- Boost - lists.preview.boost.org

shared_ptr ordering
by Matt Bergin 22 Apr '13

22 Apr '13

Back in 2008, changing shared_ptr operator< to compare by pointer value rather than owner was discussed: http://lists.boost.org/Archives/boost/2008/10/143175.php There seem to be convincing arguments for both implementations (aliasing, consistency with equality operator etc). Is boost::shared_ptr intended to be a compliant C++11 std::shared_ptr? If so, will shared_ptr be changed to comply with [util.smartptr.shared.cmp] or is the ensuing breakage too high a price to pay for 100% compliance? Thanks Matt

2 1

[ublas] GSoC 2013 Proposal - uBLAS Math Library and Performance Optimization
by Sathyam Vellal 22 Apr '13

22 Apr '13

Hello everyone. I've been in touch with Mr. David Bellot and Mr. Nasos Iliopoulos and with their guidance and support I've written this proposal for Boost.uBLAS. The proposal can be viewed here -http://goo.gl/Qo7Ry. (This is a link to a google doc; a pdf version of the proposal). Thank you -- Sathyam M Vellal P.E.S. Institute of Technology, Bangalore, india

1 0

[config][mpl][type_traits][integer] BOOST_STATIC_CONSTANT should use constexpr
by Vicente J. Botet Escriba 22 Apr '13

22 Apr '13

Hi, BOOST_STATIC_CONSTANT could be improved to make use of constexpr. This is needed for instance to make integral_constant compliant with C++11. The patch consists in changing const by BOOST_CONSTEXPR_OR_CONST in boost/config/suffix.hpp // BOOST_STATIC_CONSTANT workaround --------------------------------------- // // On compilers which don't allow in-class initialization of static integral // constant members, we must use enums as a workaround if we want the constants // to be available at compile-time. This macro gives us a convenient way to // declare such constants. # ifdef BOOST_NO_INCLASS_MEMBER_INITIALIZATION # define BOOST_STATIC_CONSTANT(type, assignment) enum { assignment } # else # define BOOST_STATIC_CONSTANT(type, assignment) static *BOOST_CONSTEXPR_OR_CONST* type assignment # endif and in each out of line static constant definition as in template <int b1, int b2> bool *BOOST_CONSTEXPR_OR_CONST* ice_eq<b1,b2>::value; The main problem is that this is a *breaking change*: the list of concerned libraries is quite big (mpl, type_traits/integer_traits, ...) and users of BOOST_STATIC_CONSTANT would need to make this modification also. An alternative is to define a BOOST_STATIC_CONSTANT_11 that includes this modification and update the Boost libraries that could make use of the new macro without introducing a breaking change. Library providing a template class containing a field with such a change that can be specialized by the user would need to either announce a breaking change or just don't use the new macro. While this mean more changes the backward compatibility can be mastered. What do you think, it is worth try to fixing this? Best, Vicente

4 17

GSOC - Boost.SQL and Boost.Thread interest
by Guillaume Nuel 22 Apr '13

22 Apr '13

Hi Boost C++ Team ! My name is Guillaume NUEL. I'm a French student in computer security (fourth year in computer science). I'm actually making a game witch uses Boost (Thread, Signal,...). I also use CouchBase for database. I really would like mixing my game development and Boost development. My project proposal Idea : I'm mainly interesting in adding a new module: Boost.SQL for managing database: MySQL, Couchbase, … If the project proposal is not suitable, I have others interests. I would also be interesting in adding new functions to Thread abilities. So I am interested in the following topics : - Boost.Thread/SyncQueue: Provide boost::sync_bounded/unbounded_queue implementations base on the C++ Concurrent Queues proposal. Trying an adaptation of the Boost.LockFree queues to follow the same interface. - Boost.Thread/ThreadPool:Provide a boost::thread_pool class scheduling arbitrary functions based on Toward Boost Async 0.2 and extend the boost::async function to manage with this new class. Projects mentored by Vicente J. Botet Escriba I'm an expert in C++ and I have good skills in OpenMP and Cuda ( GPU ) and so in parallelization of code. Homepage: http://immmortal-pc.info/Boost.html My CV is available at http://immortal-pc.info/cv.html Thank you for taking the time to read my e-mail, I look forward to receive a response from you. With regards, NUEL Guillaume

2 1

GSoC: More mentors?
by Boris Schaeling 22 Apr '13

22 Apr '13

From Monday on students can apply for GSoC projects. Mentors can already apply at <https://google-melange.appspot.com/gsoc/homepage/google/gsoc2013>. So far we have four possible mentors (for uBLAS, odeint, graph and benchmark). Are there others who would like to mentor? Then please pick a project from <https://svn.boost.org/trac/boost/wiki/SoC2013> or propose your own (like I did a few days ago: <http://article.gmane.org/gmane.comp.lib.boost.devel/240251>). Applying as a mentor doesn't mean that you must mentor a student. Among others it all depends on if and how many student applications we will receive. But having a few more mentors just in case would be nice. :) Thank you, Boris

4 6

media > audio project (gsoc or future boost)
by adrien courdavault 22 Apr '13

22 Apr '13

Hi everyone. I read the project list for the GSOC 2013 and in Ideas I've seen the mention of Audio. I would like to know who is interested in this and if someone could give some details about this line : 'Audio library (OpenAL/FMOD/etc)' see here :https://svn.boost.org/trac/boost/wiki/SoC2013#Media I would like to know if this might be related to this: http://comments.gmane.org/gmane.comp.lang.c%2B%2B.isocpp.proposals/310 Thank you in advance. Adrien

6 11

Writing timer output to file
by Mohamed Nemri 22 Apr '13

22 Apr '13

Hi, I want customize boost::timer::auto_cpu_timer to put results in a logging stream(file, console,...). What is the best way to do that? Thanks --

1 0

[GSoC] Boost.Multiprecision project
by Jyotesh Choudhari 22 Apr '13

22 Apr '13

Hello, My name is Jyotesh Choudhari and I am from India. I am pursuing my Masters in Computer Science and Technology from Indian Institute of Technology (IIT) Bombay. I went through the ideas page for GSoc 2013 and would like to work on Boost.Multiprecision (cpp_bin_float) project. Looking at the project description, I think I have understood what needs to be implemented in this project. I have also gone through the current code cpp_bin_float.hpp and I have understood some part of it. I want to know how detailed should my proposal be? What details should I include in my proposal? What should be the deliverables for the project? Thanking you. Regards, Jyotesh Choudhari

2 1

[GSoC] Boost.multiprecision project
by Flórián Deé 22 Apr '13

22 Apr '13

Hello, I'm Florián Deé from Hungary, studying at Budapest University of Technology and Economics. I'm really interested in the Boost.Multiprecision project for GSoC 2013. A little background on myself, I'm an undergraduate software engineering student mostly into C and C++, with interests in low level programming and maths. I'm currently working on a project as an outsider member of the dorm's Coder team (a poker framework for AIs), and I'm also a "consultant" for students with their C/C++/Maths assignments (mostly calculus and linear algebra). As a result, I don't really have much free time till the exams, but I've planned, seen, written and debugged lots of different projects with different coding styles. Going through the project description, I think my skill set and interests would fit in nicely, with the exception of fast Fourier transforms, which I'm only mildly familiar with, but I'd gladly read any needed extra material till summer. So I kindly seek your guidance to help me go about my application for the GSoC 2013. Looking forward to your reply. Regards, Flórián

2 1

[ublas] bugfix for matrix operation /
by Mario Mulansky 22 Apr '13

22 Apr '13

Hi, I wanted to add an elementwise divide operations for ublas matrices using operator / Unfortunately, there is operator/ defined for matrix and integral types which hides any other operator definition (matrix_expression.hpp l. 3388) and thus makes it impossible to add this functionality. I would like to add an enable_if statement there that checks if the second argument is convertible to matrix::value_type, as is done for other operators. Who would be responsible for this small change? I can provide the according patch, it's almost trivial. Mario

4 7

[gsoc-2013] Physics Library Abstraction Layer
by Preston Hamlin 22 Apr '13

22 Apr '13

I am working on putting together a draft proposal for the project. Is there a preferred format desired? I was going to use a PDF since they are quite portable. I am somewhat confused as to the scope of the project, due to the previous posts in this thread. Is the library to be built upon a specific physics library (PhysX for instance) by utilizing existing code in the respective SDK to abstract away much of the object management? Or is it meant to be built from the ground up to create generic objects, which can then be translated into an arbitrary format for an existing library? Thank-you, -Preston

2 3

library to support async/await pattern
by Valentin Milea 21 Apr '13

21 Apr '13

Is there interest for a coroutine-based library to make asynchronous APIs easier to deal with? Typically, for each task there is cruft: chaining callbacks, managing intermediate state and error codes (since exceptions don't fit this model). The code flow get inverted and becomes difficult to follow. Recent versions of F# and C# solve this problem. They implement an await operator that effectively suspends the executing method until a task completes. The compiler takes care of transforming subsequent code into a continuation. Everything runs on the main thread, with asynchronous methods spending most time awaiting. N3328 proposes resumable functions of this kind in C++. For an immediate solution we could leverage the Boost.Context/Coroutine library. The resulting code may look like this: try { task = do_a_async(...) // yield until task done task.await(); } catch (const some_exception& e) { // exceptions arrive in awaiting context } // normal code flow for (auto& task : tasks1) { task.await(); } taskAny = await_any(tasks2); taskAny.await(); ... There needs to be a representation for Awaitable tasks (similar to std::future but non-blocking). The other requirement is to have a Scheduler (run loop) in order to weave between coroutines. Benefits: - normal code flow: plain conditionals, loops, exceptions, RAII - algorithm state tracked on coroutine stack - async tasks are composable - any async API can be wrapped Cons: - must wrap async APIs (e.g. Boost.Asio) - needs std::exception_ptr to dispatch exceptions - stackful coroutines are sometimes difficult to debug I wrote an open-source library that does this: https://github.com/vmilea/CppAwait. It's far from Boost style but the concept looks sane. For a comparison between async patterns please see: https://github.com/vmilea/CppAwait/blob/master/Examples/ex_stockClient.cpp Making a Boost version would involve serious redesign. So is this worth pursuing? Thanks!

5 9

Re: [boost] GSoC: More mentors?
by Christopher Kormanyos 21 Apr '13

21 Apr '13

> From Monday on students can apply for GSoC projects. > Mentors can already apply at > <https://google-melange.appspot.com/gsoc/homepage/google/gsoc2013>. > So far we have four possible mentors (for uBLAS, odeint, > graph and benchmark). Are there others who would like > to mentor? Then please pick a project from > <https://svn.boost.org/trac/boost/wiki/SoC2013> or propose > your own. > Thank you, Boris Hello, Christopher Kormanyos (who is me) just signed up as a potential mentor for the GSoC projects with Boost.Math and Boost.Multiprecision. We have already had contact with some good candidates. Thank you. Sincerely, Chris.

1 0

GSOC - Boost.uBLAS interest
by deck léa 21 Apr '13

21 Apr '13

Hi Boost C++ Team ! My name is Lea Deck, I'm French and I study in fourth year in computer science (security in computer science), but I really like mathematics. So I look for projects that mix the two. I would be interesting in adding new mathematics properties to your uBLAS libraries, like Hadamard, Toeplitz and Block matrices, the generation of matrices, Strassen multiplications and possibly new algorithms. You also alluded to the Bernoulli numbers, but I don't see any mention in the list of new features. Is the project still available? If so, should its implementation also contain the Bernoulli polynomials? About the question of optimisation, I have some skills in Cuda (GPU) and OpenMP and so in parallelization of code. So I am able to produce a correct parallelized code. Also, my knowledge of C++ is beginning to spread to the use of C++11 ( including unordered structure map). One last question, what do you mean by "Pretty display" in your list of new features? My CV is available here <http://dla.immortal-pc.info/>. ( http://dla.immortal-pc.info/) Thank you for taking the time to read my e-mail, and I look forward to receive a response from you. With regards, DECK Léa

1 0

[GSoC 2013] SVN?
by Vicente J. Botet Escriba 21 Apr '13

21 Apr '13

Hi, would all the GSoC projects be stored under SVN? Best, Vicente

3 5

Qustion about GSOC Project
by Hardy Huang 20 Apr '13

20 Apr '13

Hi, there. I am a graduate student in Peking University. I have known Boost library for long. And I used it for several times. This year I wanted to get involved in GSOC, and I was happy when I see Boost in the list, it is my dream and it will be honor to contribute to such a powerful library. Since I am good at algorithm and data structure, knowing that there are some idea projects relative with algorithm and data structure, like radix sort, trie, I want to know something detail about the plan to develop algorithm and data structure. Thanks, I am expecting to receive some useful replies.

4 6

Re: [boost] [config][mpl][type_traits][integer] BOOST_STATIC_CONSTANT should use constexpr
by Vicente J. Botet Escriba 20 Apr '13

20 Apr '13

Le 20/04/13 14:33, Vicente J. Botet Escriba a écrit : > Le 20/04/13 13:18, Antony Polukhin a écrit : >> 2013/4/20 John Maddock<boost.regex(a)virgin.net>: >> <...> >>> No that's correct as is: >>> >>> * if BOOST_NO_INCLASS_MEMBER_INITIALIZATION is defined, then the constant is >>> really an enum and there should be no out-of-line definitions. >>> * if BOOST_NO_INCLASS_MEMBER_INITIALIZATION is *not* defined, then the >>> constant *should* have the out-of-line definition in case it's address is >>> taken/used. >> Oh, now I see, thanks. >> >>> As for the original question, this is such a pervasive breaking change, I >>> think a new macro is inevitable I'm afraid :-( >> New macro will do exactly what the old macro did, but using C++11 >> feature. Code in all Boost libraries will be changed to use new macro. >> So the old macro will be required only for users code that use macro >> and attempts to take address of constant... >> >> To me, it looks less obscuring to have one macro and a note for users >> to use BOOST_CONSTEXPR_OR_CONST for out-of-line definition; than two >> macro (which one shall be used by default?) that do the same thing but >> very slightly differ in details. >> > Just to add some information about some uses > > grep -r BOOST_NO_INCLASS_MEMBER_INITIALIZATION * | wc -l > 113 > grep -r BOOST_STATIC_CONST * | wc -l > 6598 > The figures are a little bit better :) grep -r BOOST_STATIC_CONSTANT * | grep -v svn | wc -l 3030 grep -r BOOST_NO_INCLASS_MEMBER_INITIALIZATION * | grep -v svn | wc -l 57 Best, Vicente

1 0

Trac: Add ref library
by Vicente J. Botet Escriba 20 Apr '13

20 Apr '13

Hi, Boost.Ref is not included as a library on the Trac system. Coul dit be added? Best, Vicente

1 0

[log] Permission to merge to trunk
by Andrey Semashev 20 Apr '13

20 Apr '13

Hi, I'd like to ask for permission to merge Boost.Log v2 to Boost trunk. The library has passed the review back in 2010 and was provisionally accepted [1]. It has been reworked since then, and in a few recent posts I described the changes [2][3]. [1] http://lists.boost.org/boost-announce/2010/03/0256.php [2] http://thread.gmane.org/gmane.comp.lib.boost.devel/237748 [3] http://thread.gmane.org/gmane.comp.lib.boost.devel/239313

16 43

Google Summer of Code - 2013: Boost C++ Libraries
by Ketan Akade 20 Apr '13

20 Apr '13

Hi, I am Ketan Akade, Graduate student at University of Southern California. I am majoring in Computer Science. I went through the project ideas of Boost C++ Libraries, and would like to work on it. I am particularly interested in the *Boost.uBLAS* project idea which is unifying representation of vector and matrices into a unique matrix class. Could you please let me know further details of the same? What steps should I take to start on this? PFA project proposal in which I have mentioned my academic and technical background. Thanks and regards, Ketan Akade Graduate Student, University of Southern California.

1 0

[gsoc 2013] chrono::date
by Anurag Kalia 19 Apr '13

19 Apr '13

I saw the chrono::date library proposal in the ideas page of gsoc 2013. It already seems quite solid and fleshed out; I am confident that I have understood it and can carry it out. Here it is for reference: <http://tinyurl.com/ct4tc3m> I have a few questions though. 1) Is this proposal in its final stages? The Boost chrono library docs mention in appendices that they intend to implement it in later releases. What I want to ask is if I am allowed to propose changes in the library if I want to. For example, the proposal proposes that adding a year to 'Feb 29' (leap year) should result in 'Feb 28 of next year' (non-leap year). I tend to think this is counter-intuitive. The author already gives the excellent idea of using a syntax like "last/february/2013" which also covers the case of anniversaries falling on 'Feb 29' pretty well. Moreover, it is a pretty big assumption that a person would want to celebrate the 'Feb 29' anniversaries on 'Feb 28' and not on 'Mar 1'. Moreover, adding a month to '31 January' results in an invalid date (31 February). Why shouldn't we mirror the behaviour for years? We can just make a library function to cover this issue. In short, am I allowed to propose such changes? 2) How should I apply for gsoc now that I have found my area of interest? What is the last date of application? Shall I post a draft of my application here first for review before submitting it officially? Am I allowed to? 3) I have my final examinations scheduled during three weeks in the coming month of May after which I am totally free to pursue gsoc internship. Though I will use this time to research for design and implementations of other similar libraries, I would still need to channel much time into my examinations. How am I supposed to circumvent this conflict, given that I do have ample time to dedicate to my internship later on? Thanks.Anurag.

2 1

[fusion] [range] Interest in unified functions
by Antony Polukhin 19 Apr '13

19 Apr '13

Hi, I was playing with Boost.Fusion and wrote this simple function: template <class Sequence> std::string stringize(const Sequence& seq) { std::string result; boost::fusion::for_each(seq, stringize_functor(result)); return result; } Which allows to convert many types to strings: boost::fusion::vector<cat, int, std::string> tup1(cat(), 0, "_0"); boost::tuple<cat, int, std::string> tup2(cat(), 0, "_0"); std::pair<cat, cat> cats; std::cout << stringize(tup1) << '\n' << stringize(tup2) << '\n' << stringize(cats) << '\n' But this example fails to convert the following : std::vector<cat> too_many_cats(20); All we need to do to allow conversions, is just add the following: #include <boost/range/algorithm/for_each.hpp> namespace boost { namespace fusion { template <typename Sequence, typename F> inline typename disable_if< traits::is_sequence<Sequence> , void >::type for_each(Sequence const& seq, F const& f) { ::boost::range::for_each(seq, f); } }} So now std::cout << stringize(too_many_cats) << '\n'; works and boost::fusion::for_each may work with any type, runtime length or compile time. We may also make a lot of other algorithms to make no difference between runtime and compile time classes: accumulate, find, find_if, count, all, none, begin, end, advance, distance, dereference, empty and others... The questions are: Is there interest in such features? Is there interest in adopting other algorithms and which ones? To what library shall those methods belong? -- Best regards, Antony Polukhin

1 0

[signals2] Uncombined signal
by Ivan Le Lann 19 Apr '13

19 Apr '13

Hi, I have a project where I use boost::signals2 in the spirit of the "Signal Return Values (Advanced)" section of the tutorial. That is, I have signals roughly equivalent to : boost::signals2::signal<float (), maximum<float> > sig; // slots produce values accumulated by the combiner Problem comes when I want multiple combiners on the same slots. Especially combiners of different types : You need to duplicate signals, and you need your connection setup code to know about used combiners. IMHO, having a basic signal class with no combiner makes sense. Attached is a POC patch to version 1.50, to add to boost::signals2::signal the method : slot_call_range all (BOOST_SIGNALS2_SIGNATURE_FULL_ARGS(BOOST_SIGNALS2_NUM_ARGS)) const (in boost/signals2/detail/signal_template.hpp) >From there, I think one could use any algorithm/accumulator directly on the signal.all() result. Implementation left aside, any feedback on this ? Regards, Ivan

2 4

GSoC project idea: Directory monitor for Boost.Asio (student wanted)
by Boris Schaeling 18 Apr '13

18 Apr '13

Here's another idea for a GSoC project: Turn the directory monitor from <http://www.highscore.de/boost/dir_monitor.zip> into a full-featured Boost.Asio extension. This includes understanding the source code of the existing directory monitor, getting familiar with Boost.Asio internals, enhancing the current interface, implementing the directory monitor on multiple operating systems, writing test cases and a documentation. This is a project which can be completed within a summer. It could be the right project for a student who is interested in Boost.Asio and wants to go through all steps from designing to documenting. Boris

1 0

bug in boost::container ??
by Olaf Peter 18 Apr '13

18 Apr '13

Hi, the attached example confuses me, I've got: $ g++ -Wall boost_container_bug.cpp -o boost_container_bug && ./boost_container_bug (size=4 [1, 2, 3, 4, 0, 4]) $ LANG=en g++ --version g++ (GCC) 4.6.3 20120306 (Red Hat 4.6.3-2) Copyright (C) 2011 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. Did I miss something or is it a bug? Thanks, Olaf

6 6