August 2009 - Boost-users - lists.boost.org

[Serialization] unregistered void cast - revisited
by Soumen 05 Aug '09

05 Aug '09

Hi, I'm getting "unregistered void cast" exception while serialization. I've followed the header file order recommended for serialization. i.e. #include <boost/archive/binary_oarchive.hpp> #include <boost/archive/binary_iarchive.hpp> #include <boost/serialization/serialization.hpp> #include <boost/serialization/base_object.hpp> #include <boost/serialization/is_abstract.hpp> //only in files where class is ABC #include <boost/serialization/export.hpp> //Followed by other headers like utility, map, vector, string Do I need to do something else as well? I've few derived classes for which I don't have any data. I've tried exporting those as well. But I still get the error. Regards, ~ Soumen -- View this message in context: http://www.nabble.com/-Serialization--unregistered-void-cast---revisited-tp… Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

[date_time] unexpected behaviour from microsec_clock::local_time on 32 bit system.
by werasm＠gmail.com 05 Aug '09

05 Aug '09

Hi all, I have the following little program (I'm using mingw 3.4.2 (32bit) ): #include <boost/date_time/posix_time/posix_time.hpp> #include <iostream> int main() { using namespace boost; using namespace boost::posix_time; //Time now (since epoch in seconds)... static const ptime secSinceEpoch( second_clock::local_time() ); //Time now (since epoch in micros)... static const ptime microsecSinceEpoch( microsec_clock::local_time() ); //Elapsed should be close to zero... time_duration elapsed = microsecSinceEpoch - secSinceEpoch; std::cout << "elapsed: " << elapsed.total_milliseconds() << std::endl; return 0; } It outputs: "elapsed: -1240428644900", whereas I expected it to output a positive value. It therefore seems that microsec_clock::local_time() wraps. Is 64-bits a pre-requisite for this library to work correctly or is there some remedy? Kind regards, Werner

1 0

time_facet format flags %F and %f and "."
by Mark Stapper 05 Aug '09

05 Aug '09

Hello, I've been using time_facet to create this foolish timestamp I have to send to another program. in doing so i discovered that %f returns the fractional seconds without a leading ".", and %F does this WITH a leading ".". bug or feature? Greetz, Mark

1 0

Spring layout algorithm for animation
by Kaspar Fischer 04 Aug '09

04 Aug '09

Dear list, I am interested in spring layout algorithms and have seen this: http://www.boost.org/doc/libs/1_36_0/libs/graph/doc/kamada_kawai_spring_lay… I would like to show an animation of the process of laying out the graph and wonder whether this is possible with the Kamada-Kawai algorithm, and with the Boost implementation in particular? Or does/will Boost provide another spring layout algorithm that is suitable for animation? I intend to use such an animation in a GUI where the user initially sees a small graph (a center node with neighbours, shown in a circular layout) and can double-click on nodes to add their neighbours -- at this point I want to visually refine the graph (i.e., change its layout) using a spring-based algorithm. Many thanks, Kaspar

1 0

Re: [Boost-users] prim_minimum_spanning_tree
by Line Blander Reinhardt 04 Aug '09

04 Aug '09

Well Prims is not supposed to have problems with negative cycles. However Dijkstra does. I believe the problem lies in prim_minimum_spanning_tree use of Dijkstra. Thanks Line ________________________________________ Fra: me22.ca(a)gmail.com [me22.ca(a)gmail.com] På vegne af Scott McMurray [me22.ca+boost(a)gmail.com] Sendt: 4. august 2009 18:14 Til: Line Blander Reinhardt Emne: Re: [Boost-users] prim_minimum_spanning_tree 2009/8/4 Line Blander Reinhardt <lbre(a)man.dtu.dk>: > Have done thanks. Was just wondering if it had been fixed in newer versions. > I think it fundamentally cannot be fixed, as the algorithm doesn't provide a way to detect the problematic conditions. ~ Scott

1 0

Python Boost Graph Library API
by Claudio Martella 04 Aug '09

04 Aug '09

Hello list, I'm currently trying to implement my program, which makes extensive use of graph operations, with the py bgl module. I haven't found any API in the documentation that relates to the py language bindings. Could you please point me to the API of the python bindings of BGL? Or to the proper mailling list to ask for? TIA /CM -- Claudio Martella claudio.martella(a)gmail.com

2 1

prim_minimum_spanning_tree
by Line Blander Reinhardt 04 Aug '09

04 Aug '09

Hi all Is it really true that prim_minimum_spanning_tree cannot handle negative cycles?? I was surprised to find that when I have negative cycles the boost function crashes in Dijkstra. Has this error been fixed?? Or is it just viewed as a feature?? Thanks Line

2 2

Re: [Boost-users] [boost users][tuple] foreach with tuple
by pete＠pcbartlett.com 04 Aug '09

04 Aug '09

>typedef vector<tuple<int,int> > vec_type; >vec_type l_listLeftLines; >foreach (vec_type::reference i_pair, l_listLeftLines) >{...} It is a bit of a pain to have to resort to the typedef. Because I iterate over std::maps quite a lot, I've semi-seriously toyed with the idea of a submitting a feature request for BOOST_FOREACH_PAIR( std::pair<T1,T2> const& my_pair , my_container_with_a_pair_value_type ) which through a bit of PP_CAT and PP_COMMA somehow does the "right" thing. Given the potential for confusion though, I assumed Eric would reject the request :-).

2 1

iterator_range !is_singular assertion
by Bradley Russell 04 Aug '09

04 Aug '09

We recently upgraded our software to 1.36. Still using Visual Studio 2003. Started getting these ! is_singular assertions. I think I have tracked it down to being NDEBUG. For example if my exe does not have NDEBUG defined and I call a dll that does then the assertion happens. I have found a little bit of information about it online http://www.nabble.com/Bug-in-iterator_range-when-NDEBUG-not-defined-but-boos t-asserts-are-enabled-td17161844.html This link has a proposed patch replacing #ifndef NDEBUG with #ifndef BOOST_DISABLE_ASSERTS in the range/iterator_range.hpp In limited testing this patch seems to fix the issue. I am wondering if this is a good fix or is it something more fundamental. Thanks, Bradley Russell Programmer NPC International (620)231-3390 x 122 ---------------------------------------------------------------------------- -------- NOTICE: This electronic mail message and any attached files are confidential. The information is exclusively for the use of the individual or entity intended as the recipient. If you are not the intended recipient, any use, copying, printing, reviewing, retention, disclosure, distribution or forwarding of the message or any attached file is not authorized and is strictly prohibited. If you have received this electronic mail message in error, please advise the sender by reply electronic mail immediately and permanently delete the original transmission, any attachments, and any copies of this message from your computer system.

1 0

setsockopt SO_RCVTIMEO for udp::socket in boost::asio
by Евгений Шубин 04 Aug '09

04 Aug '09

> Hello. > I tried a simple program with clear sockets: > #include <stdlib.h> > #include <sys/socket.h> > #include <netinet/in.h> > #include <arpa/inet.h> > #include <cstring> > #include <unistd.h> > #include <iostream> > > using namespace std; > > /* > * > */ > int main(int argc, char** argv) > { > if (argc < 2) > { > cout << "Few parameters" << endl; > return 1; > } > > unsigned int taddr = inet_addr(argv[1]); > if(INADDR_NONE == taddr) > { > cout << "wrong address" << endl; > return 1; > > } > > int sock = socket(AF_INET, SOCK_DGRAM, 0); > if( sock < 0 ) > { > cout << "socket error" << endl; > return 1; > } > > sockaddr_in sin; > memset(&sin, 0, sizeof(sin)); > sin.sin_family = AF_INET; > > unsigned short port = 3334; > sin.sin_port = htons(port); > > sin.sin_addr.s_addr = taddr; > int ret = connect( sock, reinterpret_cast<const sockaddr *>(&sin), sizeof(sin) ); > if ( ret < 0 ) > { > cout << "connect error" << endl; > return 1; > } > > timeval tv; > memset( &tv, 0, sizeof(tv) ); > tv.tv_sec = 5; > > ret = setsockopt( sock, SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof(tv) ); > if( ret < 0 ) > { > cout << "sockopt error" << endl; > return 1; > } > > char hello[] = "hello"; > send(sock, hello, sizeof(hello), 0); > > char msg[256]; > cout << "Recived "<< recv(sock, msg, sizeof(msg), 0) << endl; > > close(sock); > > return (EXIT_SUCCESS); > } > > This program just send buffer via connected udp socket and then wait for response. Recv may block only for 5 seconds. All work fine. > When I try to make same with boost.asio recieve method does not return if there is no response. > I use boost 1.37/ > > #include <boost/asio.hpp> > #include <iostream> > #include <vector> > using boost::asio::ip::udp; > using namespace std; > > int > main() > { > try > { > boost::asio::io_service serv; > udp::socket s( > serv, > udp::endpoint( > udp::v4(), > 3333 > ) > ); > > udp::resolver::query query(udp::v4(), "127.0.0.1", "3334"); > udp::resolver resolver(serv); > udp::resolver::iterator iterator = resolver.resolve(query); > s.connect(*iterator); > int sock = s.native(); > cout << sock << endl; > timeval tv; > memset( &tv, 0, sizeof(tv) ); > tv.tv_sec = 5; > cout << setsockopt( sock, SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof(tv) ) << endl; > > string h("hello"); > s.send(boost::asio::buffer(h)); > vector<char> v(12); > cout << s.receive(boost::asio::buffer(v)) << endl; > cout << v.size() << endl; > } > catch(boost::system::system_error &err) > { > cout << err.what() << endl; > } > } > I tested both programs with nc -lu 3334 > ooh! I use Linux. > > >

1 0