February 2007 - Boost - lists.preview.boost.org

[Feature request][Lambda] Support for pairs and tuples
by Giuseppe Ottaviano 09 Feb '07

09 Feb '07

Hi, I work often with containers of tuples or pairs, using stl and boost algorithms. Accessing elements with boost lambda is a bit cumbersome, requiring syntax like bind(&pair<int, int>::first, _1) I have written a small adaptor to make get<> work with lambda expressions, so that the syntax becomes get<0>(_1) It works for std::pair and boost::tuple. It is only a suggestion, i do not know if it handles correctly consts, references, etc, and has been tested only on gcc 4.1.2. I've posted the code on boost vault: http://www.boost-consulting.com/vault/index.php?action=downloadfile&filenam… Can something like this be implemented in future versions of Lambda? Thanks, Giuseppe Ottaviano giuott(a)tin.it

1 0

[interprocess] - wish semaphores with get_value member fn
by Oliver.Kowalke＠qimonda.com 09 Feb '07

09 Feb '07

Hello, Is it possible to provide the get_value member function for sempahore? If POSIX semaphores are used then int sem_getvalue (sem_t sem, int wert) Could be used. Regards, Oliver

2 1

Interest in Behavior Driven Development Interface (again)
by Dean Michael Berris 09 Feb '07

09 Feb '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Everyone, I've just recently decided to ask for comments (again) about the BDD interface I had been developing (on and off for the last few months). I recently picked it up again, and basing from the comments on the previous thread at http://tinyurl.com/2xmm2x there's still quite mixed reactions about BDD as an alternative to the Unit Test construct. That being said, that discussion is a good 4 months ago, and would like to ask the following questions to those who might be interested in something like this: * What degree of integration with Boost.Test be "suitable"? Though I see Boost.Test has a wealth of preprocessor macro's going about, it shouldn't be hard to integrate at that level. * What things would you like to see as part of the specification toolkit? I'm planning to put in string Regex matching support, integral type inspection (divisible by, is prime, is even/odd, etc.), and maybe pointer inspection (is null, is not null, is of type, derives from, etc.) and STL collection inspection. Input would be most appreciated. * Will a code generator (or specification file processor) be a useful addition? One which creates BDD implementations from "simple english" or using a simple specification language (a DSL)? I've re-read through the thread a few months back, and I guess the idea is "yes, it's interesting... but..." and I intend to explore the reasons for the apprehension. I've tagged release 0.1 using the git scm tool and the public (readonly) tree is available at http://orangeandbronze.com/~dean/spec.git -- while tarballs (.gz|.bz2) are available at http://orangeandbronze.com/~dean . I'd also like to make my intentions clear, I am developing this library for future review and (hopefully) inclusion into the Boost C++ Library - -- though of course, I still need to see what the community thinks about it first. Thanks in advance! - -- Dean Michael C. Berris http://cplusplus-soup.blogspot.com/ mikhailberis AT gmail DOT com +63 928 7291459 -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFFwjWOBGP1EujvLlsRAniFAJ4hZ48xkUJlhJbimjfn1qfadwZlTQCfU49t M05ECuVVkSCss9VHha0CYl8= =o9rE -----END PGP SIGNATURE-----

4 9

Boost.Build v2: Native/non-native wchar_t in msvc toolset
by Alexei Alexandrov 09 Feb '07

09 Feb '07

Hi All! I remember that in the first Boost.Build there was an option which allowed to choose whether native wchar_t should be used or typedef-ed to "unsigned short". In v2 the behavior seems to be hardcoded as "always native wchar_t". Is it possible to bring this feature back? -- Alexei Alexandrov

2 1

boost::mpi accross several computers
by Philippe Vaucher 09 Feb '07

09 Feb '07

Hello, from the tutorial ( http://www.osl.iu.edu/~dgregor/boost.mpi/libs/parallel/doc/html/) I read this sentence : "A Boost.MPI program consists of many cooperating processes (possibly running on different computers) that communicate among themselves by passing messages." How would that be done ? Can't find anything in the doc about doing it... an example would be great :) Thank you in advance Philippe

3 4

Re: [boost] [asio].[hp-ux].[1_33_1] Build related issue
by Gaurav.Jain＠iflexsolutions.com 09 Feb '07

09 Feb '07

Christopher Kohlhoff <chris <at> kohlhoff.com> writes: > > This seems to be an issue between boost::array and the Sun compiler, > and not to do with asio as such. I see the same errors in the > regression logs for boost CVS HEAD: Solaris: Actually after passing "-D_XOPEN_SOURCE=500" and "-D__EXTENSIONS__" to the compiler number of errors came down drastically which would mean that the asio code doesn't require any change. > > http://engineering.meta-comm.com/boost-regression/CVS-HEAD/developer/o > utput/mmutz-bin-boost-status-array0-test-sunpro-debug.html Regarding the regression error, for boost Sun Studio requires the user to pass following flags "-library=stlport4 -features=tmplife -features=tmplrefstatic" > Are you sure the daytime_client.cpp source you are using is the one > written for 0.3.8? That line number and error message corresponds to > the > 0.3.7 source. > > Cheers, > Chris Am afraid to say that you are right in your observation. I am now using 0.3.8 client. CC -D_XOPEN_SOURCE=500 -D__EXTENSIONS__ -library=stlport4 -features=tmplife -features=tmplrefstatic Now am getting this error : /boost/1_33_1/64/boost/asio/detail/posix_fd_set_adapter.hpp", line 35: Error: The function "memset" must have a prototype. 1 Error(s) detected. Regards, Gaurav Jain

1 0

Re: [boost] formal review request : mcs::units
by Matthias Schabel 09 Feb '07

09 Feb '07

> I uploaded another version of dimension.hpp. If there are any > problems let me know. I hope that these optimizations are sufficient. Hi Steven, This new version fixes the problems I noted with the previous version and looks very nice - I haven't had a chance to rigorously compare compile times, but it seems subjectively faster. With your permission, after I've had a chance to run a couple of torture tests, I'd like to include it for the next release... Naturally, I'll acknowledge your contribution in the documentation. Regards, Matthias

1 0

[Boost.Threads] New Feature: thread function parameter
by Eugene Lyubimkin 08 Feb '07

08 Feb '07

Hello! I want to propose a new feature in *Boost.Threads *Now class _boost::thread_ have only one constructor: /thread::thread( const function0<void>& ) /So we cannot send any parameter to thread's main function, because its prototype have no parameters. My proposal: let add constructor /thread::thread( const function1<void, void*>& )/ Parameter with type /void* /will be sent to thread function. Currently *Boost.Threads *supports Win32 API and POSIX API. Both these APIs support sending one parameter /void*/ to thread's main function. Sorry for my English. ----------------------------------------------------------- Best, Eugene Lyubimkin, the Ukrainian C++ developer.

3 2

Problem with MPL/GIL dynamic_image example on gcc-4.1.1
by Andreas Pokorny 08 Feb '07

08 Feb '07

Hi, GCC-4.1.1 gives me the attached error message while compiling dynamic_image. It boils down to: dynamic_image.cpp:14: instantiated from here boost/mpl/aux_/preprocessed/gcc/template_arity.hpp:85: error: call of overloaded 'arity_helper( aux::type_wrapper< .... >,aux::arity_tag<1>)' is ambiguous boost/mpl/aux_/preprocessed/gcc/template_arity.hpp:30: note: candidates are: char (&aux::arity_helper(...))[1] boost/mpl/aux_/preprocessed/gcc/template_arity.hpp:37: note: char (&aux::arity_helper(::aux::type_wrapper<F<T1> >,aux::arity_tag<1>))[2] boost/mpl/aux_/preprocessed/gcc/template_arity.hpp:37: note: char (&aux::arity_helper(::aux::type_wrapper<F<T1> >,aux::arity_tag<1>))[2] Is this a known mpl<->gcc-4.1.1 issue? Furthermore there are some odd warnings, and I am not sure what to think about these. The code there looks a bit fishy. I tried the example with gcc-3.4.4, gcc-4.2.0-alpha20060902 and gcc-4.3.0-alpha20061111. It compiled there, but the warning remained in the gcc-4 series. Regards Andreas Pokorny

4 3

[Review] Result of the Accumulators review
by John Phillips 08 Feb '07

08 Feb '07

Eric and all, I'll post a more detailed result, with a condensed list of suggestions and requests from the review in a few days, but since everyone who responded in the review voted for inclusion, the Accumulators library has been accepted for inclusion into boost. Many thanks to Eric for the submission, as well as Daniel Egloff, Dave Abrahams and Matthias Troyer for their parts in the design and development. Thanks as well to the reviewers (Steven Watanabe, Niitsuma Hirotaka, John Madddock, Paul Bristow, Ben FrantzDale, James Jones, Hans Meine, Topher Cooper, Hervè Brönnimann, Matt Hurd, Javier Estrada, Martin Knoblauch Revuelta, Ronald Garcia, Tom Brinkman and Bernhard Mäder) for the time and effort they were able to contribute to the review. John Phillips (Review Manager)

2 1