May 2013 - Boost - lists.preview.boost.org

[log] Is it lock free ?
by adrien courdavault 09 May '13

09 May '13

Hi I'm really interested by the new Log library. I would like to know if there is a part of this API that could be wait free, not the << operator perhaps, but a specific function to log a message without having to worry about a malloc or lock called inside the library. I would like to use this from a time critical thread. thank you

3 5

[predef] Status and review results?
by Rene Rivera 09 May '13

09 May '13

Just a quick message to mention that I've finished all the review related changes to the Predef library. And done some of the future tasks on the library. In particular the addition of endianness definitions and the moving of MinGW as a platform instead of a compiler (and hence the addition of BOOST_PLAT category of macros. And the one question I have is.. When will the review results of the library be decided on? I ask for two reasons. One, the obvious, is just to know if it's accepted or not. And to know if it's not accepted and hence that I should spend time to recreate a "boost/detail/endian.hpp" header that implements the Predef logic for endian detection. Note the project move to Github earlier this year (in preparation for the Boost git move). You can find the project at < https://github.com/grafikrobot/boost-predef>, and browse the current documentation at <http://tinyurl.com/cqqhhev>. -- -- -- Grafik - Don't Assume Anything -- Redshift Software, Inc. - http://redshift-software.com -- rrivera/acm.org - grafik/redshift-software.com -- 102708583/icq - grafikrobot/aim - grafikrobot/yahoo

8 31

[chrono/date] Two axes for a date design: validity check and representation and an essence date class/factory
by Vicente J. Botet Escriba 09 May '13

09 May '13

Hi, I would like to discuss about some alternatives to take in account the following use case "Reset a date if it resulting date is valid and report if this succeeds" when we have two axes of variation on the date design. Let me add two parameters to the possible date class. * validity: unchecked/checked * representation: serial, ymd(calendar), iso_ordinal, iso_week... date<serial, unchecked> is an unchecked serial date. It corresponds +/- to the N3344 date suggestion. date<ymd, checked> is a checked ymd date. It corresponds +/- to the H.H 2011 date proposal. WARNING !!! I'm not saying that a template class is a good thing, it is just to state clearly the features of the date. N3344 suggest a date::set_if_valid_date function. date<serial, unchecked> ud=...; if (ud.set_if_valid_date(y,m,d)) { // xx } else { // yy } But if we want to do as for the date constructors we will need multiple overloads again for this function :(. As show on this ML serial+unchecked dates validation could have some surprises. We can get the same service building a ymd date that can provide a way to be validated. //------------- // Validity check done when is_valid() is called on ymd_date. date<serial, unchecked> dd=...; date<ymd, unchecked> yd(y,m,d); if (yd.is_valid()) { // CHECK dd = yd; // xx } else { // yy } Instead of using explicitly date<ymd,unchecked> we can use a make_essence() factory that returns some kind of tuple essence_date<year,month,day> that is not a date, but its essence. It contains everything needed to build a date. This essence should be implicitly convertible to any date. The make_essence_date() factory supports all the different orders of parameters. The operator/() expression would also create an essence date. // -------------- //No validity check is done. date<serial, unchecked> ud = make_essence(y,m,d); date<serial, unchecked> ud1 = m/d/y; This essence dates support date validity, that is provides a is_valid() function. // -------------- // Validity check done when is_valid() is called on essence. date<serial, unchecked> ud; auto d = y/m/d; if (d.is_valid()) { // CHECK ud = d; // xx } else { // yy } //-------------- // Validity check done at construction of cd. serial_date<checked> cd = make_essence(m,d,y); //serial_date<checked> cd = m/d/y; Summary: IMO Having checked and unchecked dates with several representations could help to make dates interface more precise and concise. Checked dates are always valid. Dates with a specific representations would provide only, as suggested by Howard, only the functions that are efficient. The use of essence dates factories simplify the interface of all the date classes construction. Only 1 constructor by representation needed independent of the parameters order date(essence<ymd> &&); date(essence<serial> &&); date(essence<iso_ordinal> &&); date(essence<iso_week> &&); Comments welcome, Vicente P.S.The essence date classes and the contextual dates are quite related. I prefer to let this 3rd contextual axe of variation out of the current scope. I will try to integrate it soon.

2 3

[trac] Add component for Boost.Log
by Andrey Semashev 09 May '13

09 May '13

Hi, Not sure whom to ask, but could someone create a component "log" for Boost.Log in Trac? I (andysem) am the maintainer. Thanks.

2 2

JSON parsing library project for GSoC
by Olexandr Popilovsky 09 May '13

09 May '13

Hi. My name is Olexandr Popilovsky and I'm a student of Taras Shevchenko National University of Kyiv. I made a proposal for JSON parsing library and I would like to know that can I be accepted for this project and if yes, what I should to do to increase my chances (demo or smt.). thanks. waiting for your answer =)

4 6

Company copyright release examples
by Brendon Costa 08 May '13

08 May '13

I am looking into asking the company I work at for an official written "release" on the copyright for work I with to contribute to boost as per the description at: http://www.boost.org/development/requirements.html I will be developing this in my own time outside of work hours but I will need this release due to clauses in my contract and the relationship that the proposed library would have to my normal work. Are there any examples of what this release should look like for boost to accept it? I think having a sample release available on the page I linked to above would be really useful. Thanks, Brendon.

4 6

[iterator] Silent-but-deadly bug in iterator_facade category
by Gabriel Redner 08 May '13

08 May '13

Hi folks, I've stumbled across a nasty problem with iterator_facade in which code which appears correct and compiles fine can cause runtime problems. Consider an iterator_facade which just wraps a random access iterator and applies a transformation to it: char transform(int i) { return 'a'; } struct my_iterator : public boost::iterator_facade<my_iterator, char, // value_type std::random_access_iterator_tag, // *** char> // reference { my_iterator(const std::vector<int>::iterator& it) : _it(it) {} char dereference() const { return transform(*_it); } bool equal(const my_iterator& other) const { return _it == other._it; } void increment() { ++_it; } void decrement() { --_it; } void advance(int n) { std::advance(_it, n); } int distance_to(const my_iterator& other) const { return other._it - _it; } private: std::vector<int>::iterator _it; }; Now if I have an STL-style function which is overloaded on the iterator category, clearly the random-access version will be selected. template <typename TIterator> void funcImpl(TIterator it, std::input_iterator_tag) { std::cerr << "Running func treating 'it' as input iterator\n"; } template <typename TIterator> void funcImpl(TIterator it, std::random_access_iterator_tag) { std::cerr << "Running func treating 'it' as random access iterator\n"; } template <typename TIterator> void func(TIterator it) { funcImpl(it, typename std::iterator_traits<TIterator>::iterator_category()); } In the line marked ***, the docs suggest I should use boost::random_access_traversal_tag instead of std::random_access_iterator_tag - but when I do, the program no longer selects the random access version of 'func', choosing the forward-only version instead! The iterator_category of my_iterator ends up as: boost::detail::iterator_category_with_traversal<std::input_iterator_tag, boost::random_access_traversal_tag> The reason I called this a 'deadly' bug is that std::advance is suddenly a deathtrap: std::advance(it, -1); This is fine and legal for random access iterators, but if the forward-only version gets called instead, the behavior is undefined (on my system, it loops endlessly). The compiler doesn't complain either way - the problem surfaces only at runtime. Does the problem lie with iterator_category_with_traversal? It looks like it's intended in this case to be convertible to either std::input_iterator_tag or std::random_access_iterator_tag, but clearly something is wrong with the latter conversion. Any ideas? Thanks, -Gabe

7 14

[gsoc 2013] draft proposal for chrono::date
by Anurag Kalia 07 May '13

07 May '13

Hi, I mistakenly sent this draft as a reply to personal email of Vicente. I am sorry! Here it is again, verbatim. One related question though. I still haven't got the hang of a Mailing List. Is there a software takes the pain away from it. For example, I use outlook.com and '>' get appended in front of only the start of paragraph making it wholly unreadable. Thanks, Anurag Kalia. ---------------------------------------- > Date: Sat, 27 Apr 2013 18:14:09 +0200 > From: vicente.botet(a)wanadoo.fr > CC: anurag.kalia(a)outlook.com > Subject: Re: [boost] [gsoc 2013] draft proposal for chrono::date > > Le 27/04/13 16:58, Anurag Kalia a écrit : > >> I understand that you want to design your own library, I did the same, > >> but getting a coherent design would take you some time. > >> I suggest you to start with the H. Hinnant proposal as a reference and > >> state what you want to added, changed or removes. State clearly why do > >> you want these changes, what would be improved/solved. > >> > >> Don't forget to inspect the threads/discussion related to the H.Hinnant > >> proposal, and the paper N3344 Toward a Standard C++ 'Date' Class > >> http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2012/n3344.pdf > > Hi Vicente, > > > > Thanks a lot for the exhaustive reply! I will get back to you after going through all the things you pointed out and also the paper you forwarded. But I have one quick question for now, from where can I start my search for the threads and discussions around this date proposal? > > > > > > > Hi, > > There were some interesting exchanges in > https://groups.google.com/forum/?fromgroups=#!forum/kona-date-lib > > I guess that if you request to the administrator s/he will be able to > give you read access to the existing threads (there are not too much). > > Best, > Vicente > I have finally implemented some representative functions in date class and calendar to get a better idea of the proposal. I have dropped the JDN storage for another implementation that I thought of in the middle of the night. I think it is a hybrid approach between serial and field-based implementations. Off the top of my head, without any alternative implementations at hand and benchmarks too, I can't think of something inherently wrong in it. I would really like to hear its gotchas, if any. I have overhauled my proposal on the basis of it. It doesn't contain other headings like about me etc for today. I have tried to incorporate all the previous points you noted before and have also tried to keep naming consistent. I did use uppercase constants in there but they are meant to be private. Is that okay? I knew little about the benchmarks before you sent me that paper. From what I got, they just iterate over a significant period of time and note the start and ending time. Is it more than that? And should I make them before or after I have multiple implementations in hand. I could get hold of your and Howard Hinnant's implementations for them, but then the APIs would have to line up, wouldn't they? Hope you find my proposal better this time around. With an open ear, Anurag Kalia.

9 95

[GitHelp] bootstrap.bat calls missing build.bat
by legalize+jeeves＠mail.xmission.com 07 May '13

07 May '13

[Please do not mail me a copy of your followup] OK, so I am experimenting with modular boost, following the instructions here: <https://svn.boost.org/trac/boost/wiki/TryModBoost> I cloned and did the submodule update. I didn't have CMake; is this why I experience a problem? What exactly does 'cmake -P forward_header.cmake' do? When I run boostrap.bat it fails early with an attempt to call build.bat which doesn't exist. Thanks for your time. -- "The Direct3D Graphics Pipeline" free book <http://tinyurl.com/d3d-pipeline> The Computer Graphics Museum <http://computergraphicsmuseum.org> The Terminals Wiki <http://terminals.classiccmp.org> Legalize Adulthood! (my blog) <http://legalizeadulthood.wordpress.com>

3 5

[quickbook][random] Build under doc/ broken?
by John Maddock 07 May '13

07 May '13

I'm not sure if this is a random or quickbook issue, but building the doc build under doc/ fails with current quickbook trunk. Any ideas? Thanks, John.

2 1