- Boost - lists.preview.boost.org

[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

[variant] Multivisitation patch
by Antony Polukhin 18 Apr '13

18 Apr '13

Hi, There is a ticket #8459 that adds boost::apply_visitor for arbitrary amount of arguments. If there will be no objections and comments, I'll commit it to trunk at the end of next week. -- Best regards, Antony Polukhin

2 1

Using both boost 64 and 32 bit on Windows with CMake
by Klaim - Joël Lamotte 18 Apr '13

18 Apr '13

It appear that this mailing list don't accept cross messages so I'm posting a link to a discussion related to both boost and cmake. Basically I'm looking for guidance to help improving development of one application using boost in 64 and 32 bit modes by finding a way to make the FindBoost cmake module find the right binaries depending on either mode. Currently you have to specify an undocumented CMake variable manually to do this. See the full discussion: http://cmake.3232098.n2.nabble.com/boost-b2-Differenciate-32bit-and-64-bit-… Joel Lamotte

5 15

Interested in working on a project for Boost
by Ganesh Prasad 18 Apr '13

18 Apr '13

I am a Student of 2nd Semester, B.Tech in Computer Science and Engineering from National Institute of Technology, Rourkela, India. I am interested in working on a project on “ Developing Data Structures and related algorithms for a generalized Matrix and Vector Class” for the Boost Libraries as a participant of GSoC. I have some experience in working on algorithm development in C++. And I am positive that I have pretty good knowledge in advanced C/C++.I have participated in various coding contests too. Though I am just a student of Second Semester, I can assure you that I'll be working up to the requirements and expectations. But I am new to this GSoC program. I need advice and help to go on with this project. So, requesting advice and help before I start discussing about my project ideas. :) -Ganesh Prasad

2 1

[range] location of numeric algorithms
by Nathan Ridge 18 Apr '13

18 Apr '13

Hi, Is there any reason the numeric algorithms in Boost.Range (accumulate(), inner_product(), etc.) are located in boost/range/numeric.hpp rather than in files named boost/range/algorithm/<algorithm-name>.hpp like every other algorithm? I think it's confusing for users who expect the uniform file naming convention that Boost.Range applies to other algorithms, to apply to these ones too. Thanks, Nate

2 2

[iostreams] Doc errors
by Richard Hadsell 18 Apr '13

18 Apr '13

I'm just starting to read the Boost.Iostreams documentation and found a couple of minor errors. Perhaps someone would like to make these trivial corrections to help other novices in the future. (1) 2.1.4. In the implementation of container_device::write, the block that writes remaining characters at the end of the container ("if (result < n) { ... }") should insert characters from 's + result', not from 's': "container_.insert(container_.end(), s + result, s + n);". (2) 2.1.4. In the implementation of container_device::seek, there are 3 uses of "ios_bas" instead of "ios_base". -- 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