August 2014 - Boost - lists.preview.boost.org

[mpl] Proceeding with modularization
by Andrey Semashev 20 Aug '14

20 Aug '14

Hi, Now that 1.56 is out I believe we can proceed with modularization. I'd like to complete Boost.MPL decomposition into Boost.MPL.Core and Boost.MPL that I started some time ago. I can see two ways of doing this: I can keep the two parts in the separate git repositories or I can move Boost.MPL.Core to a sub- library within the same git repo. I'd prefer the second approach since this allows to create a single pull request and keeps the library history in one place, but maybe there are other considerations?

3 3

Re: [boost] [Concepts] Definition. Was [GSoC] [Boost.Hana] Formal review request
by Brian Wood 19 Aug '14

19 Aug '14

Niall Douglas writes: > > Such a design would not make good use of hardware DMA such that use > of the CPU can be completely avoided for bulk data transfers. To > invoke hardware DMA, all i/o must be done from 4Kb aligned memory > buffers. Indeed AFIO can enforce that for you because it's so > important to performance. > > Generally if you're bothering to use async i/o, you have no interest > in *ever* copying memory. That rules out all serialisation usually :) How about calling it OS-level serialization? I don't think the need for message framing goes away. Somehow the length of a message has to be calculated and put in place. -- Brian Ebenezer Enterprises - In G-d we trust. http://webEbenezer.net

1 0

GSoC: Pipeline status
by Benedek Thaler 19 Aug '14

19 Aug '14

Hi, As GSoC is coming to its end, I'd like to present you the Pipeline library: code: https://github.com/erenon/pipeline documentation: http://erenon.hu/pipeline I'll continue to work on this project, the next big thing is to implement a coroutine based execution. I hope using cooperative scheduling will enable a much better throughput/latency combination over the nice and easy to use interface we currently have. Thanks, Benedek

1 0

[serialize] Noncopyable errors in the latest 'master'
by Edward Diener 19 Aug '14

19 Aug '14

In testing out libraries which use Boost PP, with clang as the compiler, I am seeing a great deal on 'master' of: > ..\..\..\boost/archive/detail/basic_serializer.hpp:35:7: error: 'operator=' is a private member of 'boost::noncopyable_::noncopyable' > class basic_serializer : Does anybody know what this is all about ? It is occuring in a number of libraries using serialization.

1 0

[GSoC] [Boost.Hana] Status update
by Louis Dionne 18 Aug '14

18 Aug '14

Dear Boost, Summer of code is coming to its end, and I think this is a good time to give an update about Hana and what's going to happen next. In early august, I asked for a review of Hana. As a result, I got many comments and questions which helped me improve the library and also take the pulse of the C++ community. Ever since the review, I've been working on improvements and modifications to the library and its documentation. Thanks to all of those who gave comments, and especially to those who experimented with the library and reported problems/thoughts. My initial (personal) objective was to push a new library in Boost by the end of the summer. I now realize that this is unlikely to happen given the amount of work that's left and the amount of time on my hands. That being said, I still think Hana solves the problem of C++ metaprogramming in a very elegant way, and I'm not giving up on its development because GSoC ends. Hence, I will continue working on the project as much as my studies allow it and I plan to ask for informal reviews during fall. Eventually, I still hope to fulfil my initial goal of making Hana available to a wide audience via Boost. I would also like to mention that I will be presenting Hana at CppCon 2014 in september; consider attending if you followed the project this summer. Finally, I would like to thank everyone that was involved with making my participation to GSoC possible; I worked hard, had a lot of fun and I think my work (or at least derivations of it) will be useful to Boost and the wider C++ community given enough time. Regards, Louis

1 0

Debugger Visualizers discussion
by Antony Polukhin 18 Aug '14

18 Aug '14

Hi all, There is project that contains a bunch of XML files. Those files can be used to improve debugging on MSVC: For example: * using this project debug view of `boost::variant<int, float> i(1);` will show `1` instead of raw memory and internals of boost::variant. * optional<boost::container::string> will show string content or `none` instead of ram memory and some unreadable internals Maybe we should embed those files into the Boost? It would be great to have some magic bjam command `b2 install-debug` after which VisualStudio becomes aware of Boost classes. Files for pretty debugging could be also written for GDB. Will it be useful? Link o the project: https://github.com/KindDragon/CPPDebuggerVisualizers Files are located at https://github.com/KindDragon/CPPDebuggerVisualizers/tree/master/VS2013/Vis… -- Best regards, Antony Polukhin

3 3

[git] [filesystem] [system] [timer...] Remove useless SVN branches?
by Beman Dawes 18 Aug '14

18 Aug '14

Boost library repos converted from Subversion are littered with branches created by the conversion process that seem to me to be quite useless. For example, Boost filesystem has 65 branches, yet only 10 or so have any usefulness I'm aware of. A lot of the useless branches have names that begin "sandbox" or "svn-". I go back to the subversion repo when I want to look at pre-git-conversion history. Is there any point in retaining these branches? Does anyone object if I remove these branches from the libraries where I am the maintainer? What are other maintainers doing about stale branches? --Beman

3 2

clang and variadic macro warnings revisited
by Edward Diener 18 Aug '14

18 Aug '14

After making pull requests for a number of Boost libraries in order to reduce the warnings from compilation from clang using the Boost PP library, John Maddock suggested a specific solution, as Niall Douglas had suggested a general one before, that changes in the Boost PP library could get rid of these extraneous warnings. Originally I thought this would be too difficult and intrusive in Boost PP but after reading John's specific suggestion I decided to try to implement it. The suggestion John made was to surround code in Boost PP which uses variadic macros with: # ifdef __clang__ # pragma clang diagnostic push # pragma clang diagnostic ignored "-Wvariadic-macros" # endif // variadic macros # ifdef __clang__ # pragma clang diagnostic pop # endif This was actually easier than expected since all code in Boost PP which uses variadic macros is surrounded by: # if BOOST_PP_VARIADICS #endif // eventually Of course there are many files changed by this but overall it is very straightforward. This technique does work for the diagnostic "-Wvariadic-macros". Hooray ! Unfortunately the second situation involving warnings, where clang warns about: warning: empty macro arguments are a C99 feature [-Wc99-extensions] coming from what I only see as a specific place in Boost PP in variadic/elem.hpp is not conducive to the same technique. In other words surrounding that one macro with: # ifdef __clang__ # pragma clang diagnostic push # pragma clang diagnostic ignored "-Wc99-extensions" # endif #define BOOST_PP_VARIADIC_ELEM(n, ...) BOOST_PP_CAT(BOOST_PP_VARIADIC_ELEM_, n)(__VA_ARGS__,) # ifdef __clang__ # pragma clang diagnostic pop # endif does not eliminate the warning at all ( I have reported this on the clang developers mailing list ). And since that macro is heavily used in Boost PP this warning occurs quite often when compiling Boost library code using Boost PP wih clang. The only other intrusive Boost PP solution to this problem is to treat clang like gcc as far as Boost PP variadic macro code is concerned: if the end-user compiles with one of the 'std=' dialects which explicitly supports variadic macros, such as 'std=c99', 'std=c++0x, 'std=c++11', then variadic macros are turned on in Boost PP for clang, else they are turned off. Do realize that both gcc and clang support variadic macros even when one of the 'std=' dialects are not chosen, and that in fact clang has always supported variadic macros. Also changing Boost PP support for clang and variadic macros would be a regression from the way it is now where it is always supported. In retrospect, when I changed Boost PP from never supporting variadic macros for clang ( Paul's original conservative choice ) to always supporting variadic macros for clang, I gave in to an outcry that clang users should have variadic macro support in Boost PP. It would have been wiser to follow the gcc method in Boost PP. But that is past and we need a good solution now. Thoughts ? Solutions ? Brickbats <g> ?

3 3

[endian] Was: [Boost-testing] All runners on develop are failing.
by Rene Rivera 18 Aug '14

18 Aug '14

---------- Forwarded message ---------- From: Tom Kent <lists(a)teeks99.com> Date: Fri, Aug 15, 2014 at 8:13 PM Subject: [Boost-testing] All runners on develop are failing. To: Running Boost regression tests <boost-testing(a)lists.boost.org> It is looking for the endian library...which I have, but is still failing. error: Unable to find file or target named error: '../libs/endian/' error: referred to from project at error: '.' I did a: git checkout develop git pull git submodule update --init on my repository, which caused the previously empty endian directory to be populated, but still getting the error. Thoughts? Tom _______________________________________________ Boost-Testing mailing list Boost-Testing(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-testing -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

3 2

[algorithm] Interface for longest_increasing_subsequence (proposal)
by Curdeius Curdeius 17 Aug '14

17 Aug '14

Hi all, I want to propose a new algorithm to Boost.Algorithm: a well-known longest increasing subsequence <https://en.wikipedia.org/wiki/Longest_increasing_subsequence> (LIS for short). Just for a short info: given a sequence like ( 7 )( 2 )( 8 )( 1 )( 3 )( 4 )( 10 )( 6 )( 9 )( 5 ), the result will be a subsequence ( 1 )( 3 )( 4 )( 6 )( 9 ). The proposed algorithm has O(n log n) time and O(n) complexity, where n - sequence length. I have created a pull request https://github.com/boostorg/algorithm/pull/7. I have however some questions. I am not sure what should be the exposed interface. For the moment, I propose (I base it on the existing interface from algorithm/searching functions): - class longest_increasing_subsequence with methods - operator() - *compute_length* (a better name may be needed) - free function longest_increasing_subsequence_length returning the length of the LIS and taking as arguments: - a pair of iterators (first, last) - or a range - and (optionally) a comparison predicate - *free function longest_increasing_subsequence_search taking the same arguments as _length function and an Output Iterator or a tag.* - free function make_longest_increasing_subsequence taking a range and optionally a comparison predicate and returning corresponding longest_increasing_subsequence object The problematic part is the underlined longest_increasing_subsequence_search function and the way in which I should return the resulting LIS. I see possibilities like the following: 1. taking an OutputIterator argument: the best for me, generic and the most flexible solution, but there is a performance hit as we cannot reserve the memory for the output LIS. 2. returning a container with the subsequence, e.g. std::vector<value_type>: no performance hit with small objects or if the user needs a vector of values. However, not really generic as the user would be obliged to use a vector (in the aforementioned example). 3. returning a container with iterators, e.g. std::vector<InputIter>: good if a container of iterators is acceptable, it has a benefit of avoiding possibly expensive copies of sequence elements. As 2., not really generic. 4. -taking a container as an argument: the worst solution in my eyes, there is no way nice way to write generic code that handles all containers. I have coded possibilities 1-3. Thank you in advance for all the comments, corrections etc. Regards, Marek Kurdej

1 0

[testing] Proposal - regression tests results display upgrade
by Adam Wulkiewicz 17 Aug '14

17 Aug '14

Hi, Currently test failures of a new feature are always marked with the yellow color and a word 'fail'. Would it make sense to use some more descriptive color/naming scheme? In particular it would be nice to distinguish between the actual failures and the situations when the compilation of a test took too much time or an output file was too big. Would it make sense to also distinguish between compilation, linking and run failure? E.g. instead of a simple 'fail' use 'fail (c)', 'fail (l)' and 'fail (r)'. E.g. the color for actual failures could be orange and for "others" could stay yellow. Furthermore, currently it's impossible to see the results of compilation if a test passed. Would it make sense to change it? This would allow us to view and fix warnings even if the test was passed. The test passed with warnings could be marked e.g. with 'pass (w)' or the color could be set to some between yellow and this light green. If you directed me where it should be done I could try to change it. I'm guessing it would require modifying some script generating the page and some CSS? Regards, Adam

6 10

Re: [boost] [units] Pull request and RFC: information unit definitions for Boost.Units
by Curdeius Curdeius 16 Aug '14

16 Aug '14

Hi all, I considered proposing myself as a maintainer of Boost.Units for some time, but considered too inexperienced to do it on my own. Therefore I am happy that Jürgen proposed himself and I would like to help him by being a co-maintainer if it is possible and needed. As for me, I have been using the library for at least 3 years now and would like to contribute to it. For the moment, I mainly proposed small bug fixes to Boost libraries. Regards, Marek Kurdej -- Ph.D. candidate Hi Matthias, > Am Donnerstag, 31. Juli 2014, 11:25:35 schrieb Matthias Schabel: > > I would love to pass the responsibility for ongoing maintenance of > > Boost.Units on to a new torchbearer. > First a "Thank you very much" for the great library. > > My time to dedicate to this project > > has dwindled to essentially zero, and I believe it would be best for all > > involved to enlist some fresh blood, ideally someone who is using the > > library on a regular basis... Any volunteers? > As no one else has stepped up, I'd like to volunteer. I've been using > Boost.Units with great success for the last five years. For my recent Boost > contributions see https://github.com/jhunold. Mostly small bugfixes, > Boost.Build support over the last decade. > Boost.Units is getting more attentions now as c++11 is getting momentum. > Personally, I'd like to get > https://github.com/boostorg/units/pull/5 > into 1.57. > > As far as Erik's request, I don't have any objections to adding these to > the > > library, though I have not had time myself to verify that they are > correct. > It would be nice if one of the current maintainer could comment on this. > > >> Link to PR: https://github.com/boostorg/units/pull/3 > Seems obsolete, > https://github.com/boostorg/units/pull/6 > it the improved commit. > Feel free to ask questions. > Yours, > Jürgen > -- > * Dipl.-Math. Jürgen Hunold ! > * voice: ++49 4257 300 ! Fährstraße 1 > * fax : ++49 4257 300 ! 31609 Balge/Sebbenhausen > * jhunold(a)gmx.eu ! Germany

1 0

[config][type_traits] Support for final keyword in Boost?
by John Maddock 16 Aug '14

16 Aug '14

I have a couple of pull requests to add support for the final keyword in Boost.Config and an is_final trait to type traits, so my questions are: 1) Are these generally useful? 2) Do we need BOOST_NO_CXX11_FINAL_CLASS and BOOST_NO_CXX11_FINAL_OVERRIDE or will a single BOOST_NO_CXX11_FINAL do? 3) The trait appears to be unimplementable via SFINAE (anyone care to prove me wrong?) which means it's only usable on GCC and lookalikes which have an __is_final intrinsic. Should we still add this to Boost or is the answer "use GCC's native features" given that a Boost version effectively adds nothing? Thanks! John.

4 3

[docbook] Add detail namespace to Reference section
by Benedek Thaler 16 Aug '14

16 Aug '14

Hi, I have the following example code fragment: namespace boost { namespace detail { /** documentation */ class implementation {}; } // namespace detail using detail::implementation; } // namespace boost I'd like to have the class `implementation` show up in the Reference section generated by doxygen+docbook-xsl+boost build+docbook+... . Issue is, every namespace which name contains the string "detail" is excluded somehow. The direct output of doxygen has it, but not the docbook version. I couldn't find any EXCLUDE_SYMBOLS setting (it's not set in the doxyfile)or anything relevant amongst the docbook-xsl params or in the boost build scripts. Google search term "doxygen ignores detail namespace" and the like did not help. Adding documentation to the using statement and/or TYPEDEF_HIDES_STRUCT did not help. Thanks, Benedek

2 3

Should intrusive_ptr copy constructors be declared noexcept?
by James Yonan 16 Aug '14

16 Aug '14

Currently, intrusive_ptr's copy constructors are not declared noexcept. This disables the std::function small object optimization when intrusive_ptr objects are passed by value in a lambda closure. The attached patch tries to fix this by making the copy constructors conditionally noexcept, based on the noexcept status of the user-provided intrusive_ptr_add_ref function. James

1 0

[Boost][filesystem] imbue in global constructor crashes
by Peter Bindels 15 Aug '14

15 Aug '14

Hi all, When trying to use a global constructor to set the locale to utf8 for an entire project, the system failed to start due to global constructor ordering making the override happen before the std::locale object itself was initialized. Looking into the boost::filesystem code for imbue it seemed that there were a few different implementations, one of which could work. I've tried to keep the same side conditions and made this same lazy-initialization approach for all platforms, which should keep working in all the same conditions and also work for a global constructor overriding the locale. I've attached the patch. I hope that this change can be integrated into boost::filesystem or that we can have a constructive discussion on what solution we should be making. Best regards, Peter Bindels

2 1

[variant] static_visitor constructor error with boost 1.56.0 and g++ 4.5.1
by Richard Hadsell 15 Aug '14

15 Aug '14

I am trying to upgrade from Boost 1.55 to 1.56. I get this error with G++ 4.5.1: In file included from ...: .../boost/variant/static_visitor.hpp:52:24: error: 'boost::static_visitor<R>::static_visitor()' declared with non-public access cannot be defaulted in the class body .../boost/variant/static_visitor.hpp:53:25: error: 'boost::static_visitor<R>::~static_visitor()' declared with non-public access cannot be defaulted in the class body The OS is Linux Fedora 14. The compiler is gcc version 4.5.1 20100924 (Red Hat 4.5.1-4). I saw a similar problem in noncopyable.hpp, when upgrading to Boost 1.55. Andrey Semashev fixed it, and it looks like static_visitor requires the same fix. In static_visitor.hpp change line 51 from #ifndef BOOST_NO_DEFAULTED_FUNCTIONS to #if !defined(BOOST_NO_CXX11_DEFAULTED_FUNCTIONS) && !defined(BOOST_NO_CXX11_NON_PUBLIC_DEFAULTED_FUNCTIONS) I think this would be appropriate for any non-public defaulted functions, not just the ones that trip up code I happen to compile. Dick Hadsell 203-992-6320 Fax: 203-992-6001 Reply-to: hadsell(a)blueskystudios.com Blue Sky Studioshttp://www.blueskystudios.com 1 American Lane, Greenwich, CT 06831-2560 Follow Blue Sky Studios on Facebook <http://www.facebook.com/BlueSkyStudios> and Twitter <http://twitter.com/#%21/blueskystudios>

2 1

[1.56] PDF documentation available
by John Maddock 14 Aug '14

14 Aug '14

The PDF build of Boost's documentation for 1.56 is now available from the usual address: https://sourceforge.net/projects/boost/files/boost-docs/1.56.0/ Note that it includes only the subset of Boost libraries whose documentation is quickbook/docbook based. John Maddock.

6 8

[container] 1.56.0 flat_map static assert typo ?
by feverzsj 14 Aug '14

14 Aug '14

hi, list in <boost/container/flat_map.hpp>, the ctors seems to contain a wrong static assert: flat_map() : m_flat_tree() { //Allocator type must be std::pair<CONST Key, T> BOOST_STATIC_ASSERT((container_detail::is_same<std::pair<Key, T>, typename Allocator::value_type>::value)); } missing "const" ? regards. -- View this message in context: http://boost.2283326.n4.nabble.com/boost-container-1-56-0-flat-map-static-a… Sent from the Boost - Dev mailing list archive at Nabble.com.

3 3

Boost Library Writers Workshop at CPP Con 2014
by Robert Ramey 14 Aug '14

14 Aug '14

Have you been lurking out there in cyberspace all these years? Slaving away on your fantastic new library for boost? Stuck, discouraged, concerned about getting ripped apart on the mailing list? Well, this is for you !!!! The Boost Library Writers Workshop http://rrsd.com/blincubator.com/open-content/ . Share your library and get it critiqued by you peer group - other unsung authors battling the same daemons. Best of all you can attend this workshop without being registered at the conference - to attend this workshop all you have to do is ... attend!. That's right, it's absolutely free. That's right free. And not free like in free software. Really free like "free beer". (note: there is no free beer). So if you're in the area and you want to get your library off your chest because it's weighing you down, keeping you awake, ruining your love life attend this workshop. Spaces are limited so signup now at the above web page. Robert Ramey -- View this message in context: http://boost.2283326.n4.nabble.com/Boost-Library-Writers-Workshop-at-CPP-Co… Sent from the Boost - Dev mailing list archive at Nabble.com.

1 0

[GitHub] Transferring a new library's repo to boostorg?
by Beman Dawes 14 Aug '14

14 Aug '14

https://svn.boost.org/trac/boost/wiki/NewLibFromGit describes "Creating a new library from an existing Git repository". It seems to cover the case where the existing git repo isn't in the exact boost branch organizational structure required. The Boost Endian library is ready to be moved to boostorg on GitHub so it can be tested on develop, checked by release managers, and otherwise readied to be added to 1.57.0. It is already on GitHub and AFAIK has the usual boost branch and directory structure. GitHub has a mechanism to transfer a repo between users or organizations. This would seem the easiest and most direct approach for boost endian. Is there any reason not to give it a try, and if it works as advertized then add it to https://svn.boost.org/trac/boost/wiki/NewLibFromGit as an additional option? --Beman

3 6

Comments on generic matrix library with specific 3D operations -> news and documentation
by Fabien Picarougne 13 Aug '14

13 Aug '14

Hello Boost, After some holidays, I have finished to implement some functionalities of my matrix library (https://bitbucket.org/fpicarougne/template-matrix-library) I have also begin to write some documentation (http://www.polytech.univ-nantes.fr/wikinfo/gmt/) relatively poor at this stage but sufficient (I think) to give you a real idea of the library. Globally, functionalities are closed to Boost::QVM but implementation strategy is different and uses more C++11 techniques as variadic templates for example. I must also say that I took over the syntax of the swizzling operator from this library. I would appreciate if you could give me feedback on this code. The code has been compiled on MSVC2013 and gcc 4.8 without warning (except those that come from Boost.Preprocessor) by using these parameters (in gcc): -Wall \ -Wextra \ -pedantic \ -Wmain \ -Wundef \ -Wsign-conversion \ -Wuninitialized \ -Wshadow \ -Wunreachable-code \ -Wswitch-default \ -Wswitch-enum \ -Wcast-align \ -Wmissing-include-dirs \ -Winit-self \ -Wfloat-equal \ -Wdouble-promotion \ -Wzero-as-null-pointer-constant \ -Wunused-parameter Current namespace is temporary (and inherited from old library) and will be changed in the future. Code needs also to be separated in more header files (this work will be done soon). I have kept in the git repository the "test.cpp" file which allows me to test some functionalities (and incidentally allow you to evaluate the usage syntax). Fabien

2 1

[git] confusion setting moduler-boost to 'master'
by Edward Diener 13 Aug '14

13 Aug '14

Somehow my local modular-boost is confused. I have set the top-level to 'master' and each of the sub-modules to 'master'. Yet git still shows it wants to commit each of the submodules at the library level as all are marked dirty. So I try: git reset --hard master git submodule foreach --recursive git reset --hard master But still the end result is that git has marked each submodule as 'dirty' and needing to commit. Anybody have any ideas how I straighten out this mess ?

3 4

build of 1.56.0 has no tools/build/v2
by Roger Martin 12 Aug '14

12 Aug '14

Hi, The build has changed or the docs http://www.boost.org/doc/libs/1_56_0/more/getting_started/unix-variants.htm… are correct and tools/build/v2 is missing There is a bootstrap.sh in tools/build and I used it successfully And then built the libs I needed so I'm guessing the docs on getting started aged slightly. I didn't need tools/build/v2/user-config.jam this time but do see the bootstrap build of bjam produces a ./build.tool/share/boost-build/example/user-config.jam For new users maybe a little cleanup on getting started would be helpful

1 0

[git][regression] Regression is not a submodule
by Adam Wulkiewicz 12 Aug '14

12 Aug '14

Hi, I noticed that Regression isn't setup properly as a submodule of modular boost unlike other tools. It isn't present on a list displayed with: git submodule status nor in .gitmodules $ grep 'tools' .gitmodules path = tools/build path = tools/bcp path = tools/inspect path = tools/boostbook path = tools/quickbook path = tools/litre path = tools/auto_index path = tools/boostdep There is a module Regression: https://github.com/boostorg/regression/tree/develop and there is also a non-submodule directory in boostorg: https://github.com/boostorg/boost/tree/develop/tools/regression Unlike the former, the latter seems to be actively developed. I assume that the code should be contributed to the boostorg directly? Is there a reason for this? Regards, Adam

2 1