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

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