtext_oa rchive@12@ABV?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@@Z ) boost_test.obj : error LNK2019: unresolved external symbol "public: void __thiscall boost::archive::text_iarchive_impl<class boost::archive::text_iarchive>::load(class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > &)" (?load@?$text_iarchive_impl@Vtext_iarchive@archive@boost@@@archive@boost@@QA EXAAV?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@@Z) referenced in function "public: static void __cdecl boost::archive::load_access::load_primitive<class boost::archive::text_iarchive,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > >(class boost::archive::text_iarchive &,class std::basic_string<char,struct std::char_traits<char>,class std::allocator<char> > &)" (??$load_primitive@Vtext_iarchive@archive@boost@@V?$basic_string@DU?$char_tr aits@D@std@@V?$allocator@D@2@@std@@@load_access@archive@boost@@SAXAAVtext_ia rchive@12@AAV?$basic_string@DU?$char_traits@D@std@@V?$allocator@D@2@@std@@@Z ) boost_test.obj : error LNK2019: unresolved external symbol "protected: __thiscall boost::archive::detail::basic_oarchive::basic_oarchive(void)" (??0basic_oarchive@detail@archive@boost@@IAE@XZ) referenced in function "protected: __thiscall boost::archive::detail::common_oarchive<class boost::archive::text_oarchive>::common_oarchive<class boost::archive::text_oarchive>(void)" (??0?$common_oarchive@Vtext_oarchive@archive@boost@@@detail@archive@boost@@I AE@XZ) boost_test.obj : error LNK2019: unresolved external symbol "protected: __thiscall boost::archive::detail::basic_iarchive::basic_iarchive(void)" (??0basic_iarchive@detail@archive@boost@@IAE@XZ) referenced in function "protected: __thiscall boost::archive::detail::common_iarchive<class boost::archive::text_iarchive>::common_iarchive<class boost::archive::text_iarchive>(void)" (??0?$common_iarchive@Vtext_iarchive@archive@boost@@@detail@archive@boost@@I AE@XZ) boost_test.obj : error LNK2019: unresolved external symbol "public: __thiscall boost::archive::basic_text_oprimitive<class std::basic_ostream<char,struct std::char_traits<char> > >::~basic_text_oprimitive<class std::basic_ostream<char,struct std::char_traits<char> > >(void)" (??1?$basic_text_oprimitive@V?$basic_ostream@DU?$char_traits@D@std@@@std@@@a rchive@boost@@QAE@XZ) referenced in function "public: virtual __thiscall boost::archive::text_oarchive_impl<class boost::archive::text_oarchive>::~text_oarchive_impl<class boost::archive::text_oarchive>(void)" (??1?$text_oarchive_impl@Vtext_oarchive@archive@boost@@@archive@boost@@UAE@X Z) boost_test.obj : error LNK2019: unresolved external symbol "public: __thiscall boost::archive::basic_text_iprimitive<class std::basic_istream<char,struct std::char_traits<char> > >::~basic_text_iprimitive<class std::basic_istream<char,struct std::char_traits<char> > >(void)" (??1?$basic_text_iprimitive@V?$basic_istream@DU?$char_traits@D@std@@@std@@@a rchive@boost@@QAE@XZ) referenced in function "public: virtual __thiscall boost::archive::text_iarchive_impl<class boost::archive::text_iarchive>::~text_iarchive_impl<class boost::archive::text_iarchive>(void)" (??1?$text_iarchive_impl@Vtext_iarchive@archive@boost@@@archive@boost@@UAE@X Z) boost_test.obj : error LNK2019: unresolved external symbol "protected: void __thiscall boost::archive::basic_text_iarchive<class boost::archive::text_iarchive>::load_override(struct boost::archive::class_name_type &,int)" (?load_override@?$basic_text_iarchive@Vtext_iarchive@archive@boost@@@archive @boost@@IAEXAAUclass_name_type@23@H@Z) referenced in function "private: virtual void __thiscall boost::archive::detail::common_iarchive<class boost::archive::text_iarchive>::vload(struct boost::archive::class_name_type &)" (?vload@?$common_iarchive@Vtext_iarchive@archive@boost@@@detail@archive@boos t@@EAEXAAUclass_name_type@34@@Z) boost_test.obj : error LNK2019: unresolved external symbol "public: void __thiscall boost::archive::detail::basic_iarchive::load_object(void *,class boost::archive::detail::basic_iserializer const &)" (?load_object@basic_iarchive@detail@archive@boost@@QAEXPAXABVbasic_iserializ er@234@@Z) referenced in function "public: static void __cdecl boost::archive::detail::load_non_pointer_type<class boost::archive::text_iarchive,class gps_position>::load::invoke(class boost::archive::text_iarchive &,class gps_position &)" (?invoke@load@?$load_non_pointer_type@Vtext_iarchive@archive@boost@@Vgps_pos ition@@@detail@archive@boost@@SAXAAVtext_iarchive@45@AAVgps_position@@@Z) boost_test.obj : error LNK2019: unresolved external symbol "public: void __thiscall boost::archive::detail::basic_oarchive::save_object(void const *,class boost::archive::detail::basic_oserializer const &)" (?save_object@basic_oarchive@detail@archive@boost@@QAEXPBXABVbasic_oserializ er@234@@Z) referenced in function "public: static void __cdecl boost::archive::detail::save_non_pointer_type<class boost::archive::text_oarchive,class gps_position>::save::invokex(class boost::archive::text_oarchive &,class gps_position const &)" (?invokex@save@?$save_non_pointer_type@Vtext_oarchive@archive@boost@@Vgps_po sition@@@detail@archive@boost@@SAXAAVtext_oarchive@45@ABVgps_position@@@Z) boost_test.obj : error LNK2001: unresolved external symbol "struct boost::archive::version_type const boost::archive::ARCHIVE_VERSION" (?ARCHIVE_VERSION@archive@boost@@3Uversion_type@12@B) boost_test.obj : error LNK2001: unresolved external symbol "char const * const boost::archive::ARCHIVE_SIGNATURE" (?ARCHIVE_SIGNATURE@archive@boost@@3PBDB) boost_test.obj : error LNK2019: unresolved external symbol "public: __thiscall boost::archive::basic_text_oprimitive<class std::basic_ostream<char,struct std::char_traits<char> > >::basic_text_oprimitive<class std::basic_ostream<char,struct std::char_traits<char> > >(class std::basic_ostream<char,struct std::char_traits<char> > &,bool)" (??0?$basic_text_oprimitive@V?$basic_ostream@DU?$char_traits@D@std@@@std@@@a rchive@boost@@QAE@AAV?$basic_ostream@DU?$char_traits@D@std@@@std@@_N@Z) referenced in function "protected: __thiscall boost::archive::text_oarchive_impl<class boost::archive::text_oarchive>::text_oarchive_impl<class boost::archive::text_oarchive>(class std::basic_ostream<char,struct std::char_traits<char> > &,unsigned int)" (??0?$text_oarchive_impl@Vtext_oarchive@archive@boost@@@archive@boost@@IAE@A AV?$basic_ostream@DU?$char_traits@D@std@@@std@@I@Z) boost_test.obj : error LNK2019: unresolved external symbol "public: __thiscall boost::archive::basic_text_iprimitive<class std::basic_istream<char,struct std::char_traits<char> > >::basic_text_iprimitive<class std::basic_istream<char,struct std::char_traits<char> > >(class std::basic_istream<char,struct std::char_traits<char> > &,bool)" (??0?$basic_text_iprimitive@V?$basic_istream@DU?$char_traits@D@std@@@std@@@a rchive@boost@@QAE@AAV?$basic_istream@DU?$char_traits@D@std@@@std@@_N@Z) referenced in function "public: __thiscall boost::archive::text_iarchive_impl<class boost::archive::text_iarchive>::text_iarchive_impl<class boost::archive::text_iarchive>(class std::basic_istream<char,struct std::char_traits<char> > &,unsigned int)" (??0?$text_iarchive_impl@Vtext_iarchive@archive@boost@@@archive@boost@@QAE@A AV?$basic_istream@DU?$char_traits@D@std@@@std@@I@Z) Release/boost_test.exe : fatal error LNK1120: 17 unresolved externals Build Time 0:01

2 3

RE: [Boost-users] Re: Re: The Boost Community
by Simon Li 06 May '05

06 May '05

> > Be aware that dynamic sites like Plone, Mambo, etc can demand a LOT > > from your back-end (web servers, MySQL or whatever database) if the > > site gets serious traffic. > > > Can you put this in numbers so we can evaluate a prospective > Server (and using those technologies at all)? > Just my 2c. Don't forget that the aim should be to provide a site thats both functional and EASY to use. I've briefly looked at Plone, and my first impressions were that while it seems to be extremely comprehensive and have loads of features, you could spend days learning how to configure it and it took a while to work out how a user should add content to it. Admittedly I didn't spend that long on it, and it might be possible to configure things to make it easier. One of the main advantages that I can see in a Wiki is that it's relatively familiar to people, and in some respects is "self maintaining", in that it doesn't require dedicated admins to look after it. Simon This e-mail message (including its attachments) is private, is intended for the recipient named in it and may contain material which is confidential and privileged. No-one other than the named recipient may read, copy, rely on, redirect, save or alter the message or any part of it or any attachment to it in any way. VMS does not accept legal responsibility for the contents of this message. Any views or opinions presented are solely those of the author and do not represent those of VMS unless otherwise specifically stated. While reasonable effort has been made to ensure this message is free of viruses, opening and using this message is at the risk of the recipient.

1 0

Different ways of using boost
by Shetty, Diwakar 06 May '05

06 May '05

Hi all, I have come to the following conclusions about using boost. Do please let me know if I am wrong. 1) Download the boost src from web. 2) Build/Install the various libs using "bjam" (For intel, one can download the lib itself) 3) To use the boost lib in one's code, 3 ways are a) Just include the hdr files (e.g.: bind.hpp) The entire template impln is in hdr file b) Include the hdr files and boost will link the reqd libs automatically (e.g.: thread lib) c) Include the hdr file and specify the ".lib" file yourself. Also, whether a particular feature require 3a, 3b or 3c has to be figured out by yourself. Thanks, Diwakar

1 0

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 06 May '05

06 May '05

Yes, I like those the best too.. For fun here are some more: boostifymycode.com is available. boostifymycode.net is available. boostifymycode.org is available. boostmycode.com is available. boostmycode.net is available. boostmycode.org is available. getdownwithyourboostself.com is available. getdownwithyourboostself.net is available. getdownwithyourboostself.org is available. boostyou.net is available. boostyou.org is available. gotmetaprogramming.com is available. gotmetaprogramming.net is available. gotmetaprogramming.org is available. gottemplates.net is available. gottemplates.org is available. gotmpl.com is available. gotmpl.net is available. gotmpl.org is available. gottemplatemetaprogramming.com is available. gottemplatemetaprogramming.net is available. gottemplatemetaprogramming.org is available. gotstaticprogramming.com is available. gotstaticprogramming.net is available. gotstaticprogramming.org is available. getyourboostthangon.com is available. getyourboostthangon.net is available. getyourboostthangon.org is available. Though I am partial to "boostme" > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of David Abrahams > Sent: Thursday, May 05, 2005 3:33 PM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: The Boost Community > > "Brian Braatz" <brianb(a)rmtg.com> writes: > > > www.boostme.net is available. > > > > www.boostme.org is available. > > I like these. > > -- > Dave Abrahams > Boost Consulting > www.boost-consulting.com > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

2 1

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 06 May '05

06 May '05

(responding to my own post) > -----Original Message----- > From: Brian Braatz > > > -----Original Message----- > > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > > bounces(a)lists.boost.org] On Behalf Of David Abrahams > [David Abrahams Writes:] > > Hmm, I'm not sure we can convince everyone to look at the Wiki before > > going to the archive pages. > > > > [Brian Braatz Writes:] > Yes, it was a weak idea. But generally, the more we can connect the wiki > to the mailing list the better. This basic idea hinges upon somehow > providing a "really cool and compelling" way to dig around in the archives > other than just searching for keywords. > > When I click on news://news.gmane.org/gmane.comp.lib.boost.user > It opens up Outlook Express and gives me a pretty good view of the list. > (much better than the Outlook folder I use now for the list) > > We should be able to find a WEB TOOL that gives the same view. The idea is > if you make that web tool a first class citizen and let me hit the page > for digging around in the subjects of the list, and additionally > "advertise" the wiki in the header of said tool's page, then you have > connected the paradigms. > > :) [Brian Braatz Writes:] Did some more peeking. Since I use outlook for everything (or the main page's archive searching) I was not aware of this view http://lists.boost.org/boost/2005/03/date.php This speaks to what I was raising before. However it is not "in my face" (I never noticed this view existed on the site before today). I would like to amend my suggestion to be the following: 1- Formal link to the list on boost.org main page 2- Boost wiki links also a "first class citizen" link on the first page I.e. it goes from the BOTTOM of the page to the TOP- and it goes from being "Unofficial" to being one if not the FIRST thing on the list My suggestion would be to list the left pane like so: Boost Wiki Boost Mailing list (takes you to the lists view above) And then everything else Addtionally- I would add putting a link to the Wiki in the mail (previously discussed) and at the top of the "archive view" link above.

2 1

boost:bind versus bind1st
by Shetty, Diwakar 06 May '05

06 May '05

Will it be okay to say that the advantages of boost::bind over bind1st and bind2nd are 1) bind1st and bind2nd works with func obj only 2) boost::bind does not require func to have result_type and first_argument_type Thanks, Diwakar

3 2

Re: [Boost-users] The Boost Community
by JOAQUIN LOPEZ MU?Z 06 May '05

06 May '05

----- Mensaje original ----- De: Fernando Cacciola <fernando_cacciola(a)hotmail.com> Fecha: Martes, Mayo 3, 2005 6:41 pm Asunto: [Boost-users] The Boost Community > Hello people, > > In 1998, Bemans Dawes wrote the "Proposal for a C++ Library > Repository Web > Site" (you can find the paper in (www.boost.org/more/index.htm) > which > lunched the Boost we all know and love today. > The focus of the site was and is the production of high-quality > peer-reviewed C++ libraries. > Over the years, Boost grew as a community of experienced C++ > developers with > the will and means to share their expertise and work contributing > to the > highly respected Boost Libraries. Today, Boost is a somewhat > recongisable > part of the C++ programming community at large. > > Today, the way I see it, Boost as a community encompases only the > boost > developers, which are just a fraction of the larger C++ crowd. > There are > certainly Boost users, but I don't think we can say both > developers and > users form a community. I think is worth and in-the-spirit to > reach out for > ALL C++ programmers of ALL LEVELS around an "extended" Boost > Community whose > main sharing point is not just the submission of high-quality > libraries but > also the efficient and effective use of the language. > > I violently support the creation of such a site :) Let me throw in the following issues: 1. Name of the site. As I see it, it should mention the word "Boost" so as to stand clearly among other C++-devoted sites. Something like "Got Boost" (not that I particularly love this name, but you get the idea.) I propose that we do *not* launch a contest to find the name, lest we end up spending three omnths on the issue like it happend witht he Boost logo. 2. Hosting. Dave Abrahams volunteered to host the site. Is he still holding his offer? 3. Technology. I've been doing a small survey among my workmates and some of them are ardent fans of Plone (www.plone.org) an off-the-shelf community product built on top of a content management system called Zope. I lack direct experience with Plone, but from the info I've been able to gather it supports the following features: * Site structuring using the Zope backend. * News sections. * Forums (with a component called CMF.) * Look&feel tweaking thru CSS. If you want to dig deeper, Zope provides a scripting framework using Python. * Built-in hierarchical database. * User management. * Updating workflow, meaning that everybody can contibrute entries (submitting an HTML doc or with a built-in WYSIWG editor) wiki-style, but we can make these pass thru a moderating filter. * Built-in search. * Open source, free. Preferred deployment platform is Linux, but works on Windows and Mac boxes. As I said, I don't have direct experience with Plone, so if anybody out there does maybe she can assist us on wether this is a suitable product to build the site on. I pass on less tecnhical questions like what kind of content we want to include: IMHO once we have some site prototype set up, it is easier to propose ideas and let the the thing evolve organically, so to speak. Best, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

9 12

FOREACH with only const_iterators?
by Dave Steffen 06 May '05

06 May '05

Given that this is the user list and not the development list, it might be inappropriate to ask about the FOREACH library, since it hasn't gone into an official release yet. Feel free to point me to the development list. My question is fairly simple: I have a container that only provides const_iterators; it's effectively "read-only" except to a small, select group of classes. Is there a way to tell the BOOST_FOREACH macro to use const_iterator instead of iterator? (Not a huge thing - I can add iterators to the class, or I can teach FOREACH how to iterate over it as described in the docs. I'm just wondering if I'm missing something...) Thanks folks!! ---------------------------------------------------------------------- Dave Steffen, Ph.D. "The only justification for our concepts and Software Engineer IV systems of concepts is that they serve to Numerica Corporation represent the complex of our experiences; ph (970) 419-8343 x27 beyond this they have no legitimacy." fax (970) 223-6797 -- Albert Einstein dgsteffen(a)numerica.us

2 1

[MPL] How to Derive from inherit_linearly generated class?
by Rick Sloan 06 May '05

06 May '05

Hello all, Given a mpl::list typedef mpl::list<structA, structB, structC> typeList; and a wrapper struct template <class T> struct wrap { virtual void someFunction(T&) = 0; virtual ~wrap } You can generate a class by mpl::inherit_linearly<typeList, mpl::inherit<wrap<_2>,_1> >:;type generated; This is pretty much straight from "C++ Template Metaprogramming" pg. 193-194. My question is how do you generate the derived concrete class? I can override any particular function by struct Child : public generated { virtual void someFunction(structA&); } but I can't seem to generate an entire class. I apologize if this is a newbie question, but alas I am a newbie to template metaprogramming. Regards, Rick Sloan

2 1

sorted vectors as associative container?
by Dave Steffen 06 May '05

06 May '05

Hi Folks, Forgive me if this is a stupid question: I'm looking around for an associative-container-implemented-as-a-sorted-vector sort of thing, a la Scott Meyer's "Effective STL" item 23. I could swear Boost had such a thing, maybe called "vec_multiset" or something, a while back. Is there such a thing about? (I've looked at the 'multi-index' containers; they may be what I'm after, but they seem very complicated. My needs are _extremely_ simple...) Thanks! ---------------------------------------------------------------------- Dave Steffen, Ph.D. "The only justification for our concepts and Software Engineer IV systems of concepts is that they serve to Numerica Corporation represent the complex of our experiences; ph (970) 419-8343 x27 beyond this they have no legitimacy." fax (970) 223-6797 -- Albert Einstein dgsteffen(a)numerica.us ___________________ Numerica Disclaimer: This message and any attachments are intended only for the individual or entity to which the message is addressed. It is proprietary and may contain privileged information. If you are neither the intended recipient nor the agent responsible for delivering the message to the intended recipient, you are hereby notified that any review, retransmission, dissemination, or taking of any action in reliance upon, the information in this communication is strictly prohibited, and may be unlawful. If you feel you have received this communication in error, please notify us immediately by returning this Email to the sender and deleting it from your computer.

5 5

The Boost Community
by Fernando Cacciola 05 May '05

05 May '05

Hello people, In 1998, Bemans Dawes wrote the "Proposal for a C++ Library Repository Web Site" (you can find the paper in (www.boost.org/more/index.htm) which lunched the Boost we all know and love today. The focus of the site was and is the production of high-quality peer-reviewed C++ libraries. Over the years, Boost grew as a community of experienced C++ developers with the will and means to share their expertise and work contributing to the highly respected Boost Libraries. Today, Boost is a somewhat recongisable part of the C++ programming community at large. Today, the way I see it, Boost as a community encompases only the boost developers, which are just a fraction of the larger C++ crowd. There are certainly Boost users, but I don't think we can say both developers and users form a community. I think is worth and in-the-spirit to reach out for ALL C++ programmers of ALL LEVELS around an "extended" Boost Community whose main sharing point is not just the submission of high-quality libraries but also the efficient and effective use of the language. A Boost Community site would focus not directly in the Libraries produced by Boost but on C++ itself as the great language we all love and on fostering proper use of it (which includes mixing it with other languages when appropiate). It would be volunteer-driven, just like the Boost libraries are, so the site would be moderated but open for external edition (I think a wiki would work). I envision it having home page links to the fundamental resources for C++ programmers (CUJ,ACCU, "the" books, etc), blogs from developers and users (related to state of the art C++ and/or the boost libraties), important C++ related news (not just boost-library related news), reviews, etc... essentially, everything the will of volunteers can bring on to the community. The motto being "fostering proper use of C++". A minimum of moderation would be required but I assume not much. I know there are other programmer communities around (even C++ focused). The reason I propose this here (around Boost) is becasue I think that a community centered around the Boost Libraries but reaching out could stand out for its quality in the same way the Boost Libraries do. I also know that a newsgroup or mailing list is a form of community, and we have c.l.cpp.m and c.s.cpp. But they lack the verstile structure and dynamism of the today's virtual communities. Also, somehow I believe that centering a community around the Boost Libraries would spread their quality into the rest of the site; to some degree at least. I welcome comments, ctricticism and I call interested people to put forward concrete ideas. Best Fernando Cacciola SciSoft

10 17

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 05 May '05

05 May '05

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Shishir Ramam > > > > [Brian Braatz Writes:] > > This is a good idea. My simple suggestion is to do this only use > > tinyurl. > > Why not use redirect from www.boost.org/wiki to the one above? > > If the description below is accurate, it seems very simple to do - > http://www.sbrady.com/hotsource/html/redirect.html > > -shishir [Brian Braatz Writes:] Good point, a boost.org redirection is much better

1 0

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 05 May '05

05 May '05

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Vladimir Krasovsky > Sent: Thursday, May 05, 2005 1:13 AM > To: David Abrahams > Subject: Re: [Boost-users] Re: The Boost Community > > Hello David, > > Thursday, May 5, 2005, 12:17:14 AM, you wrote: > > > I actually love that. > > > www.gotboost.org > > By the way this domain is already used by someone. Can it be a > problem? > [Brian Braatz Writes:] No problem, we just organize an event where everyone on the list shows up at this guy's house at three in the morning politely asking for the domain name. ;)

1 0

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 05 May '05

05 May '05

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of David Abrahams > Sent: Wednesday, May 04, 2005 7:14 PM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: The Boost Community > > "Brian Braatz" <brianb(a)rmtg.com> writes: > > > > > My suggestion- is add to the Wiki site- something that lets a > > person browse the MAILING LIST. > [David Abrahams Writes:] > That's easily done; we can just link to the mailing list archives. We > have many choices there, actually. Just look at > http://www.boost.org/more/mailing_lists.htm and search for "archives." > > > If you can get the interface to the list working as good or better > > than the email version, then more people will start to use the web > > tool, by using the web tool, they will see the Wiki on the way > > in. In this way, the Wiki is put closer to their paradigm. > [David Abrahams Writes:] > Hmm, I'm not sure we can convince everyone to look at the Wiki before > going to the archive pages. > [Brian Braatz Writes:] Yes, it was a weak idea. But generally, the more we can connect the wiki to the mailing list the better. This basic idea hinges upon somehow providing a "really cool and compelling" way to dig around in the archives other than just searching for keywords. When I click on news://news.gmane.org/gmane.comp.lib.boost.user It opens up Outlook Express and gives me a pretty good view of the list. (much better than the Outlook folder I use now for the list) We should be able to find a WEB TOOL that gives the same view. The idea is if you make that web tool a first class citizen and let me hit the page for digging around in the subjects of the list, and additionally "advertise" the wiki in the header of said tool's page, then you have connected the paradigms. :)

1 0

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 05 May '05

05 May '05

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of David Abrahams > Sent: Wednesday, May 04, 2005 1:15 PM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: The Boost Community > > "JOAQUIN LOPEZ MU?Z" <joaquin(a)tid.es> writes: > > 2. Hosting. Dave Abrahams volunteered to host the site. > > Is he still holding his offer? > > I am. It might require getting new hosting resources soon, but at > least initially I'm sure boost-consulting.com can handle the load. [Brian Braatz Writes:] Suggestion: Setup a PayPal link to support the site. If you do, I will be one of the first to donate $$$$.

1 0

Serialization, shared_ptr and namespaces
by Jan Eickmann 05 May '05

05 May '05

Hi, I'm trying to serialize a list of shared_ptr's to objects. Here's a summary of the code I have: namespace MyNamespace { class B { //implementation plus standard serialization function }; class A { private: list< shared_ptr<B> > myBs; public: template<class Archive> void serialize(Archive & ar, const unsigned int version) { ar & BOOST_SERIALIZATION_NVP(myBs); } }; } If I just try to serialize it, I get an unregistered_class exception. So I think I have to use BOOST_SHARED_POINTER_EXPORT(_GUID) to register the shared_ptr. When I include this inside the namespace (please note that the error messages are from my real code, so they have slightly other names: BOOST_SHARED_POINTER_EXPORT(B) I get this: f:\xenocide\xenocide\src\common\simulation\simulation.h(108) : error C2059: syntax error : 'string' f:\xenocide\xenocide\src\common\simulation\simulation.h(108) : error C2146: syntax error : missing ';' before identifier 'BOOST_CLASS_EXPORT_GUID' f:\xenocide\xenocide\src\common\simulation\simulation.h(108) : error C2059: syntax error : 'string' f:\xenocide\xenocide\src\common\simulation\simulation.h(109) : error C2143: syntax error : missing ';' before '}' When I put it outside the namespace I have to use a fully qualified name including the namespace for class B, so it becomes BOOST_SHARED_POINTER_EXPORT(MyNamespace::B) Then I get (not Xenocide is here what MyNamespace is in my example): f:\xenocide\xenocide\src\common\simulation\simulation.h(115) : error C2653: '__shared_ptr_Xenocide' : is not a class or namespace name f:\xenocide\xenocide\src\common\simulation\simulation.h(115) : error C2653: '__shared_ptr_Xenocide' : is not a class or namespace name I also tried various combinations with BOOST_SHARED_POINTER_EXPORT_GUID but it didn't help. Can someone explain to me how I have to declare shared_ptr's in order for them to work with boost::serialization? I was unable to find some more indepth documentation. Thanks, Jan Eickmann

4 5

Re: [Boost-users] Compatibility between shared_ptr and Rational RoseRT
by Ross Manges 05 May '05

05 May '05

> When running the executable under Rational Purify, it reports many FIM > errors (Freeing Invalid Memory), and all of these are again due to the > deletion of objects that were owned by a shared_ptr. Is your program (the code that uses shared_ptrs) running in multiple threads? I am seeing problems with my multithreaded code where the use_count in the shared_ptrs is not accurate and causing all sorts of problems. I'm even seeing core dump backtraces showing problems in the scoped_lock where the shared_ptr use_count is getting updated. It appears that there are a lot of restrictions on how shared_ptrs can be used in a multithreaded program, and I may be breaking some of those restrictions as it is easy to do so. In any case, let me know if you find a solution to your FIM errors. I have a feeling I'm seeing symptoms of the same problem. Thanks.

4 12

Boost build help
by Eyal Susser 05 May '05

05 May '05

Hi, I just downloaded the IOStreams library, soon to be included in Boost. I needt to install it, so that all the includes will reside wherever they should. I have bjam, but I must be doing something wrong. I'm at the tired, frustrated stage where the documentation is too cryptic. Could someone please explain how I use bjam? The docs say bjam -sTOOLS=gcc my_target. What is "my_target"? Tried teh path to the Jamfile, got an error saying no suitable files can be found...please help, I'm really having a bad day here :( Eyal.

3 2

Re: The Boost Community
by Stephen W. Carson 05 May '05

05 May '05

If this discussion has got you thinking of something you'd like to see on a boost community site, why not slap it up on the existing Boost Wiki site? Put it up even if it is just a quick outline of what you would like to see. That way even if there is eventually a GotBoost site or whatever, the material is out there and can be moved to the new site. Do it now while it is on your mind: http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl?BOOST_WIKI Another thought in regards to marketing Boost Wiki (and the future site). Put the URL in your SIG when posting to boost/C++ lists and newsgroups, (see mine below, Boost Wiki desparately needs a shorter, prettier URL). Simple, cheap marketing that will quickly raise the awareness of the site! -- Stephen W. Carson, Software Engineer DirXion Software in St. Louis, MO (636) 717-2367 http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl?BOOST_WIKI

1 0

boost freebsd port compilation fails
by Maurizio Molina 05 May '05

05 May '05

Hi, I'm trying to compile the boost freeBSD port on my freeBSD machine: root@pippo/usr/ports/devel/boost > uname -a FreeBSD pippo 4.10-RELEASE-p2 FreeBSD 4.10-RELEASE-p2 #5: Thu Sep 16 12:25:56 BST 2004 root@pippo:/usr/obj/usr/src/sys/TESTBED i386 but I get the following compilation error: cd /usr/ports/devel/boost/work/boost_1_32_0 && /usr/bin/env GCC="cc -O2 -pipe " GXX="c++ -O 2 -pipe " STLPORT_ROOT=/ SHELL=/bin/sh PORTOBJFORMAT=elf PREFIX=/usr/local LOCALBASE=/usr/ local X11BASE=/usr/X11R6 MOTIFLIB="-L/usr/X11R6/lib -lXm -lXp" LIBDIR="/usr/lib" CFLAGS="-O2 -pipe " CXXFLAGS=" -O2 -pipe " MANPREFIX="/usr/local" BSD_INSTALL_PROGRAM="install -c -s -o r oot -g wheel -m 555" BSD_INSTALL_SCRIPT="install -c -o root -g wheel -m 555" BSD_INSTALL_DA TA="install -c -o root -g wheel -m 444" BSD_INSTALL_MAN="install -c -o root -g wheel -m 444" /usr/ports/devel/boost/work/boost_1_32_0/tools/build/jam_src/bin.freebsd/bjam --prefix=/usr /local "-sTOOLS=gcc-stlport" "-sBUILD=<stlport-iostream>on release <threading>single <optimiz ation>off" "-sSTLPORT_INCLUDE_DIRECTORY=/usr/local/include/stlport" "-sSTLPORT_LIB_DIRECTORY= /usr/local/lib" Segmentation fault (core dumped) *** Error code 139 Stop in /usr/ports/devel/boost. This is the gcc version I'm using: root@pippo/usr/ports/devel/boost > gcc -v Using builtin specs. gcc version 2.95.4 20020320 [FreeBSD] Any idea of what's going wrong? Thanks, Maurizio

1 0

Exception when reading from an archive
by Eyal Susser 05 May '05

05 May '05

Hi, I'm using the serialization library, but have encountered problems. I have an archiver that I use with a stringstream to serialize objects. I write some ints to the archive using "ar << my_int", and finally I archive a derived class wi ar << derived. So far, so good. In a different process, I create an input archiver, read the integers, and then pass the archiver to a different method, where I create an instance of the class I serialized and use "ar >> instance". This throws an exception. The code for the object created is one *.so, and teh code that does the creating does it with a templated class factory in a different *.so. Is that why I'm gettin the exceptions? Is there a solution? The exception is boost::archive::archive_exception what(): stream error. Thanks.

2 1

RE: [Boost-users] The Boost Community
by Brian Braatz 05 May '05

05 May '05

Jeff, Please let me offer a suggestion- I like the boost wiki- I think the problem is it is "another paradigm" separate from the list. People can only handle so many paradigms. Now, if you "connect the two", then the Wiki will get more people using it. My suggestion- is add to the Wiki site- something that lets a person browse the MAILING LIST. If you can get the interface to the list working as good or better than the email version, then more people will start to use the web tool, by using the web tool, they will see the Wiki on the way in. In this way, the Wiki is put closer to their paradigm. While I think my suggestion is a good idea, I must give the credit to KMART for putting the display of "canned meat" right in front of the bathroom entrance. :) > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Jeff Garland > Sent: Wednesday, May 04, 2005 1:10 AM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] The Boost Community > > I'll just respond in general to the thread. It's great to see some > enthusiasm > from the boost user community. I'm afraid to say, however, that I'm a bit > of > a skeptic. Having started the Boost Wiki in 2001 after a burst of similar > discussion and the initial creation of the Boost-user mailing list, I > consider > the wiki to be basically a failed experiment. That's not to say that the > Boost_User Wiki hasn't been very useful to many people (especially Boost > developers). But I'd say that there has never been sustained "user" > support to > help grow the Wiki and make it more useful for Boost users. The initial > vision was that users would contribute 'tutorial' and other useful > information. It's been sparse at best. > > So my thought is that before we go and grow another technological solution > and > diluting our resources even more, we need to really think about what the > 'users' want to achieve and why the current site doesn't meet those needs. > Why > don't more "users" contribute to the current Wiki? What is it that users > really want to contribute to Boost? Why not build on some of the good > things > on the current wiki (like the 'effective xyz' pages? Why not band > together > and help work on the current Wiki? If it's additions to docs there are > certainly many ways that can be done currently and is done all the time by > a > variety of users. It's alot of work to do it well I believe -- more than > I > can do. And if we want to augment with other technologies we can -- I'm > fully > open to expanding the hosting I already provide. > > As for the wiki spam problem, I believe it to be a non-issue at this > point. > The problem really started to take off in mid-2004, but a series of > measures > have been implemented to reduce and resolve the problem. Yes, spam still > appears from time to time, but it is quickly and completely reversed. The > spammers that get thru initially get blocked out because the new content > filter stops them from posting links to spammer sites after an admin > update. > The more organized and dangerous spammers seem to have moved on to easier > targets. And at this point, any open content system is going to have to > face > this problem -- so be prepared. > > Jeff > > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

2 1

Re: [Boost-users] Exception when reading from an archive
by kauai 05 May '05

05 May '05

--- Eyal Susser <eyal.susser(a)gmail.com> wrote: > Hi, > I'm using the serialization library, but have encountered > problems. > > I have an archiver that I use with a stringstream to > serialize > objects. I write some ints to the archive using "ar << > my_int", and > finally I archive a derived class wi ar << derived. So > far, so good. > > In a different process, I create an input archiver, read > the integers, > and then pass the archiver to a different method, where I > create an > instance of the class I serialized and use "ar >> > instance". This > throws an exception. > > The code for the object created is one *.so, and teh code > that does > the creating does it with a templated class factory in a > different > *.so. Is that why I'm gettin the exceptions? Is there a > solution? > > The exception is boost::archive::archive_exception > what(): stream error. > > Thanks. > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > Are you registering all your classes you are serializing? Your input archiver may not know about all class types in your archive. Use the macro BOOST_CLASS_EXPORT(ClassName) to export a GUID for every class. This should be in a .cpp file outside of any functions. include <boost/serialization/export.hpp> to use it. This include should be placed after any archiver includes. Also within your class (if it is a container for other classes) you will need to do this in your serialize function: ar.register_type(static_cast<ClassName*>(NULL)); // if they are pointers Have fun, -K __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

1 0

smart 'memory' ?
by Gottlob Frege 04 May '05

04 May '05

I find sometimes the journey is as entertaining as the destination, so let me indulge with how I got to what I'm thinking... So, I'm looking at some code. there is an object like FooManager, that manages a list<Foo *>. It has functions like RemoveFoo(Foo *), AddFoo(Foo *), etc. (Ah, so many thoughts - why not just "Remove", instead of "RemoveFoo", since we know the context, why even wrap the list at all (this one is easy - the FooManager does more than just wrap a list), etc). Anyhow, RemoveFoo(Foo *) linearly searches the list for the matching Foo * (by pointer comparison). 1. So I think, hey, really, this should be a set, not a list. And by looking at what else we do with the list, yeah, a set would be better. 2. And really, shouldn't it be a list of smart_ptr<Foo> or somesuch? (probably intrusive_ptr, as Foo has its own ref-counting, but whatever.) 3. And what about this pointer comparison? 4. etc. Just so much rewriting could be done. How do you decide where to stop! So, about the pointer comparison: Since an allocator can (and, surprisingly to some, often does) reuse the same pointer that was previously allocated and freed, how do I even know that the Foo * that should be removed, is really the "same" Foo * that is in the list. (Well, in my case, if I assume decent programming, and since the Foo's are ref-counted, no one should have a pointer to a deleted Foo, but that is assuming decent programming). So, my first thought is that each Foo should have a unique, non-recycled ID (GUID, ever incrementing int/long/etc, or something). Then I would know if I _really_ had the same Foo. And so I could do that. But we are now actually at the point where I will probably skip the whole idea and assume decent programming exists and there is no real problem. Then, I started wondering - I've had this problem before, thus where's the general solution. And what would it be? And *should* it be (ie should it exist)? So that's my question. What would it be. Would it be an allocator that included space for, and initialized, the ID, and then also special pointer type, where the pointer comparison checked the ID? Is it just dumb to try to be robust against bad programming (to this extent)? Would the Object ID be useful beyond that - ie to distinquish between: 1. these 2 objects are equal and 2. these 2 objects are the same object (ie there is only 1 object, not 2) Thoughts?

1 0

RE: [Boost-users] The Boost Community
by Brian Braatz 04 May '05

04 May '05

Excellent ideas. I hope this "happens" and is not another good idea that never gets implemented. :) > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Foster, Gareth > Sent: Wednesday, May 04, 2005 1:41 AM > To: boost-users(a)lists.boost.org > Subject: RE: [Boost-users] The Boost Community > > Hello all, > > As a C++ developer, and a developer who is coming to appreciate the > usefulness of Boost, I think this is a great idea. > > There isn't really a focal point on the web for C++ developers, at least > not > in the sense that Java developers have in java.sun.com, Mono developers > have > in go-mono.net, and as menitoned by somebody else, PHP developers have in > php.net. Boost certainly has enough clout to make this happen. > > The idea of having the following C++/Boost oriented things on one site is > fantastic. > > * blogs for people with influence over the development of the C++ standard > (think planet), > * blogs for heavy boost contributors (think planet), > * documentation for C++/Boost with user anotations + > * forum > * news section feat. latest compiler news and c++/boost developments etc > > + see recent work murray cumming did in adding anotation facilities > to gtkmm docs > > I am sure there are more ideas, but that summarises the best of them I > hope. > > Useful links that might help get this off the ground. > > - http://www.planetplanet.org/ > // will help with blogs > - http://oswd.org > // will help with site design > - http://www.murrayc.com/temp/gtkmm_book_with_comments/html/ch13s02.php > // murray's annotation effort > > It would be nice to get the site looking good right from the outset, > whilst > a "Name Competition" might not merrit a two month wait, I think perhaps a > site design competition would. Look at sites like the Firefox page, a real > central C++/Boost resource looking that professional would have to be good > for C++ as a language. > > I hope my comments and thoughts add something to the discussion. > > All the best, > > Gaz > > > -----Original Message----- > > From: Fernando Cacciola [mailto:fernando_cacciola@hotmail.com] > > Sent: 03 May 2005 17:42 > > To: boost-users(a)lists.boost.org > > Subject: [Boost-users] The Boost Community > > > > Hello people, > > > > In 1998, Bemans Dawes wrote the "Proposal for a C++ Library > > Repository Web > > Site" (you can find the paper in > > (www.boost.org/more/index.htm) which > > lunched the Boost we all know and love today. > > The focus of the site was and is the production of high-quality > > peer-reviewed C++ libraries. > > Over the years, Boost grew as a community of experienced C++ > > developers with > > the will and means to share their expertise and work > > contributing to the > > highly respected Boost Libraries. Today, Boost is a somewhat > > recongisable > > part of the C++ programming community at large. > > > > Today, the way I see it, Boost as a community encompases only > > the boost > > developers, which are just a fraction of the larger C++ > > crowd. There are > > certainly Boost users, but I don't think we can say both > > developers and > > users form a community. I think is worth and in-the-spirit to > > reach out for > > ALL C++ programmers of ALL LEVELS around an "extended" Boost > > Community whose > > main sharing point is not just the submission of high-quality > > libraries but > > also the efficient and effective use of the language. > > > > A Boost Community site would focus not directly in the > > Libraries produced by > > Boost but on C++ itself as the great language we all love and > > on fostering > > proper use of it (which includes mixing it with other languages when > > appropiate). > > It would be volunteer-driven, just like the Boost libraries > > are, so the site > > would be moderated but open for external edition (I think a > > wiki would > > work). > > I envision it having home page links to the fundamental > > resources for C++ > > programmers (CUJ,ACCU, "the" books, etc), blogs from > > developers and users > > (related to state of the art C++ and/or the boost libraties), > > important C++ > > related news (not just boost-library related news), reviews, etc... > > essentially, everything the will of volunteers can bring on to the > > community. The motto being "fostering proper use of C++". > > A minimum of moderation would be required but I assume not much. > > > > I know there are other programmer communities around (even > > C++ focused). The > > reason I propose this here (around Boost) is becasue I think that a > > community centered around the Boost Libraries but reaching > > out could stand > > out for its quality in the same way the Boost Libraries do. > > I also know that a newsgroup or mailing list is a form of > > community, and we > > have c.l.cpp.m and c.s.cpp. But they lack the verstile structure and > > dynamism of the today's virtual communities. Also, somehow I > > believe that > > centering a community around the Boost Libraries would spread > > their quality > > into the rest of the site; to some degree at least. > > > > I welcome comments, ctricticism and I call interested people > > to put forward > > concrete ideas. > > > > Best > > > > Fernando Cacciola > > SciSoft > > > > > > > > _______________________________________________ > > Boost-users mailing list > > Boost-users(a)lists.boost.org > > http://lists.boost.org/mailman/listinfo.cgi/boost-users > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

RE: [Boost-users] Re: The Boost Community
by Brian Braatz 04 May '05

04 May '05

> -----Original Message----- > Sent: Tuesday, May 03, 2005 6:29 PM > > "Eric Goebelbecker" <egoebelbecker(a)gmail.com> escribió en el mensaje > news:7611ef5a05050311052664ea@mail.gmail.com... > > I'm new to Boost and have been lurking on the list for a short time. > > > > This sounds like a good idea to me. Although I mean lean toward a > > portal type site, only because wikis have been compromised frequently > > of late, and a volunteer site doesn't need the additional headaches. > > > > I may be able to help out with web hosting space. I have an "extra" > > account and a low-budget hosting company. > > > > Best, > > Fernando Cacciola > SciSoft [Brian Braatz Writes:] #1 GO GO GO ! This is an excellent idea #2 I suggest if you cannot get a donated host, that you setup a PayPal link on the site. This would allow people to donate. I for one would donate money to keep the site running.

1 0

RE: [Boost-users] The Boost Community
by Foster, Gareth 04 May '05

04 May '05

I might be wrong here, but I think the gist of the discussion was that coming up with something bigger than just a documentation Wiki was the way forward. I am new to boost, I had no idea there was a Wiki. A site with documentation as well as blogs and news on C++ and Boost would pull more traffic than a Wiki alone, in doing so it would hopefully become established, this would in turn drive the expansion of the Wiki, leading to a general trend towards the resource being indispensible for C++ developers. Ideally, you want the situation where C++ student of one year, A, upon being asked by C++ student of 1 month, B, where he can find std::cout documentation on line, will instantly point to one place, as he would with PHP etc. I'll admit, as soon as I saw the word "skeptic", the enormity of the task dawned on me, maybe it is more than a little pie in the sky, then again, it would be great if successful. The anoying part of this is surely that all the tools to build a great community site are always there, its just the theming and pooling of the technologies that is tough. Gaz > -----Original Message----- > From: Jeff Garland [mailto:jeff@crystalclearsoftware.com] > Sent: 04 May 2005 09:10 > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] The Boost Community > > I'll just respond in general to the thread. It's great to > see some enthusiasm > from the boost user community. I'm afraid to say, however, > that I'm a bit of > a skeptic. Having started the Boost Wiki in 2001 after a > burst of similar > discussion and the initial creation of the Boost-user mailing > list, I consider > the wiki to be basically a failed experiment. That's not to > say that the > Boost_User Wiki hasn't been very useful to many people > (especially Boost > developers). But I'd say that there has never been sustained > "user" support to > help grow the Wiki and make it more useful for Boost users. > The initial > vision was that users would contribute 'tutorial' and other useful > information. It's been sparse at best. > > So my thought is that before we go and grow another > technological solution and > diluting our resources even more, we need to really think > about what the > 'users' want to achieve and why the current site doesn't meet > those needs. Why > don't more "users" contribute to the current Wiki? What is > it that users > really want to contribute to Boost? Why not build on some of > the good things > on the current wiki (like the 'effective xyz' pages? Why not > band together > and help work on the current Wiki? If it's additions to docs there are > certainly many ways that can be done currently and is done > all the time by a > variety of users. It's alot of work to do it well I believe > -- more than I > can do. And if we want to augment with other technologies we > can -- I'm fully > open to expanding the hosting I already provide. > > As for the wiki spam problem, I believe it to be a non-issue > at this point. > The problem really started to take off in mid-2004, but a > series of measures > have been implemented to reduce and resolve the problem. Yes, > spam still > appears from time to time, but it is quickly and completely > reversed. The > spammers that get thru initially get blocked out because the > new content > filter stops them from posting links to spammer sites after > an admin update. > The more organized and dangerous spammers seem to have moved > on to easier > targets. And at this point, any open content system is going > to have to face > this problem -- so be prepared. > > Jeff > > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

3 2

RE: [Boost-users] genetic programming
by Lyon, Richard 04 May '05

04 May '05

Hi, Are you talking about genetic programming or optimisation using genetic algorithms? There is a big difference. There are some pretty good c++ libraries available for genetic algorithms. Galib is one that springs to mind. Solving graph problems using genetic algorithms is a bit hit and miss. Some randomising techniques have been used to solve things like mst efficiently. These are different to a genetic algorithm. As I understand it genetic programming is the concept of using genetic algorithms to write code which will solve an optimisation problem. Rather a difficult problem. Well that enough bignoting, thanks ... Richard -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Martin Jynge Sent: Wednesday, 4 May 2005 10:31 PM To: boost-users(a)lists.boost.org Subject: [Boost-users] genetic programming Hi. Ayone interested in gp using boost?

2 1

serialized memory allocation using boost::pool
by Davi de Castro Reis 04 May '05

04 May '05

Hi, I am trying to improve the performance of a std::list instance in a multi-thread application using boost::pool, but the syncronization overhead is killing me. The point is that all access to my std::list are serialized, so I would like my pool to not use mutexes when retrieving memory in the pool, which is what I suppose is happening (got evidence through valgrind). From the template parameters of boost pool, I think this is possible, but I couldn't figure out how to do this. Any tips here? pseudo code: struct X; std::list<X *, boost::fast_pool_allocator<X *> > l; lock(); //fill l with 1000 X * instances unlock(); lock(); for (uint32 i = 0; i < 1000; ++i) { X *x = l.front(); l.erase(l.begin()); l.push_back(x); } unlock(); Thanks in advance, Davi de Castro Reis

1 0

Boost.Test and Visual Studio .Net 2003
by Robert Cober 04 May '05

04 May '05

I am trying to learn how to use boost.test. I have a solution with two projects in it: MyApp - this is the application I am building Test - this is the project for the test framework. Everything builds ok. Simple tests run fine, errors can be clicked on, all is well. My question has to do with how the Test project can refer to the code within MyApp. If I want to refer a class or function defined in MyApp, do I have to manually add the appropriate .h and .obj files to Test? Is there any way to have Test be able to automatically refer to the files of other projects within this workspace? Thanks in advance for any help......

3 2

Re: [Boost-users] genetic programming
by Martin Jynge 04 May '05

04 May '05

Quoting steve ahlgren <steveahlgren(a)gmail.com>: > I've messed around with genetic optimization routines and have > applied > them to a number of different applications with excellent success. I > would love to see a "real" toolkit available in something like boost. It shure would be fancy. > What algorithms have you implemented? Well, the old function regression was my first implementation, but since then the code has changed a lot. I'm currently working on the Serengeti-problem (with lions and a gazelle, you know). It's not running yet, because of design dilemmas.. urk.. There are two things I have difficulties with: first i'd like to put values in the edges. But my implementation uses different value types in one graph, like a double and a vector of doubles. How does one implement that in boost? Second, it seems that I have to use indices with each node, in order to put the right edges in the right slots. So what kind of container should be used? > I'll have to dig through my old > code to see what I have lying around. I'd love to see what you've got. > Great idea; I know! regards, Martin Jynge

1 0

genetic programming
by Martin Jynge 04 May '05

04 May '05

Hi. Ayone interested in gp using boost? I've seen a few earlier posts on trees, but my main interest is a true graph approach.. and it's a little difficult implementing nodes having several outputs (as opposed to a single one i tree-gp). Looping is also a concern as you might guess. I don't have any usable code yet as I'm still learning boost, but I will soon try to port my non-boost code to boost (likely with some help, though) Any input would be highly appreciated :) regards, Martin Jynge --

1 0

RE: [Boost-users] Re: [boost::fsm] Accessing most recently posted event
by Marcin Tustin 04 May '05

04 May '05

> -----Original Message----- > From: Andreas Huber [mailto:ahd6974-spamgroupstrap@yahoo.com] > Sent: 04 May 2005 12:06 > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: [boost::fsm] Accessing most > recently posted event > > Marcin Tustin <mt <at> it-innovation.soton.ac.uk> writes: > > Sounds pretty good. Apart from ahving a member instead > of a function, > > and not bothering to clear the variable, this is pretty what I have > > done. Of course, I am still jury-rigging things for a > technology demo > > at this stage, rather than building a bulletproof library (which I > > leave for later, or better, other people). > > I see. Anyway, I've added triggering_event() to the to-do list... > I'll probably have something more bulletproof in about a month. If you haven't done this by then, I'll submit my patches to the boost list.

1 0

RE: [Boost-users] Re: [boost::fsm] Accessing most recently posted event
by Marcin Tustin 04 May '05

04 May '05

> -----Original Message----- > From: Andreas Huber [mailto:ahd6974-spamgroupstrap@yahoo.com] > Sent: 03 May 2005 14:11 > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Re: [boost::fsm] Accessing most > recently posted event > > Hi Marcin > const event_base * current_event() const > > If the state entry was triggered by an event, current_event() > returns a pointer to the event that triggered the transition. > Otherwise, it returns 0 (e.g. the state was entered as a > result of calling state_machine::initiate(). > We could also name this triggering_event()... > > Thoughts? Sounds pretty good. Apart from ahving a member instead of a function, and not bothering to clear the variable, this is pretty what I have done. Of course, I am still jury-rigging things for a technology demo at this stage, rather than building a bulletproof library (which I leave for later, or better, other people).

2 1

Regex & Autotools
by Foster, Gareth 04 May '05

04 May '05

Hello, Is there a sample autotooled project somewhere that would allow proper linking to Regex? Despite my best efforts with the macro provided for autoconf, I couldn't get my project to link. A sample working project is always the best way to go. Cheers, Gaz ********************************************** * http - http://planetearthworm.com/ * * mail home - earthworm(a)planetearthworm.com * * mail work - gareth.foster(a)siemens.com * **********************************************

2 1

RE: [Boost-users] The Boost Community
by Foster, Gareth 04 May '05

04 May '05

Hello all, As a C++ developer, and a developer who is coming to appreciate the usefulness of Boost, I think this is a great idea. There isn't really a focal point on the web for C++ developers, at least not in the sense that Java developers have in java.sun.com, Mono developers have in go-mono.net, and as menitoned by somebody else, PHP developers have in php.net. Boost certainly has enough clout to make this happen. The idea of having the following C++/Boost oriented things on one site is fantastic. * blogs for people with influence over the development of the C++ standard (think planet), * blogs for heavy boost contributors (think planet), * documentation for C++/Boost with user anotations + * forum * news section feat. latest compiler news and c++/boost developments etc + see recent work murray cumming did in adding anotation facilities to gtkmm docs I am sure there are more ideas, but that summarises the best of them I hope. Useful links that might help get this off the ground. - http://www.planetplanet.org/ // will help with blogs - http://oswd.org // will help with site design - http://www.murrayc.com/temp/gtkmm_book_with_comments/html/ch13s02.php // murray's annotation effort It would be nice to get the site looking good right from the outset, whilst a "Name Competition" might not merrit a two month wait, I think perhaps a site design competition would. Look at sites like the Firefox page, a real central C++/Boost resource looking that professional would have to be good for C++ as a language. I hope my comments and thoughts add something to the discussion. All the best, Gaz > -----Original Message----- > From: Fernando Cacciola [mailto:fernando_cacciola@hotmail.com] > Sent: 03 May 2005 17:42 > To: boost-users(a)lists.boost.org > Subject: [Boost-users] The Boost Community > > Hello people, > > In 1998, Bemans Dawes wrote the "Proposal for a C++ Library > Repository Web > Site" (you can find the paper in > (www.boost.org/more/index.htm) which > lunched the Boost we all know and love today. > The focus of the site was and is the production of high-quality > peer-reviewed C++ libraries. > Over the years, Boost grew as a community of experienced C++ > developers with > the will and means to share their expertise and work > contributing to the > highly respected Boost Libraries. Today, Boost is a somewhat > recongisable > part of the C++ programming community at large. > > Today, the way I see it, Boost as a community encompases only > the boost > developers, which are just a fraction of the larger C++ > crowd. There are > certainly Boost users, but I don't think we can say both > developers and > users form a community. I think is worth and in-the-spirit to > reach out for > ALL C++ programmers of ALL LEVELS around an "extended" Boost > Community whose > main sharing point is not just the submission of high-quality > libraries but > also the efficient and effective use of the language. > > A Boost Community site would focus not directly in the > Libraries produced by > Boost but on C++ itself as the great language we all love and > on fostering > proper use of it (which includes mixing it with other languages when > appropiate). > It would be volunteer-driven, just like the Boost libraries > are, so the site > would be moderated but open for external edition (I think a > wiki would > work). > I envision it having home page links to the fundamental > resources for C++ > programmers (CUJ,ACCU, "the" books, etc), blogs from > developers and users > (related to state of the art C++ and/or the boost libraties), > important C++ > related news (not just boost-library related news), reviews, etc... > essentially, everything the will of volunteers can bring on to the > community. The motto being "fostering proper use of C++". > A minimum of moderation would be required but I assume not much. > > I know there are other programmer communities around (even > C++ focused). The > reason I propose this here (around Boost) is becasue I think that a > community centered around the Boost Libraries but reaching > out could stand > out for its quality in the same way the Boost Libraries do. > I also know that a newsgroup or mailing list is a form of > community, and we > have c.l.cpp.m and c.s.cpp. But they lack the verstile structure and > dynamism of the today's virtual communities. Also, somehow I > believe that > centering a community around the Boost Libraries would spread > their quality > into the rest of the site; to some degree at least. > > I welcome comments, ctricticism and I call interested people > to put forward > concrete ideas. > > Best > > Fernando Cacciola > SciSoft > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0

[boost::fsm] Accessing most recently posted event
by Marcin Tustin 03 May '05

03 May '05

Is there a way that an event triggered on entry to a state can access the event that caused that state to be entered? I'm thinking of adding a line to the process_event method in a wrapper class derived from state_machine that stores the event object received as a way of doing this. 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

config-files with program_options
by Anselm Kunstein 03 May '05

03 May '05

Hi, I'm trying ro read a config file with the program_options library, but I can't figure out how to read several sections in a config file whithout knowing (the exact) section names. For example I have a config file with N [blockn] sections (n=1...N) in my config file but I don't know N: [block1] name="Name section 1" [block2] name="Name2 section 2" . # there might or might not be more sections with the 'name' option . Is there any support for wildcards in option names so I could write something like options_description Opts_Desc("Config-File Options"); Opts_Desc.add_options() ("*.name", value<string>(), "Name of the block"); or is there a workaround for this problem? Thanks a lot in advance, Anselm Kunstein

3 2

[MPL] is lambda<apply<...> >::type wrong?
by Vladim�r Marko 03 May '05

03 May '05

[using boost 1.32.0 with gcc 3.4.2 (mingw)] It seems to me that is_same< lambda<apply<type-list> >::type, apply<type-list> >::type::value is allways true. Is this intended or a bug? The exercise 1-6 in the MPL tutorial, http://www.boost.org/libs/mpl/doc/tutorial/exercises.html , reads "Build a lambda expression that has functionality equivalent to twice. Hint: mpl::apply is a metafunction!" where twice is defined as template <typename F,typename X> struct twice : apply<F,typename apply<F,X>::type> { }; and I think the obvious solution should be apply<_1,apply<_1,_2> > . The problem is that the expression apply< apply<_1,apply<_1,_2> >, add_pointer<_1>,int >::type doesn't compile at all. This is surprising since for example the expression apply< twice<_1,twice<_1,_2> >, add_pointer<_1>,int >::type yields int**** as expected. Is there some fundamental difference between the metafunctions twice and apply? My tests suggest that the compilation failure is caused by the property of lambda<apply<...> >::type described at the top. This seems pretty odd to me since lambda<T>::type is supposed to have a nested template class "apply" if T is a placeholder-expression and lambda<apply<type-list> >::type::apply refers to the boost::mpl::apply even if the type-list contains a placeholder-expression (making the whole apply<type-list> a placeholder-expression). Did I miss something? Thanks, Vladimir Marko

2 1

dijkstra_visitor
by Andreas Fabri 02 May '05

02 May '05

Hello, I wrote my own dijksta visitor for the shortest path algorithm, and on an event I wanted it to write something in a bundled property of a vertex. However this didn't compile because the graph is a const&. Is there a way to solve this problem? Thanks in advance, andreas

2 1

[conversion] RoundEven odd behaviour
by Vyacheslav E. Andrejev 02 May '05

02 May '05

Hello All I tried to use boost numeric conversion library and faced with strange feature of RoundEven that I can not comprehend or find expanation in the documentation. First I thought that this policy rounds float number to nearest integer, because RoundEven::round_style::value == std::round_to_nearest. But! This policy for example returns 2.0 for 3.0 as argument! cout << boost::numeric::converter< int , double , boost::numeric::conversion_traits<int, double> , boost::numeric::silent_overflow_handler , boost::numeric::RoundEven<double> >::convert(2.9) << endl; cout << boost::numeric::converter< int , double , boost::numeric::conversion_traits<int, double> , boost::numeric::silent_overflow_handler , boost::numeric::RoundEven<double> >::convert(3.0) << endl; cout << boost::numeric::converter< int , double , boost::numeric::conversion_traits<int, double> , boost::numeric::silent_overflow_handler , boost::numeric::RoundEven<double> >::convert(3.1) << endl; Output of the program above will be: 3 2 3 What is the cause of such strange behaviour? This is a bug or there is something that I don't understand? -- Vyacheslav E. Andrejev System Architect, Excelsior, LLC

2 1

How to get type given pointer to base?
by Delfin Rojas 02 May '05

02 May '05

Hello all, I hope somebody can help me with this. What I want to do is the following: Let's say I have a templated class called Foo template<class U> class Foo { public: Foo(U * p) : m_p(p) {} ~Foo() {} U* getP() const { return m_p; } private: U* m_p; }; Now let's say I want to have a collection of pointers to Foo. What should I do? The possibilities I can think of are the following: 1) Create a class FooBase and have Foo inherit from FooBase. FooBase is non-templated so there should be no problem with having a collection of pointers to FooBase where I can put instances of Foo. The problem with this approach is that given a FooBase* in that collection I could not use the getP method because I would not know what type of Foo it holds. 2) Create a boost::variant of all the possible types of Foo<U>* I want to have. Then have a collection of this FooVariant. Using static visitation I can obtain the original type of Foo<U>* in a given FooVariant. The problem with this approach is that I get an ICE from VC++ 7.1 when the type of U gets too complex (shared_ptr to another type for example). Have I overlooked something? Is there a better technique for what I want to achieve? Any help would be great, Thanks Delfin Rojas

3 7

searching boost::teebuf : public std::streambuf
by Elmar Loos 01 May '05

01 May '05

Hello, in the archive of this mailinglist I found some Mails from year 2000, that a teebuf derived from std::streambuf should be inserted into boost. I couldn't find anything like this in the current boost version. Could you tell me something about the status of that class, please? Bye Elmar

2 1

Re: [Boost-users] Re: pimpl and boost
by Gottlob Frege 30 Apr '05

30 Apr '05

> Date: Fri, 29 Apr 2005 10:56:14 +0200 > From: David Gruener <gruenedd(a)idmt.fraunhofer.de> > Subject: Re: [Boost-users] Re: pimpl and boost > To: boost-users(a)lists.boost.org > Message-ID: <4271F6AE.20509(a)idmt.fraunhofer.de> > Content-Type: text/plain; charset=us-ascii; format=flowed > > Pavel Antokolsky aka Zigmar wrote: > > auto_ptr - compiled with warning about destroying incomplete type > > As i already mentioned, > very little to no benefit here over plain pointer. > > > shared_ptr - compiled and worked - the destructor of > > Yes, but you still have to provide assignment and copy > constructor for your standard cases. So, as i stated, little > benefit here over plain pointer too and an ideological problem that > shared_ptr used in this way, as serveral times suggested, > isn't a shared pointer anymore. > Again, whats needed is a pointer like Griffiths's grin_ptr > or Peter's impl_ptr. I'm looking forward to the > policy pointers. > > --David > I agree, if/when the need next arises, I will find/write an impl_ptr<>. I will also go with the technique mentioned earlier of ONLY putting the data in the impl_ptr, and not forwarding all the functions. (Still think it should be part of the language though. C++ proclaims to enable data-hiding, but it really doesn't. The speed/data-hiding trade-off that C++ originally made was fine for the times, but now that trade-off isn't so necessary...)

2 1

Re: pimpl and boost
by Peter Dimov 30 Apr '05

30 Apr '05

Kyle Alons wrote: >> I wrote such a pointer several years ago: >> >> http://lists.boost.org/MailArchives/boost/msg21930.php >> http://groups.yahoo.com/group/boost/message/15313 >> http://groups.yahoo.com/group/boost/files/impl_ptr/impl_ptr.html >> >> but I never proposed it for Boost. >> >> The reason that I didn't propose it is that I don't use it. :-) > > Is the source code for this still available anywhere? I get > 'document not found' for the links on > http://groups.yahoo.com/group/boost/files/impl_ptr/impl_ptr.html. > Thanks. It's there: http://groups.yahoo.com/group/boost/files/impl_ptr/

1 0

[Mac OS X] Problem compiling boost
by Daniel Lidström 29 Apr '05

29 Apr '05

Hello, I get linking errrors while trying to compile the program_options. Here is the output: Daniels-Monster:/Users/daniellidstrom/Desktop/boost_1_32_0> bjam --with-program_options stage ...found 468 targets... ...updating 4 targets... gcc-Link-action bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/libboost_program_options-gcc-d -1_32.dylib g++: unrecognized option `-shared' ld: Undefined symbols: _main set -e DYLD_LIBRARY_PATH=bin/boost/libs/program_options/build/ libboost_program_options.dylib/gcc/debug/shared-linkable- true:$DYLD_LIBRARY_PATH export DYLD_LIBRARY_PATH "g++" -g -fPIC -shared -o "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/libboost_program_options-gcc-d -1_32.dylib" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/cmdline.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/config_file.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/options_description.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/parsers.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/variables_map.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/value_semantic.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/positional_options.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/utf8_codecvt_facet.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/convert.o" "bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/winmain.o" ...failed gcc-Link-action bin/boost/libs/program_options/build/libboost_program_options.dylib/ gcc/debug/shared-linkable-true/libboost_program_options-gcc-d -1_32.dylib... ...skipped <@boost!lib>libboost_program_options-gcc-d-1_32.dylib for lack of <@boost!libs!program_options!build/libboost_program_options.dylib/gcc/ debug/shared-linkable-true>libboost_program_options-gcc-d-1_32.dylib... I'm using gcc 3.2 (the one that came with XCode 1.5). Strange thing is the man page says -static is an option for both gcc and g++. How can I successfully compile the dynamic-link library? The static-link library was compiled successfully though. Also, I added -Wno-long-double to the g++ options, to get rid of an annoying warning. The gcc bjam file says there is no gcc or g++ on Darwin, but there certainly is for me. I'm using Mac OS X 10.3.9 with gcc 3.2 from XCode 1.5 installation. -- Daniel SBG AB

2 1

shipping software depending on boost
by Wieger Wesselink 29 Apr '05

29 Apr '05

Hi, I have written some code for a software package that my group at the university is distributing, and it has dependencies on boost. Now my boss is getting very worried about the extra complications this brings for a user to install the software (note that the dependencies are not only header files). The package is currently installed using an autoconf script. I wonder how others deal with this situation. Is it best to ship the complete boost library as part of the installation, and handle the compilation of the used boost libraries by autoconf? Thanks, Wieger -- _______________________________________________ Surf the Web in a faster, safer and easier way: Download Opera 8 at http://www.opera.com Powered by Outblaze

2 1

pimpl and boost
by David Gruener 29 Apr '05

29 Apr '05

Hi, i'm wondering if there is a nice smart pointer in boost to implement PImpl. I found an implementation here: [1]. So, does boost provide a smart pointer which does play nice with pimpl, so that one doesn't need to provide a destructor, copyconstructor and assignment operator for the default cases? --David [1] http://www.octopull.demon.co.uk/arglib/TheGrin.html.

10 17

redefinition of `class boost::algorithm::iterator_range<IteratorT>'
by Michael Garriss 28 Apr '05

28 Apr '05

this program: #include <boost/algorithm/string.hpp> #include <boost/range.hpp> produces this error: In file included from /home/mgarriss/dev/tp-compiled/include/boost-1_32_0/boost/range.hpp:20, from range.cxx:2: /home/mgarriss/dev/tp-compiled/include/boost-1_32_0/boost/range/iterator_range.hpp:79: error: redefinition of `class boost::algorithm::iterator_range<IteratorT>' /home/mgarriss/dev/tp-compiled/include/boost-1_32_0/boost/algorithm/string/iterator_range.hpp:50: error: previous definition of `class boost::algorithm::iterator_range<IteratorT>' /home/mgarriss/dev/tp-compiled/include/boost-1_32_0/boost/range/iterator_range.hpp:380: error: `template<class IteratorT> boost::algorithm::iterator_range<IteratorT> boost::make_iterator_range(IteratorT, IteratorT)' conflicts with previous using declaration `template<class IteratorT> boost::algorithm::iterator_range<IteratorT> boost::algorithm::make_iterator_range(IteratorT, IteratorT)'

2 1

bgl question: in which namespace to put the functions
by Andreas Fabri 28 Apr '05

28 Apr '05

Hello, Developping code so that BGL algorithms can run on the vertices and edges of CGAL triangulations, I am wondering if there are rules or recommendations for in which namespace the functions source, target, vertices, out_edges ... shall go. In boost, because it is boostish code, or in CGAL, because that is the namespace where the triangulation lives in? thank's in advance, andreas

3 3

Re: pimpl and boost
by Gottlob Frege 28 Apr '05

28 Apr '05

> Today's Topics: > > 1. Re: pimpl and boost (Gennadiy Rozental) > 2. Re: Re: Re: pimpl and boost (David Gruener) > 3. Re: Re: Re: pimpl and boost (Jeff Flinn) > > ---------------------------------------------------------------------- > I so wish that pimpl was somehow part of the language. ie you don't worry about or see the pointer at all, nor do you have to forward all the function calls. I don't see why the compiler couldn't implement pimpl for me. (Of course, for cases where you want multiple implementations at run time, etc, you still do that yourself, but the typical case the compiler could do). Maybe: // h file class Foo interface { void GetX(); }; // cpp file: class Foo continue // some keyword { int x; }; Foo::GetX() { return x; } And just like implementing the same function twice, the linker could warn about 2 implementations of the same class. P.S. this could also allow you to keep some members defined in the h file, (and others in the cpp impl) if you wanted the speed of inlines for some variables, or public/protected variables, etc. Just a thought. Maybe I should complain elsewhere...

2 1

Program_options and XML
by Jason Stewart 28 Apr '05

28 Apr '05

Has anyone implemented an XML parser for boost::program_options? I have just started looking into using this library but I need to read/write XML files. Before digging in and doing it myself I thought I'd ask if anyone has looked at the problem. Thanks, Jason Stewart

3 3

BGL::johnson_all_pairs_shortest_path question
by Andreas Fabri 28 Apr '05

28 Apr '05

Hello, What is the non-named parameter version of this algorithm. In the online manual it says Weight (and this is probably a typo, as there is a template parameter WeightMap). Also what are the arguments id1 and zero good for? andreas

1 0

[BGL] check edge existence in adjacency_matrix in constant time?
by Felipe Magno de Almeida 28 Apr '05

28 Apr '05

Sorry if I have overlooked something in the docs, but Im really not finding it... Is there someway to check the existence of an edge in adjacency_matrix in constant time knowing its vertices? An explanation or a pointer to the right place in the docs would be really appreciated. Thanks in advance, -- Felipe Magno de Almeida UIN: 2113442 email: felipe.almeida at ic unicamp br, felipe.m.almeida at gmail com, felipe at synergy com I am a C, modern C++, MFC, ODBC, Windows Services, MAPI developer from synergy, and Computer Science student from State University of Campinas(UNICAMP). To know more about: Unicamp: http://www.ic.unicamp.br Synergy: http://www.synergy.com.br current work: http://www.mintercept.com "There is no dark side of the moon really. Matter of fact it's all dark."

1 1

Re: [Boost-users] Compatibility between shared_ptr and Rational RoseRT
by Dave.Ware＠seleniacomms.com 28 Apr '05

28 Apr '05

Don't worry about having hijacked my original thread. The examples you've been discussing have been very helpful to my code as well. We haven't solved all of our problems yet, but it is much better. Thanks ------------------------------------------------------------ This email and any attached files contains company confidential information which may be legally privileged. It is intended only for the person(s) or entity to which it is addressed and solely for the purposes set forth therein. If you are not the intended recipient or have received this email in error please notify the sender by return, delete it from your system and destroy any local copies. It is strictly forbidden to use the information in this email including any attachment or part thereof including copying, disclosing, distributing, amending or using for any other purpose. In addition the sender excludes all liabilities (whether tortious or common law) for damage or breach arising or related to this email including but not limited to viruses and libel.

1 0

Function pointers, result_of<> and function_traits<>
by Delfin Rojas 27 Apr '05

27 Apr '05

Hello all, I am trying to get the return type of a member function given its signature. Consider the following example from the boost result_of test: struct X {}; int main { using namespace boost; typedef int (X::*mem_func_ptr)(float); BOOST_STATIC_ASSERT((is_same<result_of<mem_func_ptr(X,char)>::type, int>::value)); } This test compiles/runs with no problems. Now, what I want is the return type of a function with signature "int (X::*)()" I added the following to the test: typedef int (X::*empty_mem_func_ptr)(); BOOST_STATIC_ASSERT((is_same<result_of<empty_mem_func_ptr(X)>::type, int>::value)); This does not compile. I get an error boost_1_32_0\boost\utility\result_of.hpp(43) : error C2825: 'F::template': cannot form a qualified name. I tried to do this same thing with function_traits but it seems function traits does not support member function pointers at all. So this is all I have. If somebody can help me I would appreciate it. Thanks -delfin

1 1

const-correctness and lambda expression
by Matthias Kaeppler 27 Apr '05

27 Apr '05

Hi, in my program, I have this function to compute the sum of sizes of a bunch of files: Glib::ustring DirBrowser::get_total_file_size() { using namespace boostfs; // boostfs = boost::filesystem using namespace lambda; // lambda = boost::lambda boost::intmax_t sum = 0; std::for_each(m_first_file,m_files.end(),sum+=bind(&file_size,*_1)); return make_size_human_readable( sum ); } The line with for_each computes the sum of file sizes, by iterating over a container of pointers to boostfs::pathS, starting at the first file (there are also directories, before the files), till the end, invoking the lambda functor on each. Since this method does not change the state of the object, I wanted to declare it const, but suddenly the lambda expression starts complaining: Somehow, those path objects inside the lambda expression become const now, too, and it doesn't compile anymore. How come? -- Matthias Kaeppler

2 1

RE: [Boost-users] Problems with boost::filesystem::path continued
by Johnson, Donald W MVK Contractor 27 Apr '05

27 Apr '05

It seem that due to sheer luck mixing the runtimes was not causing problems. Objects allocated in the dll where always deleted in the dll and objects allocated in the application likewise deallocated in the application. I know because I had forced things to work using BOOST_FILESYSTEM_NO_LIB and manually linking to the library, and the application worked and all the tests ran fine. The application was linked to /MT or /MTd so that was a disaster waiting to happen the first time I called a function across dll with a dynamicly allocated argument that was held and eventually dealocated on the dll side. I was being saved because that particular situation never occurs. Thanks for the help I thought the error message was refering to the runtime that boost itself was linking against, so was trying to find how to make boost link against the dynamic library. Thanks Donald J -----Original Message----- From: John Maddock [mailto:john@johnmaddock.co.uk] Sent: Wednesday, April 27, 2005 4:44 AM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Problems with boost::filesystem::path continued >I found the main problem, I had edited the wrong boost/config/user.hpp >file. > When I set BOOST_FILESYSTEM_DYN_LINK in the correct user.hpp file I >get this error. > > c:\Boost\include\boost-1_31\boost\config\auto_link.hpp(269): fatal > error > C1189: #error : "Mixing a dll boost library with a static runtime is > a really bad idea..." > > #if (defined(_DLL) || defined(_RTLDLL)) && defined(BOOST_DYN_LINK) # > define BOOST_LIB_PREFIX #elif defined(BOOST_DYN_LINK) # error "Mixing > a dll boost library with a static runtime is a really bad idea..." > #else > # define BOOST_LIB_PREFIX "lib" > #endif > > So wher am I supposed to set the _DLL symbol or do I need to disable > auto_link. If so how? The message tells you what you've done wrong: you can not use a dll that passes C++ objects between itself and your application unless you are also using a dll runtime library. If you do manage to do this your code will very likely crash and burn at runtime (due to recourses being allocated and freed by differing C++ runtimes). If you switch your code generation options (in the VC++ IDE) to use a dll runtime then the error message will go away. John.

1 0

minimal perfect hash
by Davi de Castro Reis 27 Apr '05

27 Apr '05

Hi all, I recently found in the archives a discussion about minimal perfect hashing in boost, in the beginning of January (2005). Does someone one have any news about it? I have written myself a F/OSS C library for minimal perfect hashing, and I am willing to contribute for any efforts to make this data structure part of the boost toolkit. Thanks in advance, Davi

2 2

Re: Placeholder Arguments
by Joe McCay 27 Apr '05

27 Apr '05

David Abrahams <dave(a)boost-consulting.com> writes: > Are you asking about runtime lambda expressions as in the Boost Lambda > library, or compile-time lambda expressions as in MPL? > > -- > Dave Abrahams > Boost Consulting > www.boost-consulting.com I was talking about MPL because that is the only one I have studied at the moment. What is the difference between the two outside of the obvious runtime/compile-time difference? I was trying to figure out how a particular placeholder gets its arguments to choose from when deciding the argument (argument 1 for _1 etc.). Thank you. Joseph L. McCay

2 1

Re: Boost Logging Library?
by Jwahar Bammi 27 Apr '05

27 Apr '05

I dont understand the need for the "Proposed Boost logging library". A perfectly good and flexible logging library exists in the form of log4cxx, which is much more flexible than the proposal, has a ton of tools, appenders, is compatible with log4j/log4net etc etc (for those of us who live in multiple worlds). log4cxx is going through a almost complete rewrite, but is still eminently usable. see http://logging.apache.org/log4cxx/ cheers, > > >Vladimir Prus wrote: > > >>Hi Matthias, >> >> >> >> >>>on http://www.torjo.com/ is a "Boost logging library" available for >>>download, but it doesn't exist on the official boost website. >>> >>>Will this library be in a future release of boost, or has someone just >>>abused the well-known name for pushing his work? >>> >>>A logging library would be a very nice thing to have in boost, but >>>before I use this one, I want to make sure it will be available publicly >>>in the official boost packages. >>> >>> >>This library was not reviewed, and I'm not sure if the review was requested. >>So, it's "Proposed Boost logging library", and is not an official part of >>Boost. >> >> >>- Volodya >> >> > >In fact, John Torjo has requested a formal review at boost.devel just >today, see his post at >http://article.gmane.org/gmane.comp.lib.boost.devel/122414 > >- Andreas > > -- Jwahar Bammi bammi(a)memento-inc.com

5 4

Problems with boost::filesystem::path continued
by Johnson, Donald W MVK Contractor 27 Apr '05

27 Apr '05

I found the main problem, I had edited the wrong boost/config/user.hpp file. When I set BOOST_FILESYSTEM_DYN_LINK in the correct user.hpp file I get this error. c:\Boost\include\boost-1_31\boost\config\auto_link.hpp(269): fatal error C1189: #error : "Mixing a dll boost library with a static runtime is a really bad idea..." #if (defined(_DLL) || defined(_RTLDLL)) && defined(BOOST_DYN_LINK) # define BOOST_LIB_PREFIX #elif defined(BOOST_DYN_LINK) # error "Mixing a dll boost library with a static runtime is a really bad idea..." #else # define BOOST_LIB_PREFIX "lib" #endif So wher am I supposed to set the _DLL symbol or do I need to disable auto_link. If so how? Thanks, Donald Johnson

2 1

Need a method to compare types of data objects at runtime
by Stephen Torri 26 Apr '05

26 Apr '05

If I have a abstract class with two classes that implement it how can I make a check that tests for the type of the implementation classes? For example: class Base { public: virtual void run (){} virtual void stop (){} }; class Gas_Pedal : public Base { public: virtual void run (){} void set_rate (unsigned int); private: unsigned int m_rate; }; class Brake_Pedal : public Base { public: virtual void stop (){} void set_pressure (unsigned int); private: unsigned int m_pressure; }; When I have a client that receives a Base object I want to be able to check at run-time. So that if I am expecting a Gas_Pedal object and I receive something else I can perform some recovery action. The software I am designing contains an untold number of arrangements of items. Each item expect a distinctive input type. If it receives something that it does not know how to handle or need I want to throw an error. In order to do this I would like to do something like: void process (Base* sobj) { if (! (isExpectedType(sobj)) { throw NotForMeException (); } Brake_Pedal* bp = dynamic_cast<Brake_Pedal>(sobj); bp.set_pressure (50); } So perhaps my design is flawed that I am running into this situation. I am attempting to keep the interface for the Base class simple. The underlying types could have different data and access methods because they differ so widely in what they will do. This example was given to keep things simple. Stephen

3 2

Problems with boost::filesystem::path
by Johnson, Donald W MVK Contractor 26 Apr '05

26 Apr '05

Boost::filesystem::path is failing to create any paths. If always throws and exception stating invalid path. In the main function I call boost::filesystem::path::default_name_check( boost::filesystem::native ); boost::filesystem::initial_path(); The attempts to construct paths are in the pattern boost::filesystem::path(some_path_name); Specificly the error occurs here in constructing a path void path::m_path_append( const std::string & src, name_check checker ) { // convert backslash to forward slash if checker==native // allow system-specific-root if checker==no_check || checker==native assert( checker ); assert( src.size() == std::strlen( src.c_str() ) ); // no embedded 0 if ( src.size() == 0 ) return; std::string::const_iterator itr( src.begin() ); // [root-filesystem] # ifdef BOOST_WINDOWS if ( (checker == no_check || checker == native) && src.size() >= 2 ) // This fails { ... Since fore some reasion it seemed to be failing to get the native checker I tired this instead boost::filesystem::path(some_path_name,boost::filesystem::native); But the above test still fails. This seems to be related to the filesystem code being used in the main application and a DLL. I added this line to boost/config/user.hpp #define BOOST_FILESYSTEM_DYN_LINK So the library should be dynamicly linked and it being used in two places shouldn't be a problem. If I trace the function in a debugger all calls to the filesystem library look like this > FESMLibDebug.dll!boost::filesystem::path::default_name_check(bool (const std::basic_string<char,std::char_traits<char>,std::allocator<char> > &)* new_check=0x004573e2) > FESMLibDebug.dll!boost::filesystem::path::m_path_append(const std::basic_string<char,std::char_traits<char>,std::allocator<char> > & src={...}, bool (const std::basic_string<char,std::char_traits<char>,std::allocator<char> > &)* checker=0x004573e2) So it doesn't look like im getting 2 copies of boostfile system. I previously was getting calls to $(APPNAME)!boost::filesystems and $(DLLNAME)!boost::filesystems. However, if I incluede the source files of the dll directly into the application the problem dissapears. Anyone know what I missed? This is with Visual C++ 2003 and Windows XP. Thanks Donald Johnson

2 1