- Boost - lists.boost.org

[log] Patch: rotating file collector (Was: Boost.Log: Any Comments?)
by Alexander Arhipenko 23 Apr '13

23 Apr '13

Hi, Just to remind the context, I'm pasting content from previous mail thread. On Tuesday 09 April 2013 18:43:05 Alexander Arhipenko wrote: >> ... >> [snip] >> ... >> >> It would be very nice to have rotation feature work the same way as in e.g. >> python logging library or Apache web server: >> * latest file has the name filename.log >> * previous file has the name filename.log.1 or filename.log.YYYY-MM-DD >> AFAIK, it's impossible to achieve in current implementation. >> >> Andrey, what do you think about this? >The library does not provide this behavior out of the box. But the file sink >uses a file collector object to collect rotated files. This collector can be >implemented by the user (it should derive from sinks::file::collector >interface) and it should be possible to implement it the way you describe. >Would you like to submit such implementation? I've added implementation of rotating file collector - the patch against boost log trunk attached. This is not the final version, I'm going at least to finish most of the TODOs. However, it would be nice to receive feedback and questions from library author as well as from people interested in this feature. Andrey, could you please find several minutes to review the attached patch? Also, I would like to discuss following question with you : rotate_file invocation in ~text_file_backend - I would prefer to have this call during text_file_backend construction phase. Regards

1 0

std::optional<T>
by Fernando Cacciola 23 Apr '13

23 Apr '13

Hi People, I'm pleased to announce that thanks to the outstanding effort of Andrzej Krzemienski who wrote the standard papers, and Ville Voutilainen who champion it during the meetings, we now have std::optional<T> in C++14! Cheers -- Fernando Cacciola www.scisoft-consulting.com

11 15

Plugin/Extension : DLL/so loading?
by Mathias Gaunard 23 Apr '13

23 Apr '13

I remember there were a couple of projects (Boost.Plugin or Boost.Extension) to support dynamic loading of a DLL/so plugin in a portable fashion in Boost. What has happened to these, and is there some code I can use somewhere?

3 4

Boost Graph Library: why are `source()`/`target()` non-member functions?
by Max Moroz 22 Apr '13

22 Apr '13

Why is `source()` not a member function of the graph class? (See http://www.boost.org/doc/libs/1_53_0/libs/graph/doc/index.html for description, and boost/graph/adjacency_list.hpp and boost/graph/adjacency_matrix.hpp for implementation.) I thought the reason was that it was a generic algorithm. But it's not. There's no interface (beyond `source()` itself) exposed by an edge descriptor that would allow `source()` to avoid digging into the implementation. If I interpret the code correctly, `source()` reads directly from the internal data structure of the edge descriptor (`detail::edge_base.m_source`), which I am pretty sure is the implementation detail (`m_source`, despite being strangely exposed as a public member, is not the standard interface; otherwise it would have been defined as such, and `source()` would have been entirely redundant.) I think I'm missing something, but what?

2 3

GSOC
by Flórián Deé 22 Apr '13

22 Apr '13

Personal Details: Name: Flórián Deé University: Budapest University of Technology and Economics Major: Engineering Information Technology Degree Program: B.Sc. Email: deeflorian(a)gmail.com Availability: -starting after the exams, best case: June 10, worst case: June 21 -ending on at the 2nd week at the university: Sept. 15 -planned time spent on GSoC: 32 to 56 hours per week, as needed -other factors: two 3-day long trips - would catch up afterwards with extra work Background information: During my B.Sc. studies I'll already have finished all courses that are related to calculus, linear algebra, number theory, combinatorics and digital technology until summer, so in my opinion, all the educational background needed is present with the exception of the Signals and Systems course in next semester, which would cover digital Fourier transforms in detail. I'd gladly make up for this by reading the needed material. As for my programming background, I've fallen in love with C and C++, which now I consider to be my main languages. 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++ assignments. As a result, I've seen, written and debugged lots of different (admittedly not so enormous) projects with different coding styles. I'm interested in contributing to the Boost Libraries because two of my friends used a quite a lot of them while researching, and their feedback was always positive. The C++11 implementations also sounded nice, so adding to this codebase as a summer project would be an honor. I wanted to do a Solar System simulator during secondary school because I thought that I could calculate the acceleration (and approximate path) of a pen somehow shot near Earth. Then after the first console-based C++ implementation I realized that doubles aren't sufficiently precise to determine if an object is near or inside a planet, so I tried to write a "double double" class - which was too childish to work both code and mathematicalwise. Since then I know that I'd needed a precise mathematical library, and that radix-2 is my best friend when calculating on CPU-s, so I think that this project would be an ideal fit. Rating of knowledge: C++ - 4 C++ Standard Template Library - 4 Boost C++ Libraries - 2 Subversion - 3 Development environments: Eclipse, Visual Studio Documentation tools: Doxygen Project proposal, schedule and milestones: I'd like to work on the Boost.Multiprecision project described on the wiki page. As the description there is more than detailed, I wouldn't like to copy all that text here, but I'd follow the main steps described. I think that the I/O and the +, - operators should be finished, and the multiplication already started by the end of July. The *, / operators should be finished by the end of August at least with the Karatsuba multiplication, interoperable with Boost.Math, and the FFT multiplication should also be started by then. The final goal would be FFT multiplication, with passed correctness tests of wide variety from the beginning of the project.

1 0

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