March 2015 - Boost - lists.boost.org

Library for await and stackless coroutine emulation
by TONGARI J 10 Mar '15

10 Mar '15

Hi there, I just want to announce my recent work on a small library called `CO2`, which intend to emulate await/yield and stackless coroutine. https://github.com/jamboree/co2 This library depends on Boost.Preprocessor, while I'm not proposing this library to Boost at this moment, if there's much interest in this library, I'd think about it. The idea is based on N4286: http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2014/n4286.pdf The library is still work in progress, but the core features are complete, so if you're interested in the resumable function proposal and don't have a tool to experiment with, this library may serve the purpose. You can find a more detailed description on the github page, if you want to know some number about performance, from a rough test borrowed from Boost.Coroutine (symmetric/performance_switch.cpp), the avg cpu cycles for context-switch are: Boost.Coroutine: 40 cycles CO2: 16 cycles built-in await: 9 cycles (compiled with MSVC14, x64) Feedbacks will be appreciated. Thanks

2 2

Release 1.58. - Permission to merge Boost.Integer
by Vicente J. Botet Escriba 10 Mar '15

10 Mar '15

Hi, I forget to do the merge to master of the last 5 commits in https://github.com/boostorg/integer/commits/develop Please, let me know if I can do the merge? Best, Vicente

5 5

[interprocess] BOOST_NO_SWPRINTF not honored...
by Joseph Van Riper 09 Mar '15

09 Mar '15

This might seem pedantic, but with Microsoft's compilers magnify this issue a bit. boost/interprocess/detail/win32_api.hpp, line 2255 (as of boost 1.57) uses std::sprintf even when BOOST_NO_SWPRINTF is defined. This causes a problem for some folks (like me), as VC++ 2013 (and possibly other Microsoft compilers) have set a pragma deprecating this function, which leads to outright compiler errors when compiled with #pragma strict_gs_check(on), as is required in some environments who are picky about security. I confess, I have not checked to see if interprocess still does this in the latest code, but if it doesn't, I wonder if it could use an alternative (Microsoft recommends sprintf_s, but you might know of something better). - Trey

1 0

[1.58.0] Release Notes
by Daniel James 09 Mar '15

09 Mar '15

Hi, The beta release is coming up soon, so please can you add your release notes as necessary. Easiest way is probably to create a pull request at: https://github.com/boostorg/website/blob/master/feed/history/boost_1_58_0.q… thanks, Daniel

3 3

[gsoc15]Request for feedback: Enhanced vector and deque containers proposal
by Sandilya Jandhyala 08 Mar '15

08 Mar '15

Hello Boost, My name is Sandilya Jandhyala, and I'm interested in the 'Enhanced vector and deque containers' project idea. I've prepared a first draft of my proposal. It can be viewed here: https://github.com/jysandy/boost-gsoc-proposal/blob/master/README.md I haven't included my solution to the programming competency test, since I don't want it to be copied. I'd appreciate feedback/suggestions/criticisms, since I feel that the proposal still has some room for improvement. I'd also appreciate suggestions for additional features. Thanks, Sandilya

1 0

GSoC project " Enhanced vector and deque containers "
by ahmad samir 08 Mar '15

08 Mar '15

Good morning guys I'm Ahmad, I want to contribute in " Enhanced vector and deque containers ", I'm working on the programming competency test, but I have a question about the main idea * The idea is " Write full-fledged and enhanced C++11 versions of std::deque and std::vector with extra functionality that gives the programmer more control over buffers, segments and iteration " . * Is there specific features or new functionality you want to add or I should search for some new features in other languages like " java & python " which isn't in C++ vector and deque containers, or search for what can give the programmer more control on these containers ?.

1 0

GSoC 2015 Paritipation
by vladiksonic＠mail.ru 08 Mar '15

08 Mar '15

Dear Boost, I am eager to participate in GSoC 2015 in one of the two projects (from this page: https://svn.boost.org/trac/boost/wiki/SoC2015 , 1 or 5). As far as I understood - the first one is tougher, and (as for me) conditions are clearer, and the second one is (seemingly) simpler. I am totally new to Boost (never used it, actually, but that is because I mostly used C++ in programming contests). I am familiar with algorithms (from contests), and testing (manual, both for stability and performance), but I am not very familiar with C++11 (but I can compensate for that, because I have knowledge of other languages (C#, Javascript (nodejs)) and older C++ versions), and automated tests. At first, I'd like to ask - what should I choose? My main goals are: 1) To enter the community 2) Learn C++11 new features (I will start early on that, obviously) 3) And to learn how to code open-source programs (up till now I only have closed-source ones, apart from one project, which has not-so-good code quality - see here: https://github.com/hotab/OpticalBuilder ) 4) Plus understand better how to maintain library code (I already have some understanding of that, but I want to sharpen it). Both can give me that. I'm looking forward to hearing from you, Vladislav Babkin

1 0

[dynamic_bitset] Maintainer available?
by Riccardo Marcangelo 07 Mar '15

07 Mar '15

Hi Boost developers, Does anybody know if the maintainer of dynamic_bitset is available? I made some contributions to dynamic_bitset on GitHub last year but there doesn't appear to have been any activity in the repository for many months. My pop_back() commit was accepted into the 'develop' branch over 5 months ago and, despite passing tests, has yet to be merged to 'master'. I also implemented capacity(), reserve(), and shrink_to_fit() which have been sitting in a pull request waiting to go into 'develop' since November. It would be nice to get these into 1.58.0 if possible. Kind regards, Riccardo -- View this message in context: http://boost.2283326.n4.nabble.com/dynamic-bitset-Maintainer-available-tp46… Sent from the Boost - Dev mailing list archive at Nabble.com.

2 1

GSoC project " Enhanced vector and deque containers "
by ahmad samir 07 Mar '15

07 Mar '15

Good morning guys I'm Ahmad, I want to contribute in " Enhanced vector and deque containers ", I'm working on the programming competency test, but I have a question about the main idea * The idea is " Write full-fledged and enhanced C++11 versions of std::deque and std::vector with extra functionality that gives the programmer more control over buffers, segments and iteration " . * Is there specific features or new functionality you want to add or I should search for some new features in other languages like " java & python " which isn't in C++ vector and deque containers, or search for what can give the programmer more control on these containers ?.

1 0

[GSOC2015] uBLAS Matrix Solver Project
by Rajaditya Mukherjee 07 Mar '15

07 Mar '15

Hi, My name is Raj and I am Phd student in Computer Graphics. I am interested in tackling the problem of uBLAS Matrix Solver and in order to write my proposal, I am looking for inputs for which of the following algorithms will be most useful for prospective users in boost-numeric library. Here is a categorical list of all the prospective ones which will bring uBLAS updated to other commercial libraries like Eigen/Armadillo. Please let me know your preferences.... *David Bellot* : As a potential mentor, do you have any specific additions or deletions for this list? This could also be useful for other candidates pursuing this project. *DENSE SOLVERS AND DECOMPOSITION* : 1) *QR Decomposition* - *(Must have)* For orthogonalization of column spaces and solutions to linear systems. (Bonus : Also rank revealing..) 2) *Cholesky Decomposition* - *(Must have)* For symmetric Positive Definite systems often encountered in PDE for FEM Systems... 3) *Householder Method* - Conversion to tridiagonal form for eigen solvers. *SPARSE SOLVERS AND PRECONDITIONERS* : 1) *Conjugate Gradient* - *(Must have)* For symmetric Positive Definite systems, this is the kryvlov space method of choice. Both general and preconditioned variants need to be implemented for convergence issues. 2) *BiCGSTAB* *(Needs introspection)* - For non symmetric systems.. 3) *Incomplete Cholesky Decomposition* *(Good to have)* - For symmetric Positive definite sparse matrices, to be used as preconditioner as extension to (1) for preconditioned CG Methods ... 4) *Jacobi Preconditioner* *(Must have)* - As prerequisite for step(1). *EIGEN DECOMPOSITION MODULES (ONLY FOR DENSE MODULES)**:* 1) *Symmetric Eigen Values* - *(Must have)* Like SSYEV Module in Lapack - That is first reduction to a tridiagonal form using Householder then using QR Algorithm for Eigen Value computation. 2) *NonSymmetric Eigen Values* - *(Good to have)* Like SGEEV module in Lapack - using Schur decompositions as an intermediate step in the above algorithm. 3) *Generalized Eigen Values* - *(needs introspection)* I use this in my research a lot and its a good thing to have.. ** Computing Eigen Decomposition of sparse modules needs special robust numerical treatment using implicitly restarted arnoldi iterations and may be treated as optional extensions.

3 3