- Boost - lists.boost.org

Separate Boost newsgroups for individual Boost libraries
by Edward Diener 17 Jul '08

17 Jul '08

I would like to suggest that Boost have separate NGs for each individual library. I also think, for each individual library, a separate user and developer NG should exist. While this will lead to many more NGs, it will make it much easier for users/developers to have concerns addressed for a particular Boost library, and to have the implementor of that library address those concerns. Also of course it will make it easier for people interested, or working with, a particular Boost library, to go to the appropriate NG to read about any issues brought up regarding that library. This is especially important as more Boost library are developed. If a post deals with issues that can not constrained to a single library, it can go to the current 'user' or 'devel' NG as appropriate.

5 5

[smart_ptr> C++0x unique_ptr ?
by Loïc Joly 17 Jul '08

17 Jul '08

Hello, I was wondering whether there was any plan to implement a unique_ptr in boost, that would replace scoped_ptr, and would default to scoped_ptr on compilers lacking move semantic? Best regards, -- Loïc

9 21

Fwd: [GIL] Point concept
by Felipe Magno de Almeida 17 Jul '08

17 Jul '08

---------- Forwarded message ---------- From: Felipe Magno de Almeida <felipe.m.almeida(a)gmail.com> Date: Thu, Jul 10, 2008 at 3:28 AM Subject: [GIL] Point concept To: boost-users(a)lists.boost.org Hi, PointNDConcept is defined as having axis_value member functions. But the GIL code in the boost trunk doesn't comply with this requirement. A free axis_value function is just useless, because it can't be found by ADL when template parameters are passed explicitly. Also, member template axis doesn't model Metafunction. Am I missing something? Regards, -- Felipe Magno de Almeida -- Felipe Magno de Almeida

3 3

[1.36.0] Libraries with unmerged changed - more accurate results.
by Beman Dawes 17 Jul '08

17 Jul '08

Below is a more accurate list of headers and libraries whose header files appear to differ between the trunk and branches/release. The most likely causes are: * Changes that should be merged into branches/release for 1.36.0. If that is the case, developers need to merge these changes right away! * Work in process that should *not* be merged into branches/release for 1.36.0. Please let me know that that is the case so I can stop worrying:-) * Although it seems to be working much better, there still may be issues with the script used to generate the report. Please let me know so the script can be improved. Thanks, --Beman Headers: any.hpp concept_check.hpp enable_shared_from_this.hpp intrusive_ptr.hpp lexical_cast.hpp make_shared.hpp memory_order.hpp operators.hpp pointer_cast.hpp pointer_to_other.hpp rational.hpp scoped_array.hpp scoped_ptr.hpp shared_ptr.hpp smart_cast.hpp static_assert.hpp static_warning.hpp swap.hpp tokenizer.hpp weak_ptr.hpp Libraries: asio bimap circular_buffer config date_time detail filesystem function function_types graph interprocess intrusive iostreams iterator math mpl numeric optional parameter pool program_options property_tree python regex serialization signals spirit statechart system test tr1 type_traits units utility

9 20

[any] small buffer optimizations?
by Thorsten Ottosen 16 Jul '08

16 Jul '08

Hi, As eplained in http://www.codeproject.com/KB/cpp/dynamic_typing.aspx boost::any is not exactly optimal w.r.t. efficiency. Diggins implementation avoid allocation for objects smaller or equal to sizeof(void*). Personally I think this is too small a buffer; Boost.Function uses at least two words; I think it might be worth to go for 3 words. Would anyone be interested in implementing this? best regards -Thorsten

4 3

Help with Access Violation in simple serialization
by Tiago da Silva 16 Jul '08

16 Jul '08

Hi there, I am having some serious problems when trying to instantiate either a boost::archive::xml_oarchive or a boost::archive::text_oarchive: everytime I receive an "Unhandled exception at XXX (msvcr90d.dll) in XXX : Access violation writing location" I have simplified a test case up to this: structure I want to save (in whatever.h): typedef struct s_pt { int x; int y; bool operator< (const struct s_pt &spt2) const { return (x < spt2.x) && (y < spt2.y); } bool operator== (const struct s_pt &spt2) const { return (x == spt2.x) && (y == spt2.y); } template<class Archive> // turns into a serializable component void serialize(Archive &ar, const unsigned int version) { using boost::serialization::make_nvp; ar & make_nvp("x", x); ar & make_nvp("y", y); } } POINT2D; saving routine in main.cpp std::ofstream ofs("savedProject.imag3d"); boost::archive::xml_oarchive oa(ofs); const structs::POINT2D p = {111, 222}; oa << boost::serialization::make_nvp("p2d", p); but I get this error in boost::archive::xml_oarchive oa(ofs); I have tried already both xml_oarchive and text_oarchive... but the demo.cpp provided by http://www.boost.org/doc/libs/1_35_0/libs/serialization/doc/index.html involving the bus routes works!!! what am I doing wrong??? I use Visual Studio 2008. Thanks in advance, Tiago

3 7

[GIL] 3d image support
by Falco Hirschenberger 16 Jul '08

16 Jul '08

Hi, As I read from the documentation, GIL supports N-dimensional image concepts but lacks an Implementation of 'PointNDConcept', 'RandomAccessNDImageViewConcept', 'RandomAccessNDImageConcept' and 'RandomAccessNDLocatorConcept' for 3d images. Has anybody implemented these as extension to GIL or is it on the TODO list? What do you think will be the effort to implement them? Is it a no-brainer by just implementing the concepts similar to the 2d counterparts or will there be some difficulties from the design? Greetings, Falco

3 2

[1.35.0] Serialization demo.cpp problem
by Scott Woods 16 Jul '08

16 Jul '08

Hi, Just trying to get a comparison going between several different techniques for serialization. After downloading 1.35.0 and using the beautiful auto-installer it was looking pretty good. Trying to compile demo.cpp hasnt gone quite so well; Deleting intermediate files and output files for project 'demo - Win32 Debug'. --------------------Configuration: demo - Win32 Debug-------------------- Compiling... demo.cpp c:\program files\boost\boost_1_35_0\boost\serialization\utility.hpp(50) : error C2989: 'is_bitwise_serializable<struct std::pair<T,`template-parameter258'> >' : template class has already been defined as a non-template class c:\program files\boost\boost_1_35_0\boost\serialization\utility.hpp(50) : error C2988: unrecognizable template declaration/definition c:\program files\boost\boost_1_35_0\boost\aligned_storage.hpp(59) : fatal error C1506: unrecoverable block scoping error Error executing cl.exe. demo.exe - 3 error(s), 0 warning(s) I've searched the mail arachives. There have been a few issues around demo.cpp but they seem unrelated. Anyone know whats going on? Windows XP, Microsoft Visual C++ 2005. Thanks, Scott

1 0

[statechart] test failures on the release branch
by Richard Webb 15 Jul '08

15 Jul '08

Merging the changes in Function to the release branch has caused some failures in the statechart tests (e.g. http://tinyurl.com/6hmxod) So i guess that Statechart needs to be merged as well now? Thanks, Richard Webb

2 2

Re: [boost] Bind question, is there any way to implicitly go through a reference_wrapper?
by Michael Goldshteyn 15 Jul '08

15 Jul '08

"Peter Dimov" <pdimov(a)pdimov.com> wrote in message news:00c501c8e6b1$b3a11f60$6607a80a@pdimov2... > > It doesn't because of the mem_fn behavior that bind inherits. The > argument is not a reference to Crayon and so is assumed to be a smart > pointer to Crayon. mem_fn then tries to invoke get_pointer() on it to obtain a Crayon*. > You should be able to make it compile by adding a get_pointer overload > for > reference_wrapper: > > namespace boost > { > > template<class T> T* get_pointer( reference_wrapper<T> const & r ) { > return r.get_pointer(); > } > > } That would be an awesome addition, since reference_wrapper is an oft-used part of boost and there's really no reason for it not to be automatically traversed, making bind constructs that involve it a little simpler. Thanks, Michael Goldshteyn

1 0