- Boost - lists.boost.org

[gsoc]
by Matt 22 Mar '09

22 Mar '09

Hello, All I am Matthew Chandler, and I am looking to help out for the Google summer of code. I've looked through the ideas page and am interested in working on the Devector container. -Matt

2 1

[C++0x] The cpp0x branch has been merged into trunk
by Beman Dawes 21 Mar '09

21 Mar '09

The cpp0x branch has been merged into trunk, enabling a bunch of BOOST_NO_* macros for C++0x feature tests. This should have no impact on existing code, so be sure to let me know if it causes any problems. --Beman

7 19

[asio] possible multi-threaded bug with async_write_some()
by Brian and Victoria 21 Mar '09

21 Mar '09

I'm running asio (boost 1.38) multi-threaded, and noticed that on a context switch between threads, my handler for async_write_some() will sometimes get invoked with no data transferred. My handler would call async_write_some() again if there was remaining data, but this caused (most of) the data to get sent twice since there appears to still be a call to the handler pending in the work queue. I'm not sure if this behavior is a bug or not, but the solution was to check for 0 data transferred and exit the handler if so. -Brian

1 1

threadpool lockfree_channel
by Tim Blechmann 21 Mar '09

21 Mar '09

hi oliver, i just saw, that you checked in a lockfree fifo implementation to the boost sandbox. it looks like an implementation of the michael/scott queue, isn't it? if so, i suppose, you are missing a thread-safe memory reclamation mechanism ... some time ago, i wrote a boost-style implementation of this data structure, not sure, if you came across it [1] ... maybe it would make sense to join our efforts in implementing a lockfree queue and get it into boost? best, tim [1] http://tim.klingt.org/git?p=boost_lockfree.git;a=summary -- tim(a)klingt.org http://tim.klingt.org Music is the can opener of the soul. It makes you terribly quiet inside, makes you aware that there's a roof to your being. Henry Miller

9 21

[static] A useful static variable generator.
by Dongfei Yin 21 Mar '09

21 Mar '09

Hi, Some time, the sequence of static variable initialization always make something bad. the piece of code below will useful about this problem. class static_ { public: template <int N, class T> static T& var() { static T instance; return instance; } private: ~static_() {} }; so, if a static variable of std::string is wanted, I will use: static_::var<0, std::string>() = ":)"; the first template parameter is used for when many static variable of same type is needed. most important thing is all the variable will be initialized before use. Just for fun. Dongfei.

3 2

[geometry] robustness approaches
by Fernando Cacciola 20 Mar '09

20 Mar '09

Hi boost geometers ;) When Barend first presented his geometric library proposal I argued that the so-called "epsilon-based" approach used by the library is just fundamentally wrong, and most importantly, totally unnecessary since much better methodologies exist. It deeply worries me to find that the GGL proposal haven't been updated in this regard, because the instrumentation of the proper techniques have a significant impact on the overall design. Robustness can't be treated as an implementation detail to be deal with later on (and I'm sorry if I haven't made this point clear when I warn about it before). If I where to review the library today I would have to strongly vote for rejection because of this. Here is a simple example of the fundamental problems I see all over the proposed library: There is a point_in_circle algorithm, which (as I can see from its implementation), essentially returns whether the distance from the point to the center is smaller than the circle radius. No doubt *that* is the one and only correct *mathematical* formulation for the predicate, but is one that just doesn't stand in practice at all. Computing *that very particular predicate ROBUSTELY* has been the subject of active research during the past decade, to such an extent, that pretty much every paper or article on the subject of robustness in geometric computing uses the in-circle test as a common example. The library proposed by Lucannus, OTOH, at least acknowledges the problem appropriately and approaches a solution in the right direction (there is still some important design-influencing considerations but are not showstopers) Here is a starting point http://www.mpi-inf.mpg.de/departments/d1/ClassroomExamples/ Best -- Fernando Cacciola SciSoft Consulting, Founder http://www.scisoft-consulting.com

8 36

casting a shared_ptr from void with custom deleter
by Kenny Riddile 20 Mar '09

20 Mar '09

I'm trying to do something like this: boost::shared_ptr<const Foo> fooPtr( boost::static_pointer_cast<const Foo>(typeErasedFoo), CustomDeleter<const Foo>() ); where typeErasedFoo is of type: boost::shared_ptr<const void> The first statement compiles if I remove the custom deleter, but if I leave it in, I get errors saying: cannot convert parameter 2 from 'CustomDeleter<T>' to 'boost::detail::sp_enable_if_convertible_impl<true>::type' Is there any way to do what I'm trying to do? I no longer need typeErasedFoo after the cast if that helps.

3 3

extending asio with unix signal handlers
by Dmitry Goncharov 20 Mar '09

20 Mar '09

Hi, What do you think of extending asio with unix signal handling? The extension lets the library user to have a handler invoked when a unix signal arrives. void on_signal(boost::system::error_code const& error) { if (!error) cout << "sigint received" << endl; } int main(int argc, char const* argv[]) { boost::asio::io_service ios; boost::asio::posix::signal_handler<SIGINT> sigint(ios); while (true) { ios.reset(); sigint.async_wait(boost::bind(on_signal, _1)); ios.poll(); } } This lets the user have a unix signal handler of an arbitrary signature. This also removes the "async-signal safe calls" restriction. Please, find the implementation along with two examples enclosed. BR, Dmitry

4 4

Re: [boost] [gsoc][student] Proposal
by Kasra 20 Mar '09

20 Mar '09

> You're proposal has significant overlap with Crypto++ (public domain). >You might want to investigate some sort of code sharing, or at least > have a look at how they've done things. Thanks for mentioning Crypto++, I am aware of the library and have used it few time before (though it is a pain to compile and _link_). The library isn't exactly in public domain however, I have decided to use the algorithms code (i.e. mmx, sse2 instruction set and etc). The rest of boost.crypto library is designed to use templates over inheritance (the approach Wei Dai has used on Crypto++). With Best Regards --- Kasra Nassiri

1 0

[fusion] compiler errors with empty boost::tuple
by Rutger ter Borg 20 Mar '09

20 Mar '09

I am experiencing a compiler error if I try to use a fusion join function on empty tuples. The operation I am trying to do does work with fusion::vectors. See attached; I guess this is a bug? Kind regards, Rutger ter Borg

3 5