December 2010 - Boost - lists.preview.boost.org

NVCC CUDA SUpport
by Joel Falcou 31 Dec '10

31 Dec '10

Here is a few tickets regarding nvcc support: https://svn.boost.org/trac/boost/search?q=nvcc+cuda #4431: Bugs: stdint.h not available in CUDA nvcc + visual studio (new) <https://svn.boost.org/trac/boost/ticket/4431> #4356: Patches: small change to allow mpl code to compile with nvcc (the compiler for ... (new) <https://svn.boost.org/trac/boost/ticket/4356> #4052: Patches: NVCC and boost/type_traits/has_new_operator.hpp (closed: fixed) <https://svn.boost.org/trac/boost/ticket/4052> #4046: Bugs: Boost.System does not compile on Cuda (closed: invalid) <https://svn.boost.org/trac/boost/ticket/4046> as the originator of the nvcc config patch and havign the ressource to test and patch these issues, would the owner of said ticket see any problem if I reassign them to me ?

4 6

Your Boost Needs You!
by John Maddock 31 Dec '10

31 Dec '10

There's been a lot of discussion lately about improving participation in Boost, this is a call to arms for a *small* initial experiment which I hope will grow into something bigger, perhaps after some tweaking to the initial experiment. The aim is to merge the ideas put forward for a Boost Guild, with mine for a roving hit squad to tackle neglected libraries. It goes like this: * Someone, lets call him a manager/mentor/maintainer puts forward a library that has a large number of open tickets against it and which maybe hasn't seen maintenance in some time. * The manager creates a copy of that library in the sandbox for folks to work on. * The volunteers either submit patches or - preferably - commit patches direct to the sandbox copy. Work continues for a month or two until everyone agrees that they have a release ready update (tested locally by the volunteers). * The manager reviews the changes and either: merges the changes to Trunk, or files a clear list of things to do to make the update acceptable. * If the Trunk tests don't pass (and the fix isn't obvious/trivial) the manager throws the problem back to the volunteers, otherwise merges to Release. The aim is to: * Minimize the "manager/maintainer/mentor" bottleneck by having one review of a whole set of changes to one library, rather than having to deal with lots of little patches from all the over the place. * Maximize the chances that the volunteers achieve something meaningful in a defined set of time, by defining clear goals from the outset, and selecting a manageable target from the many folks could choose - currently we suffer from the "too much to choose from or do, so why do anything" dilemma. * The manager undertakes the responsibility for stewarding the update through to the release branch - frankly this is only possible by the mentor dealing with one library at a time, otherwise the "many small patches" problem quickly becomes overwhelming. * Give recognition to volunteers for their efforts, first by providing the "I did that" feeling from working towards a specific target, second by adding acknowledgements or in the case of partial (or complete!) rewrites, co-authorship to the docs. Like I said the aim is to start small with an experiment (low risk, low expectations) to see how it goes... then refine the process... and hope we can grow it into something larger over time. For this I'm prepared to act as the mentor/manager for an update to the pool library (though I'm certainly open to other suggestions, and if someone else would like to take on the manager/mentor role then by all means fell free! :-) ). I chose this library because it's seen essentially no maintenance (and certainly no maintainer) since it was written, and because on the face of it, it looks to be a manageable task for folks to take on. So... let me know if you're interested, and if there's enough people I'll define the goals and get things started, Regards, John.

9 13

Re: [boost] [Boost-commit] svn:boost r67496 - trunk/boost/spirit/home/phoenix/detail
by Steven Watanabe 31 Dec '10

31 Dec '10

AMDG On 12/29/2010 3:38 PM, joel(a)boost-consulting.com wrote: > Author: djowel > Date: 2010-12-29 18:38:19 EST (Wed, 29 Dec 2010) > New Revision: 67496 > URL: http://svn.boost.org/trac/boost/changeset/67496 > > Log: > Added specialization for T const* in reference_type The code says T* const. Which should it be? > Text files modified: > trunk/boost/spirit/home/phoenix/detail/type_deduction.hpp | 6 ++++++ > 1 files changed, 6 insertions(+), 0 deletions(-) > > Modified: trunk/boost/spirit/home/phoenix/detail/type_deduction.hpp > ============================================================================== > --- trunk/boost/spirit/home/phoenix/detail/type_deduction.hpp (original) > +++ trunk/boost/spirit/home/phoenix/detail/type_deduction.hpp 2010-12-29 18:38:19 EST (Wed, 29 Dec 2010) > @@ -254,6 +254,12 @@ > typedef T& type; > }; > > + template<typename T> > + struct reference_type<T* const> > + { > + typedef T const& type; > + }; > + > template<typename C> > struct const_reference_type > { In Christ, Steven Watanabe

2 1

[MPL] web doc suggestion.
by Mostafa 31 Dec '10

31 Dec '10

Hi, From: http://www.boost.org/doc/libs/1_45_0/libs/mpl/doc/refmanual/classes.html, it's not readily apparent that in addition to vector_c, list_c, set_c, there are also vector, list, and set classes. I suggest also mentioning and linking to them in the aforementioned url. Thanks, Mostafa

2 2

[Fusion] begin/end ADL issues with C++0x range-based for
by Michel MORIN 30 Dec '10

30 Dec '10

Hi, In some cases, Fusion Sequence cannot be used in C++0x range-based for (available on GCC 4.6 Pre-Release). For example, this code #include <vector> #include <boost/fusion/include/vector.hpp> int main(int argc, char* argv[]) { typedef boost::fusion::vector<int, int> sequence; for (auto x : std::vector<sequence>()) {} return 0; } does not compile. The C++0x range-based for for (auto x : std::vector<sequence>()) {} is equivalent to { auto&& rng = std::vector<sequence>(); for ( auto it_begin = begin(rng), it_end = end(rng); it_begin != it_end; ++it_begin ) {} } Note that rng has std and boost::fusion as its associated namespaces and the unqualified begin/end function call triggers ADL. The reason for the compiler errors is that the call to begin/end is ambiguous between std::begin/end and boost::fusion::begin/end. (Also boost::fusion::begin/end tries to instanciate result_of::begin/end::<std::vector<sequence> > which leads to a compiler error.) These compiler errors can be avoided by using SFINAE (lazy_enable_if with "is_sequence" metafunction) to boost::fusion::begin/end: // <boost/fusion/sequence/intrinsic/begin.hpp> template <typename Sequence> inline typename lazy_enable_if< traits::is_sequence<Sequence> , typename result_of::begin<Sequence> >::type const begin(Sequence& seq) { return result_of::begin<Sequence>::call(seq); } I attached a patch (against trunk) and a test case. The test case does not compile before applying the patch, but compiles fine after applying the patch. Joel and Christopher, what do you think about this change? Regards, Michel

6 14

[icl] doc filenames
by John Maddock 30 Dec '10

30 Dec '10

Just noticed while doing an update that the documentation filenames for the ICL library are *very* long, and certainly won't pass our inspection script tests (although apparently these have stopped being run :-( ). Assuming that this is quickbook generated docs, the trick is to give each section an ID rather than rely on the auto-generated ones, for example: [section:foo Some long title that generates a really long file name...] Just a quick heads up, John.

4 4

[msm] triggering of a transition on a subclass event does not work on sub state machines in case of favor_compile_time policy is used.
by Richard Szabo 30 Dec '10

30 Dec '10

Hi All I have run it to a problem that feature : Event Hierarchy see doc : http://svn.boost.org/svn/boost/trunk/libs/msm/doc/HTML/ch03s02.html#d0e1108 Does not work in case of sub state machine with favor_compile_time policy. No trans get called on the sub state machine even it has a row with event type to the event base class. This now hit me badly because compiler runs out of heap if favor_compile_time policy is not used but when it is used I can not dispatch based on base type and I have about 150 different event so declare a transition for it is a no go. Is there any workaround for this problem ?. Br. Richie

1 0

[PATCH 1/2] Fixed warning: ignoring return value of 'size_t fread(void*, size_t, size_t, FILE*)'
by Vasyl' 30 Dec '10

30 Dec '10

--- boost/uuid/seed_rng.hpp | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-)

2 1

QuickBook question
by Marshall Clow 29 Dec '10

29 Dec '10

I am working one some quickbook docs and they start out like this: > [library The Boost Algorithm Library > [quickbook 1.5] > [id algorithm] > [dirname algorithm] > [purpose Library of useful algorithms] > [category algorithms] > [authors [Clow, Marshall]] > [copyright 2010 Marshall Clow] > [source-mode c++] > [license > Distributed under the Boost Software License, Version 1.0. > (See accompanying file LICENSE_1_0.txt or copy at > [@http://www.boost.org/LICENSE_1_0.txt]) > ] > ] > Later (in the same file), I want to drop a link. So, I write: > [link algorithm.all.none_of_if none_of_if1] (where all is a section, and none_of_if is a heading in that section) but this doesn't work! Some investigation shows that > [link the_boost_algorithm_library.all.none_of_if none_of_if2] does work - even though the documentation says that the first element in the link is the document id, not the document title. Am I misunderstanding something? -- Marshall

2 2

Inspection script not running?
by John Maddock 29 Dec '10

29 Dec '10

Title says it all really, Cheers, John.

2 1