October 2014 - Boost - lists.preview.boost.org

NT² eand Boost.SIMD release 3.1
by Joel FALCOU 17 Oct '14

17 Oct '14

NT2 is a Matlab inspired, high-level API for scientific computation in C++. Boost.SIMD is a candidate library for inclusion in Boost dealing with the portable encpsulation of SIMD instructions sets. This following releases updates NT2 and Boost.SIMD libraries to version 3.1 This new release of NT2 and Boost.SIMD focused on fixing performance issues and add some new support for architecture and parallel algorithms. The main new features include: - Linear Algebra functions can now use GPGPU through the MAGMA library whenever possible. - Reduction and Scan operations now works with more shared memory back-end - Altivec support is now complete for the VMX ISA We wish to thanks Ian Masliah for his work on MAGMA, Antoine Tran Tan for the shared memory support and Pierre Esterie participation to the Altivec ISA completion. Special thanks to Tim Blenchmann and khegeman for their help and patch submisisons throughout this development. The release Changelog is here : https://github.com/MetaScale/nt2/releases/tag/3.1.0 The files can be download from : http://nt2.metascale.org/downloads/ The Documentation (in progress) is here : http://nt2.metascale.org/doc/html/index.html **Next release announcment** This release is the last C++03 compliant version of both NT2 and Boost.SIMD. The next release - 4.0 - will be C++11 only. The new minimal version of supported compilers will be : - g++ 4.6 - Microsoft Visual Studio 12 (2013) - clang 3.4 - icc 12.1 The main reason for this change is to be able to properly fix the compile-time issue that one may experience. It'll also allow us to use some C++11 feature to simplify or streamline some of the user interface. * The NT2 Team

5 7

Is there something like shared_array<T> but with iterators?
by david.hagood＠gmail.com 15 Oct '14

15 Oct '14

I am doing signal processing, and in order to avoid data copies I need something that: 1) can adopt a buffer of data (e.g. take ownership of a T*) 2) Release that buffer via an appropriate method (e.g. take a deallocator object). 3) Provide operator[] over that buffer 4) provide shared ownership semantics So far, that's shared_array<> to a T [sic] However, I need one more thing: 5) provide iterator support (begin(), end(), cbegin(), cend()) The current shared_array<> doesn't provide that - is there anything else that does?

3 2

[interprocess] BOOST_INTERPROCESS_SHARED_DIR_PATH from env variable
by Jakub Szymanski 11 Oct '14

11 Oct '14

Hi, Will it be possible to have shared_dir not hardcoded? Currently it is determine by some tricks (some involve event log parsing on windows, which fails a lot). The only practical solution for us was to use BOOST_INTERPROCESS_SHARED_DIR_PATH define. boost/interprocess/detail/shared_dir_helpers.hpp inline void get_shared_dir(std::string &shared_dir) { #if defined(BOOST_INTERPROCESS_SHARED_DIR_PATH) shared_dir = BOOST_INTERPROCESS_SHARED_DIR_PATH; #else get_shared_dir_root(shared_dir); #if defined(BOOST_INTERPROCESS_HAS_KERNEL_BOOTTIME) shared_dir += "/"; get_bootstamp(shared_dir, true); #endif #endif } That has limitations tough as it makes it very difficult to test software in isolation (it requires access to the same folder). Similarly it makes is difficult to run program of machines with different file structure. Can it be configurable by some environment variable ? cheers --jakub

2 3

[Modularization] master and develop on super project.
by Kohei Takahashi 10 Oct '14

10 Oct '14

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi all, Is there any policies for merging develop into master and sending pull-request on super project (boost.git)? Before 1.56.0 release, each libraries merged develop into master, though boost.git didn't. It caused to remain some defects in master (also release archive), which are already fixed in develop branch. e.g. - - https://svn.boost.org/trac/boost/ticket/5677 - -- https://svn.boost.org/trac/boost/ticket/6946 (dup of #5677) - - https://svn.boost.org/trac/boost/ticket/10507 and more tickets might exist... The wiki[1] describes about each libraries, but about super project is still absent. [1] https://svn.boost.org/trac/boost/wiki/StartModWorkflow Thanks, Kohei Takahashi -----BEGIN PGP SIGNATURE----- Version: GnuPG v2 iQIcBAEBAgAGBQJUN8IGAAoJEGGyPxPRp/zoW3QP/RBWJ7dcmS8TEr7SpQW97QW4 17chTXKLkqHboHxCJuswXb1rlnmBxxooFItA5kFudEu7U1x5OekPfw+oroQP7v+C CgS5eHanitjprkijvJO2SS61a/YdF/1hMJuHUWzd62hI5ZXWQvNryeJtNl0K0hFb H9EhniJltvBxjOOdvjUsd9XCPby1afi5jOCUjC3J08X6WruRuI5BhCNElVQOSxMG yzaxuLeFDy8L9CICAU8RNEkBEjY1n5rYAVnbETX1WSSoDGEHblRUTOT7JQwSuWWx 0ymdz5cHisM+QoGuLUI0BQS9mayik3vAsre68Ve/xxTHU3LaRRVvaLJIH0YrRRwX fQmyJVGtIbUIhgfmdK8mq050wNOwdzefVfGDkq/XNo+wwlwA6vS4D5gqzfthNL6S HdXI8TIJtdRUYk66vFVcL5uf3PjoW4cZpWYceq1h1EOsIfXuzpr6uO0NDhqEFD+D I3vSmGPV1eAPYpQHP7jeOePMn1EPItU0sVqLk0f6ymafyqnaPU52zhcS45VMS85Q krk+ixO5bND6IU0EgYDn7QZHrTdACDcH5UGscBmxrSUwJ4DrgB6q8mAG/Ek0lZBK /UyXOu3k+ZtrYQIu5aayIyANp6p/EKGNgT7F7/4t0WZzxleskEPzKX9XKoPJnjQ9 aZrV3aXQSRrlEeBhU4fk =KrkS -----END PGP SIGNATURE-----

1 0

[config] Need some clang testing help...
by John Maddock 09 Oct '14

09 Oct '14

Folks there are persistent bug reports relating to clang with libstdc++ not correctly detecting the std lib feature set, see https://svn.boost.org/trac/boost/ticket/7473 for a lengthy discussion on the issue, and why it's been unsolvable. Anyhow, I've just committed a tentative patch in https://github.com/boostorg/config/commit/b36566fe04a89103b7ef5569c03e3a26e… and I'd be grateful if as many folks as possible can test this with their clang/linux/libstdc++/C++11 setups as they can. It works well for me locally, but that's only one possible clang configuration. BTW the easiest way to patch your Boost version is probably to download a replacement libstdcpp2.hpp from here: https://raw.githubusercontent.com/boostorg/config/b36566fe04a89103b7ef5569c… Many thanks in advance, John.

3 8

[range] How does remove_erase() work?
by Stephan T. Lavavej 08 Oct '14

08 Oct '14

Hi, I'm updating my Uniform Container Erasure proposal [1], and I was informed that Boost.Range has a similar-ish remove_erase(). I have a very simple question about its implementation, so I can describe it accurately in my revised proposal, but I don't really want to look at erase.hpp's guts. Does remove_erase() just use the erase-remove idiom, or does it contain special cases for the vector-like family, list-like family, and map-like family? (As N4009 explained, the erase-remove idiom is suboptimal for list, and won't compile for forward_list and the map-like family.) Thanks, STL [1] http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2014/n4009.htm

3 4

It's that time again already - Boost 1.57.0 schedule
by Marshall Clow 08 Oct '14

08 Oct '14

It’s that time again. Here’s the schedule for the upcoming 1.57 release. If you’ve got changes that you’ve been putting off making, now’s the time to get busy. * 13-Oct Last day for significant changes * 20-Oct Beta build * 27-Oct Automatic merging turned off, build an RC * 3-Nov Release Boost 1.57.0 Questions? Ask them here. — The Boost release managers. P.S. We’ll be updating the calendar on boost.org/development soon.

5 5

Update on status of C++ 11 only Boost fork/modular distro
by Niall Douglas 08 Oct '14

08 Oct '14

Dear all, You may remember me saying these past six months that I was working on the tools necessary to fork Boost into a C++ 11 only distro which also lets you use Boost libraries individually standalone without the rest of Boost. I have got a first alpha version of my clang AST based local bindings generator working, and you can see it at https://github.com/ned14/local-bind-cpp-library. This tool lets you generate a set of local namespace bindings for some implementation of some library such that your library implementation code need not care what the underlying implementation of dependent library facilities is. This way your Boost library could be using either Boost or the C++ 11 STL and you don't need to care (most of the time). Now I have the first alpha version working and generating a reasonable set of C++ 11 STL bindings (https://github.com/ned14/local-bind-cpp-library/tree/master/include/s tl11), I'll start with porting over my own Boost libraries to it, so Spinlock, Expected, non-allocating future-promise and AFIO will be the first, and with those AFIO should become capable of being completely standalone from Boost whilst still also able to compile against Boost. After that I'll try porting the entire of Boost.Thread over such that Boost.Thread no longer needs Boost. I think that would be hugely useful for people wanting to use potential C++ 17 threading functionality without needing to draw in all of Boost as a dependency, and that might greatly widen the empirical experience new threading facilities might get before the 17 standard. Once I've succeeded with that, if the present situation with Boost has not improved by then, I'll press ahead with a new Boost distro website which score ranks Boost modules according to the scoring manifest I outlined on this list some months ago, and lets you download standalone Boost distros of your choice. Obviously only libraries ported to be standalone capable would be available, so I would hope that if I can demonstrate an easy port with Boost.Thread, others may follow with their libraries. All this work occurs in my free time of a few hours per week, so progress will be slow but steady. Comments on the local namespace binding tool are welcome. Comments on the proposed C++11/14 good practice idiom of binding dependent library APIs into the local namespace are also welcome, because as far as I can tell no one here has discussed a formal set of conventions to make best use of inline namespaces for new libraries. I'll put it another way - local namespace binds are a poor man's version of Herb's proposed C++ ABI facility. As much as they aren't quite as good, they are probably good enough (relative to Herb's most recent proposal). Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

3 7

Fusion and constexpr, C++11 support entirely broken on develop
by Mathias Gaunard 07 Oct '14

07 Oct '14

Hi folks, I recently decided to move all of my code to C++11, and I usually use the develop version of each library so that I can have the latest developments and can easily contribute. However, Fusion appears to be fundamentally broken on develop whenever C++11 is enabled with GCC. It works fine with C++03. I don't think this is a situation that should ever happen; while develop is for libraries undergoing development, it shouldn't be totally broken either. In particular, the problem is that it uses BOOST_CONSTEXPR in contexts where C++11-constexpr is invalid, but where it is valid in C++14. Should there be a separate BOOST_CONSTEXPR14 macro?

3 3

[EXTERNAL] Tickets notifications
by Adam Wulkiewicz 07 Oct '14

07 Oct '14

Hi, AFAIU the notifications about new tickets and updates are sent only to the owner of the library. This is inconvenient since many libraries are developed by a group of people, e.g. Geometry. I'm aware that anyone can get updates if he's on the Cc list but people which are curious about development aren't added to this list automatically. Ideally it would be great if any user could subscribe to get notifications for any library like it's possible on GitHub. Or maybe moving to GitHub mentioned by John in one of the other threads would be the best solution? Anyway, is it somehow possible to receive notifications as a non-owner automatically? If the answer was no, would sending notifications to the library mailing list be a reasonable solution? AFAIU this could be done by adding a "fake" owner with a mailing list email. Regards, Adam

2 1