May 2012 - Boost-users - lists.preview.boost.org

Updating filesystem to V3 (boost 1.49)
by Shai Shasag 06 May '12

06 May '12

I'm in the process of updating my code to use filesystem V3. I replaced all the V2 deprecated functions with V3 calls. I defined the macros: BOOST_FILESYSTEM_VERSION=3 BOOST_FILESYSTEM_NO_DEPRECATED When compiling the file libs/filesystem/v3/src/path.cpp I get the following error: #error both BOOST_FILESYSTEM_DEPRECATED and BOOST_FILESYSTEM_NO_DEPRECATED are defined The reason for this error is that in filesystem/v3/config.hpp has the following: // BOOST_FILESYSTEM_DEPRECATED needed for source compiles # ifdef BOOST_FILESYSTEM_SOURCE # define BOOST_FILESYSTEM_DEPRECATED # endif So i'm puzzled as to why BOOST_FILESYSTEM_DEPRECATED is forced when compiling path.cpp? And how can I avoid theses erros when compiling filesystem V3? Thanks Shai Boost version: 1.49 Compiler: gcc version 4.2.1, Xcode 3.2.5 on Mac OS 10.6.8

2 1

random number generators seeding with 0 and 1
by Nelis Willers 05 May '12

05 May '12

Dear Boost Users While testing my RNG I noticed that the use of seed value 0 yields the same results as using a seed value of 1. It seems that this is the specified behaviour for rand() in the C runtime library: http://stackoverflow.com/questions/8049556/whats-the-difference-between-sra… I could find no reference to this in my (oldish) draft C++0x spec. Also no mention in the boost random doc page. I tried to unravel the random include files, but got lost. Is this behaviour required by the specification or just following tradition? More important, is this behaviour guaranteed in all libraries? thanks! nelis

2 1

Building boost to NOT USE built-in Visual Studio 2010 TR1 random
by Nelis Willers 05 May '12

05 May '12

Dear Boost-Users boost 1.49.0 and Visual Studio 2010 / Visual Studio 2005 / & g++ 4.6.3 on (Win7 and Ubuntu 12.04) Build boost libs locally, using the local compilers. The objective is to get exactly the same random number generator behaviour for the same application built with VS2005, VS2010 and Gnu compilers. Building the same code, I see the* same normal & uniform* results under gnu and VS2005 builds (presumably boost random is used under VS2005). Building the same code, I see the different *normal results* under gnu and VS2010 builds (presumably the native VC lib is used under VS2010). So I would like to use (at least) boost's random library instead of the native Visual C++ 2010 library. >From boost docs I learned that it is possible to set up the boost config files such as to use/not use the native compiler/library TR1 implementation: http://www.boost.org/doc/libs/1_49_0/libs/config/doc/html/index.html#boost_… http://www.boost.org/doc/libs/1_49_0/libs/config/doc/html/boost_config/boos… http://www.boost.org/doc/libs/1_49_0/doc/html/boost_tr1/usage.html#boost_tr… I cannot however, figure out how to do this from the above docs. BOOST_HAS_TR1_RANDOM comes into it, but I fail to get the big picture. Can anybody please advise? thanks in advance! nelis

3 2

Bimap compare function
by Tyler Conant 05 May '12

05 May '12

I noticed that the comparator for a bimap can only be a Functor Type. It does not appear that I can pass a function object or a function pointer, as a STL Set would allow. Is this correct or did I miss something. For my application I could really use the ability to pass in comparator which is decided at runtime. Does anyone have any suggestions on a work around?

3 2

[mpl] Why no wrapper for pointers as non-type template parameters?
by Philipp Moeller 04 May '12

04 May '12

I'm a little surprised that the MPL does not offer a wrapper for pointers as constants. They sometimes come in handy to specify lists to different callable objects. Something along the lines of: template<typename T, T N> struct ptr_wrap { static const T value; typedef ptr_wrap<T, N> type; typedef T value_type; }; template<typename T, T N> typename ptr_wrap<T, N>::value_type const ptr_wrap<T, N>::value = N; already does the job for me, but it could possibly be enhanced with a helper to use the correct syntax for pointers to member. Could this be an addition to the library or is there a better way? Cheers, Philipp Moeller

1 0

[Serialization] Linker errors issue with using binary_wiarchive.hpp and binary_woarchive.hpp
by Lord Metritutus Merkwurdichliebe 04 May '12

04 May '12

Hello there, I've never used one of these Mailing List conversations before so I apologise for any mistakes. To the point: I'm having issues trying to get the wide character version of Boost's serialization libaries working. When I attempt to compile the result is a multitude of Linker errors. After Googling failed to turn up anything I posted the question on Stack Overflow, where a reply suggested there may be 'a bug in Boost.Serialization'. Stack Overflow features nice syntax highlighting so it is easy to see the code, I pieced it together from other normal character examples: http://stackoverflow.com/questions/10439984/serialization-example-of-boost-… If there is a rule against posting external links, I apologise. Or if people want me to paste the code directly in to an email for this mailing list I will do so. The Stack Overflow question was originally based on my thought that I had made a mistake in my code. If this IS the case then I do apologise, and I would appreciate knowing just where I've gone wrong. Regards.

3 4

Re: [Boost-users] [Serialization] Linker errors issue with using binary_wiarchive.hpp and binary_woarchive.hpp
by Lord Metritutus Merkwurdichliebe 04 May '12

04 May '12

I believe I was asked this in the Stack Overflow question I linked as well. Assuming that all the .lib files in the stage/lib folders are the built libraries (I could be wrong so please correct me if that is the case), and assuming that libboost_serialization-vc100-mt-1_47.lib or libboost_serialization-vc100-mt-gd-1_47.lib are the libraries I need, I believe that both of the latter things you mentioned are satisfied. Regards. From: bryan.bagnall(a)navy.mil To: boost-users(a)lists.boost.org Date: Thu, 3 May 2012 18:36:03 -0700 Subject: Re: [Boost-users] [Serialization] Linker errors issue with using binary_wiarchive.hpp and binary_woarchive.hpp From: http://www.boost.org/doc/libs/ It looks like the serialization library requires you to build boost (it is not header only) Instructions can be found here: http://www.boost.org/doc/libs/1_49_0/more/getting_started/windows.html http://www.boost.org/doc/libs/1_49_0/more/getting_started/unix-variants.html If you have built boost already, make sure to include the boost libs in your linker path. Regards -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Lord Metritutus Merkwurdichliebe Sent: Thursday, May 03, 2012 5:09 PM To: boost-users(a)lists.boost.org Subject: [Boost-users] [Serialization] Linker errors issue with using binary_wiarchive.hpp and binary_woarchive.hpp Hello there, I've never used one of these Mailing List conversations before so I apologise for any mistakes. To the point: I'm having issues trying to get the wide character version of Boost's serialization libaries working. When I attempt to compile the result is a multitude of Linker errors. After Googling failed to turn up anything I posted the question on Stack Overflow, where a reply suggested there may be 'a bug in Boost.Serialization'. Stack Overflow features nice syntax highlighting so it is easy to see the code, I pieced it together from other normal character examples: http://stackoverflow.com/questions/10439984/serialization-example-of-boost-a rchive-binary-woarchive-hpp-and-or-boost-archive <blockedhttp://stackoverflow.com/questions/10439984/serialization-example-of -boost-archive-binary-woarchive-hpp-and-or-boost-archive> If there is a rule against posting external links, I apologise. Or if people want me to paste the code directly in to an email for this mailing list I will do so. The Stack Overflow question was originally based on my thought that I had made a mistake in my code. If this IS the case then I do apologise, and I would appreciate knowing just where I've gone wrong. Regards. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

[Polygon] Multiple Holes
by Bagnall, Bryan D CIV SPAWARSYSCEN-PACIFIC, 56220 04 May '12

04 May '12

Hello, I'm trying to use Boost::polygon to do Boolean operations in my project. My project uses contours that I'm grabbing with OpenCV, processing with Boost::polygon, then writing out to an ESRI shapefile. OpenCV can give me a list of contours within contours within contours...etc, and shapefiles can be written out this way (see link for picture http://icode-mda.googlecode.com/svn/trunk/examples/09_KMLshp/shapefilesWithH oles/results/blobsWithHoles.png ) I haven't been able to jam all of the data in a single polygon_with_holes_data polygon. Is it true that boost::polygon::polygon_with_holes_data can only have an exterior contour and multiple holes, but no contours inside of the hole? In other words, could a single boost polygon support a continent with a lake, with an island, with a pond, with a small island... ? Is there another type that would support this? Maybe the Geometry library? My current work around is to have (1)the continent with a lake as one polygon, (2)the island with a pond as another,(3)the small island as another... Thanks, Bryan

1 0

[Serialization] Issues with C++/CX in MSVC 11
by Brent Lewis 04 May '12

04 May '12

In porting a project from iOS to Windows 8, I'm continuing to use Boost 1.49 to save the state of the program. I'm having what seems to be problems with a compiler bug. I'm wondering if anyone has found a workaround? Thought it's probably not necessary, I'll explain exactly what I'm seeing in detail in case it's helpful. If one creates a new Visual Studio project from the C++ Metro style Direct3D application template, simply adding these lines to the main function works just fine: std::stringstream memStream(std::ios_base::binary | std::ios_base::in | std::ios_base::out); boost::archive::binary_oarchive oa(memStream); However, when I add the rest of the project files (even if I don't instantiate any of the classes - just adding the modules does it), the two same lines in the main function crash. I eliminated all the code from the modules that refer to serialization and archive classes, because I noticed that there was a lot of static initialization of singletons of templated archive classes. Even with minimal static initializers, still, the code fails to run. Closest thing I can tell is that the compiler is not correctly translating the this pointer when moving from a class's initializers into a superclasse's constructor when multiple inheritance is involved (and perhaps lots of templating is required too). To be specific, in one specific run of the application (note that I've removed the template parameters for brevity): Inside binary_oarchive_impl ctor, the this pointer is: 0x0091fcb8 moving into its superclass constructor, basic_binary_oprimitive, the this pointer is: 0x0091fcc1 That would suggest that the superclass is 9 bytes into the derived class, however: back inside the binary_oarchive_impl ctor, using the debugger: (boost::archive::basic_binary_oprimitive*)this : 0x0091fcc4 offsetof(basic_binary_oprimitive, m_sb) = 0 (offsets retrieved by inspecting/altering memory, since you can't get the offset of a reference type) offsetof(binary_oarchive_impl, m_sb) = 12 Which clearly indicates that the superclass is 12 bytes into the derived class. In the superclass, it initializes m_sb - which is a reference to the stream buffer the archive writes to. Stepping back out of the superclass constructor, back into the derived class constructor, m_sb points to the wrong memory, of course. Has anyone else experience this with C++/CX? Is there an existing bug report? We've got a Microsoft contact that's helping us iron out issues. Does anyone know a concise repro that'll help MS solve this problem faster? Thanks for any information.

2 2

[Range] Weak concept assertions.
by Robert Jones 03 May '12

03 May '12

Here's a snippet from boost/range/algorithm_ext/push_back.hpp... namespace boost { namespace range { template< class Container, class Range > inline Container& push_back( Container& on, const Range& from ) { BOOST_RANGE_CONCEPT_ASSERT(( SinglePassRangeConcept<Container> )); BOOST_RANGE_CONCEPT_ASSERT(( SinglePassRangeConcept<const Range> )); BOOST_ASSERT( (void*)&on != (void*)&from && "cannot copy from a container to itself" ); on.insert( on.end(), boost::begin(from), boost::end(from) ); return on; } } // namespace range using range::push_back; } // namespace boost Shouldn't the first concept assert be asserting that 'Container' is in fact a container, not a range? Thx - Rob.

1 0