November 2007 - Boost-users - lists.preview.boost.org

[multi_array] performance of multi_array
by frederic.bron＠alcan.com 22 Nov '07

22 Nov '07

I have just replaced some vector< vector< vector<T> > > by a multi_array and I have seen a huge time penalty. I looked further in the documentation and discovered the BOOST_DISABLE_ASSERTS macro which did not really improved the speed. Where can I find ALL macros I have to use (for any boost library) to have the best speed (building a release)? Regards, F. Bron Avis : Ce message et toute pièce jointe sont la propriété d'Alcan et sont destinés seulement aux personnes ou à l'entité à qui le message est adressé. Si vous avez reçu ce message par erreur, veuillez le détruire et en aviser l'expéditeur par courriel. Si vous n'êtes pas le destinataire du message, vous n'êtes pas autorisé à utiliser, à copier ou à divulguer le contenu du message ou ses pièces jointes en tout ou en partie. Notice: This message and any attachments are the property of Alcan and are intended solely for the named recipients or entity to whom this message is addressed. If you have received this message in error please inform the sender via e-mail and destroy the message. If you are not the intended recipient you are not allowed to use, copy or disclose the contents or attachments in whole or in part.

1 0

[BGL] Properties Complexity
by David A. Greene 21 Nov '07

21 Nov '07

What's the complexity of accessing a property bundle for an adjacency list? Does it change as the user changes the selector for the vertex list structure (vecS vs. listS, etc.)? I could imagine property access being constant-time for vecS vertex lists, because the vertex space can be represented with a dense set of numbers corresponding to the vertex's position in the vector. The property map can thus be a simple vector or even a C array indexed by vertex number. With listS, on the other hand, it's hard to see how one could get constant-time property map lookup. I have a problem where vertex property lookup happens very frequently. What kind of tradeoffs should I be thinking about? -Dave

1 0

[Boost Interprocess] - Shared Memory File - Can we specify the path to create
by Srinivas Gardas 21 Nov '07

21 Nov '07

Hi, When creating a 'SharedMemory File', is there a way to specify the 'Path' where the file should be created? Right now, it is created by default in the Temp folder local machines. In case of a 'Citrix Machine' I am not able to figure out where this file is created? I'd like specify the path to create the 'shared memory' file so that it is the same irrespective of the user logged into the machine. Any help is appreciated. Thanks & Regards, Srinivas Gardas

2 1

[asio] RS232 serial communications
by Bob Warren 21 Nov '07

21 Nov '07

Has anyone built RS232 support into asio, or can anyone give me some guidelines on how to do this?

1 0

[random] fast uniform integer random distribution
by frederic.bron＠alcan.com 21 Nov '07

21 Nov '07

Hi, I face a speed problem. I need a uniform distribution of integers between 0 and N, N around 10^11. I tried 2 methods with very high speed efficiency difference (time x8). The 1st one is based on uniform_smallint (the slowest method) and the 2nd one is based on a real distribution casted to an integer. Is there a problem with the 2nd method which is much quicker? Fred common code: typedef boost::mt19937 base_generator_type; base_generator_type generator; generator.seed(static_cast<unsigned int>(std::time(0))) ; method 1 (using uniform_smallint): boost::uniform_smallint<uint32_t> uni_dist((uint32_t)0, (uint32_t)10000000000); boost::variate_generator<base_generator_type*, boost::uniform_smallint<uint32_t> > uni(&generator, uni_dist); for (unsigned long k=0 ; k<100000000UL; ++k) uni() ; method 2 (using uniform_real and casting): boost::uniform_real<double> uni_dist(0., 100000000001.); boost::variate_generator<base_generator_type*, boost::uniform_real<double> > uni(&generator, uni_dist); for (unsigned long k=0 ; k<100000000UL; ++k) (uint32_t)uni() ; F. Bron ___________________________________ Frédéric Bron (frederic.bron(a)alcan.com) Unité ALT, Alcan CRV, BP 27, 38341 Voreppe téléphone : +33 4 76 57 81 72 télécopie : +33 4 76 57 80 99 Avis : Ce message et toute pièce jointe sont la propriété d'Alcan et sont destinés seulement aux personnes ou à l'entité à qui le message est adressé. Si vous avez reçu ce message par erreur, veuillez le détruire et en aviser l'expéditeur par courriel. Si vous n'êtes pas le destinataire du message, vous n'êtes pas autorisé à utiliser, à copier ou à divulguer le contenu du message ou ses pièces jointes en tout ou en partie. Notice: This message and any attachments are the property of Alcan and are intended solely for the named recipients or entity to whom this message is addressed. If you have received this message in error please inform the sender via e-mail and destroy the message. If you are not the intended recipient you are not allowed to use, copy or disclose the contents or attachments in whole or in part.

1 0

[boost] [interprocess] offset_ptr to volatile int?!
by Lars Hagström 21 Nov '07

21 Nov '07

Hi, I'm trying to use a offset_ptr<volatile int> to point to a piece of shared memory that can be changed by another process without other synchronization mechanisms. But I get this kind of errors: offset_ptr.hpp:116: error: invalid conversion from 'volatile void*' to 'const void*' offset_ptr.hpp:116: error: initializing argument 1 of 'void boost::interprocess::offset_ptr<T>::set_offset(const void*) [with PointedType = volatile int]' Basically it is not possible to call set_offset when the pointed type is volatile. Is this a feature or a bug? (If it is a bug I'll happily enter it into trac). Lars

2 2

Boost Filesystem Error, am I Linking?
by Chris Miller 21 Nov '07

21 Nov '07

More trouble. Eclipse CDT tells me that when I try and use boost::filesystem::initial_path<boost::filesystem::path>() that in include/boost/filesystem/operations.hpp on line 503 that I have a "undefined reference to 'boost::filesystem::detail::get_current_path_api(std::string&)'" I went to Google and found a forum thread where the person wasn't linking against Boost's Filesystem library. I looked in the lib/ section under my MinGW installation, however, and found libboost_filesystem.a, right where it should be. Is there still the potential for it to not be linking, or am I just not using something correctly? I tried a simpler usage of the same function and got the same problem. I should warn you, I'm still rather green at C++, all my previous programming training being in Java. I'm still getting acclimated to C++, pretty quickly, but still every once in a while I run into a wall and need someone to pick me up and steer me in the right direction, so to speak. Thanks in advance for any help you can give!

4 4

Can Boost help
by Lalit Seth 21 Nov '07

21 Nov '07

Hi, I have a java application which does PDF Priting which I can execute like java -classpath <path to lib> PDFPrinter <filename> <printername> <Copies> <Orinetation> This works well on windows and linux. I want to execute the same from C++ code which will spool a new process and execute this. I am having a windows service which will actually perform this. I need help for a) Spooling a new process using boost so that it works for Windows and Linux. b) Security issues like under windows service will be running under account 'SYSTEM' new process spooled will be again under SYSTEM user hence what all is needed to make sure System has access to printers and/or specific printer. Note: I have tried std::system but again it behaves different for c++ console application and c++ service under windows. Thanks

1 0

[boost-users] pointers container library generic function.
by chun ping wang 20 Nov '07

20 Nov '07

HI i would I write a generic function as template parameter that accepts any kind of boost pointers containers library. like for example. template <typename T, template <typename ELEM, typename = std::allocator<ELEM> > class CONT > extern typename boost::enable_if<boost::mpl::or_<boost::is_integral<T>, boost::is_convertible<T, unsigned long> >, CONT<T> >::type str_to_numlist(const std::string&, const T&); accept any vector, list, deque.. i want to replace vector list, deque with ptr_vector, ptr_list, ptr_deque.. etc.

1 0

[multi-index] dereferencing the end iterator
by Ovanes Markarian 20 Nov '07

20 Nov '07

Hello Joaquín, today I faced some problem, when debugging the code. namespace mi = boost::multi_index; typedef mi::multi_index_container < vx_type , mi::indexed_by < mi::hashed_unique<mi::tag<vx_hash>, hash_extractor, dummy_hasher> , mi::ordered_non_unique<mi::tag<vx_kind>, kind_extractor > > > some_t; When created a view for vx_hash tag and going to change the invalid iterator like view.end(). Dereferencing view.end() fails as awaited, but this code which calls my TypeModifier: view.modify(view.end(), my_modifier); This will result in my_modifier call with dereferenced iterator, which contains badly initialized object. My environment VC++ 2005 Express, boost 1.34.1 With Kind Regards, Ovanes

2 5