- Boost - lists.boost.org

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

uploaded Channel - a lib for (distributed) message passing and event dispatching to boost vault
by Yigong Liu 20 Mar '09

20 Mar '09

Hello there, I just upload a updated version of my Channel library to Vault under Main/"Distributed Computing". This version contains some update for Channel to work with the latest boost distributions and change to boost license, since there are several requests for me to update and upload to Vault. A brief description: Channel procides name space for distributed message passing and event dispatching. Its major components (message IDs, routing algorithms...) are configurable as template parameters. As a namespace shared by peer threads, channels support publish/subscribe scope control, message filtering, and translation.. Thanks Yigong

2 4

Question on ABI tags and autolinking
by Philip Lowman 20 Mar '09

20 Mar '09

Hi, I'm working on some cleanup of CMake's FindBoost module and had a few questions about Boost and it's ABI tags. 1. Are the ABI tags as defined on the webpage below always defined in this order in generated libraries? http://www.boost.org/doc/libs/1_38_0/more/getting_started/windows.html#libr… 2. Regarding the "-s" ABI tag, the current CMake module has a deficiency and only searches for it under Windows and only after not being able to find a non "-s" library there. On other platforms it doesn't search for the "-s" at all. I'm curious what the defaults are for building Boost against static C/C++ system libraries using Boost's present build system? Does this happen on compilers other than VS? I just want to know if there is any history to this oddity or if it's just a bug. 3. Are "-p", "-n", and "-y" ABI tags persistent across all generated libraries filenames once one of them is enabled? In other words, the generated Boost library filenames vary only on "-s" (static runtime) , "-g" (debug standard/runtime libs) , and "-d" (debug build of boost library)? 4. None of this changes (or will change) under the CMake build of Boost, I trust? 5. Is there a consolidated list of Boost libraries online for which MSVC/Borland/etc. auto-linking is always against the shared library? I'd like to be able to link the end-user to a list of these, if possible. Thanks. -- Philip Lowman

2 3

[tuple][io] compiler error on empty tuple
by Rutger ter Borg 19 Mar '09

19 Mar '09

Apparently boost::tuple's io-support doesn't include empty tuples. See attached, this results in a compiler error using g++ 4.3.3. I think it shouldn't give an error, since "b" is a valid instance. Kind regards, Rutger ter Borg

2 2