- Boost - lists.preview.boost.org

[git] regression testing
by Joaquin M Lopez Munoz 05 Dec '13

05 Dec '13

Hi, Congrats on the recent Git switchover. Two questions: * Are regression test scripts yet adapted to the new repository? * For SVN, test columns mark the latest commit the tests are done against. How will we do with Git? Looks like commits are identified by 40-digit numbers, which are considerably harder to visually track than nice consecutive SVN changeset numbers. Joaquín M López Muñoz Telefónica Digital

3 3

Git hardlinks, and developing
by John Maddock 05 Dec '13

05 Dec '13

OK so I've just been hit by my first Git SNAFU. 1) I checked out modular Boost as per https://svn.boost.org/trac/boost/wiki/TryModBoost 2) Changed libs/multiprecision to the develop branch also as per https://svn.boost.org/trac/boost/wiki/TryModBoost 3) Added some pending changes from the old SVN Trunk that I didn't commit before the changeover. 4) Ran the tests and everything failed, even though it was passing on SVN :-( The issue was that the headers under boost/ were now *copies* of the last release and no longer pointers to the new "develop" code. Interestingly some of the hard links were updated when I ran bjam, but apparently not all :-( Running bjam -a fixes the issue, but it's a very easy trap for the unwary.... John. PS, one other very annoying issue for VC++ users: 1) Open a test file. 2) Select a #included boost file, right click and select "Open document". 3) Document opens, but gives you a copy under boost/ not the "real" header under libs/mylib/. 4) Run away screaming after you've spent an XXX hours editing the wrong file....

8 13

Possible issue with Boost.Atomic (and consequently Boost.Thread)
by Peter Dimov 04 Dec '13

04 Dec '13

In https://svn.boost.org/trac/boost/ticket/9466 it is reported that using Boost.Thread on AIX without defining BOOST_THREAD_DONT_USE_ATOMIC causes an assertion failure in boost/smart_ptr/detail/lwm_pthreads.hpp (pthread_mutex_lock failing with EINVAL.) As I explain in the ticket, what I think happens is that Boost.Atomic falls back to its generic lock pool implementation, which uses a static array of 41 boost::lightweight_mutex objects. However, since lightweight_mutex has a constructor calling pthread_mutex_init and a destructor calling pthread_mutex_destroy, using call_once, directly or indirectly, outside main can lead to accessing a mutex object before it's constructed or after it's destroyed, depending on static initialization order. If I'm correct in this assumption, then lockpool.hpp/.cpp need to be fixed to either use boost/smart_ptr/detail/spinlock_pool.hpp*, or use statically initialized pthread mutexes directly. *) Which is basically what it was designed for, see the comment: // spinlock_pool<0> is reserved for atomic<>, when/if it arrives

2 4

[util] Any interest in address_of_first_nonarray_elem ?
by Mostafa 04 Dec '13

04 Dec '13

Working with built-in array objects I find myself repeatedly needing to get the address of the first non-array element. This is so I can do generic array comparisons and copies without resorting to multiple loops. For example: template <typename T, std::size_t N> bool generic_array_equal(T (&lhs)[N], T const (&rhs)[N]) { return std::equal( address_of_first_nonarray_elem(lhs), address_of_first_nonarray_elem(lhs) + total_extent<T[N]>::size, address_of_first_nonarray_elem(rhs)); } Any interest in adding such a function to the util library? Mostafa

5 8

Gitflow (was Git: maintaining super-project)
by Mateusz Loskot 04 Dec '13

04 Dec '13

On 4 December 2013 12:03, Jeremy Ong <jeremycong(a)gmail.com> wrote: > Maybe I should have said "github" world. Many organizations and open source > projects choose to make it the lowest branch in the hierarchy specifically > because it *is* the default. I wasn't aware that boost uses nvie's gitflow. > There are many that feel it's a bit overengineered, but if it's specified > somewhere that that is what we're using then eschewing the defaults is fine. Use of GitHub is orthogonal to development workflow. Somewhat 'traditionally', GitHub is perceived as an incubator of fast yet agile Ruby/... web shops which don't rely on a typical release model, but they constantly roll releases. Boost is not a rolling release based project! I will conclude with Scott Chacon's blog (http://scottchacon.com/2011/08/31/github-flow.html) """ For teams that have to do formal releases on a longer term interval (a few weeks to a few months between releases), and be able to do hot-fixes and maintenance branches and other things that arise from shipping so infrequently, git-flow makes sense and I would highly advocate it’s use. For teams that have set up a culture of shipping, who push to production every day, who are constantly testing and deploying, I would advocate picking something simpler like GitHub Flow. """ Best regards, -- Mateusz Łoskot, http://mateusz.loskot.net

1 0

BOOST_NO_CXX11_ATOMIC?
by Peter Dimov 04 Dec '13

04 Dec '13

Now is perhaps not a very appropriate time to ask, but why don't we seem to have a macro for the <atomic> C++11 header?

6 13

[container] Doc regressions in 1.54 and 1.55
by Jared Grubb 04 Dec '13

04 Dec '13

I went to look up the API on "flat_set" and found that the header documentation after 1.53 is not helpful. I think this is unintentional, so I thought I'd point it out: http://www.boost.org/doc/libs/1_53_0/doc/html/boost_container_header_refere… 1.53 docs: "class flat_set" is shown in the <flat_set.hpp> section; the link takes you to the API reference -- perfect! http://www.boost.org/doc/libs/1_54_0/doc/html/boost_container_header_refere… 1.54 docs: "class flat_set" only appears in <container_fwd.hpp> section; the link takes you to the forward decl only and not a reference header. However, other references to "flat_set" (for example, in the free functions like operator==) are hyperlinked and do take you to the right header reference file.. http://www.boost.org/doc/libs/1_55_0/doc/html/boost_container_header_refere… 1.55 docs: "class flat_set" is not listed at all. The API header reference does not exist and none of the references to flat_set are hyperlinked. Jared

2 1

[GitHub] README policy?
by Mateusz Loskot 04 Dec '13

04 Dec '13

Hi, A little trivia, apparently GitHub wishes every repository has a readme: "We recommend adding a README to this repository to help give people an overview of your project." Is there any policy about if such README is required and what it should contain? Best regards, -- Mateusz Łoskot, http://mateusz.loskot.net

1 0

[modular boost] non-linked headers
by Edward Diener 04 Dec '13

04 Dec '13

On Windows after the .\b2 headers step, there are these directories under the modular-boost/boost diectory which are not symbolic links: detail graph numeric pending I can understand 'detail' but should not the others have been created as symbolic links to their appropriate libraries header files ?

6 36

[IdentityType] Const omission undocumented.
by Mostafa 03 Dec '13

03 Dec '13

#include <boost/utility/identity_type.hpp> #include <utility> int main() { BOOST_IDENTITY_TYPE(( std::pair<int,int> const )) x; // This compiles, so the const qualification in the // above macro parameter is ignored. x = std::make_pair(0,1); return 0; } The omission of const qualification in BOOST_IDENTITY_TYPE types is undocumented. This omission maybe deduced from the documented implementation, but, as noted, that is not part of the library api and is subject to change. I therefore propose that this behaviour be documented as part of the library's public api. Mostafa

4 5

[swap] std::swap should be pulled in from <utility> in C++11
by Joaquin M Lopez Munoz 03 Dec '13

03 Dec '13

Hi, boost/utility/swap.hpp #includes <algorithm> to bring std::swap in, which is OK for C++03 but not so in C++11: std::swap has been moved to <utility> as explained in [diff.cpp03.library]. Should I create a ticket? Best regards, Joaquín M López Muñoz Telefónica Digital

2 1

[git] Write (i.e. push) permission
by Beman Dawes 03 Dec '13

03 Dec '13

Boost library maintainers, please remember you need to apply for write permission for your library. See https://svn.boost.org/trac/boost/wiki/ModAdmin As of a couple of minutes ago, all pending requests for write permission have been completed. --Beman

1 0

[git] unable to push
by Oliver Kowalke 03 Dec '13

03 Dec '13

Hi, I tried to push my commits to the remote branch 'develop' but I get an error: ERROR: Permission to boostorg/context.git denied to olk. fatal: The remote end hung up unexpectedly Is the boostorg repo closed for commits? so long, Oliver

2 1

[git] Effect of "git checkout develop" on submodule?
by Beman Dawes 03 Dec '13

03 Dec '13

If I do this: git clone --recursive git@github.com:boostorg/boost.git modular-boost cd modular-boost/libs/math git checkout develop What commit of math develop branch should be checked out? The current head? The head at the time of the git clone? Some other head? --Beman

3 8

[ANN] C++Now 2014: 5 Days to Submissions Deadline
by Boris Kolpackov 03 Dec '13

03 Dec '13

Hi, Only 5 days left before the submissions deadline for C++Now 2014! C++Now is a general C++ conference for C++ experts and enthusiasts. It is not specific to any library/framework or compiler vendor and has three tracks with presentations ranging from hands-on, practical tutorials to advanced C++ design and development techniques. For more information about C++Now, see the conference's website: http://cppnow.org/about/ Have you learned something interesting about C++ (e.g., a new technique possible in C++11)? Or maybe you have implemented something cool related to C++ (e.g., a C++ library)? If so, consider sharing it with other C++ enthusiasts by giving a talk at C++Now 2014. For more information on possible topics, formats, etc., see the call for submissions: http://cppnow.org/2013/10/21/2014-call-for-submissions/ Boris

1 0

modular-boost comile error
by Vicente J. Botet Escriba 03 Dec '13

03 Dec '13

Hi Im getting this error when compiling several tool sets on the same command line. Best, Vicente ./b2 toolset=clang-3.2xl,darwin-4.8.1x -l 300 -j6 modularized layout build Building the Boost C++ Libraries. Performing configuration checks - symlinks supported : yes - hardlinks supported : yes - 32-bit : no - 64-bit : yes - arm : no - mips1 : no - power : no - sparc : no - x86 : yes - has_icu builds : no warning: Graph library does not contain MPI-based parallel components. note: to enable them, add "using mpi ;" to your user-config.jam - zlib : yes - iconv (libc) : no - iconv (separate) : yes - icu : no - icu (lib64) : no - compiler-supports-ssse3 : yes - lockfree boost::atomic_flag : yes - compiler-supports-avx2 : yes - gcc visibility : yes - long double support : yes warning: skipping optional Message Passing Interface (MPI) library. note: to enable MPI support, add "using mpi ;" to user-config.jam. note: to suppress this message, pass "--without-mpi" to bjam. note: otherwise, you can safely ignore this message. - zlib : yes - 32-bit : no - 64-bit : yes - arm : no - mips1 : no - power : no - sparc : no - x86 : yes - has_icu builds : no - zlib : yes - iconv (libc) : no - iconv (separate) : yes - icu : no - icu (lib64) : no - compiler-supports-ssse3 : no - compiler-supports-avx2 : no - lockfree boost::atomic_flag : yes - gcc visibility : yes - long double support : yes - zlib : yes /Users/viboes/github/modular-boost/tools/build/v2/build/virtual-target.jam:1099: in virtual-target.register-actual-name from module virtual-target error: Duplicate name of actual target: <pstage/lib>libboost_atomic.dylib error: previous virtual target { clang-darwin%clang-darwin.link.dll-libboost_atomic.dylib.SHARED_LIB { clang-darwin%clang-darwin.compile.c++-lockpool.o.OBJ { lockpool.cpp.CPP } } } error: created from libs/atomic/build/boost_atomic error: another virtual target { darwin%darwin.link.dll-libboost_atomic.dylib.SHARED_LIB { darwin%darwin.compile.c++-lockpool.o.OBJ { lockpool.cpp.CPP } } } error: created from libs/atomic/build/boost_atomic error: added properties: <toolset-darwin:version>4.8.1x <toolset>darwin error: removed properties: <toolset-clang:platform>darwin <toolset-clang:version>3.2xl <toolset>clang /Users/viboes/github/modular-boost/tools/build/v2/build/virtual-target.jam:484: in actualize-no-scanner from module object(file-target)@16688 /Users/viboes/github/modular-boost/tools/build/v2/build/virtual-target.jam:134: in class(a)virtual-target.actualize from module object(file-target)@16688 /Users/viboes/github/modular-boost/tools/build/v2/build-system.jam:720: in load from module build-system /Users/viboes/github/modular-boost/tools/build/v2/kernel/modules.jam:289: in import from module modules /Users/viboes/github/modular-boost/tools/build/v2/kernel/bootstrap.jam:139: in boost-build from module /Users/viboes/github/modular-boost/boost-build.jam:17: in module scope from module

3 3

[conversion] GitHub boostorg repos now open for regular use
by Beman Dawes 02 Dec '13

02 Dec '13

Please be cautious at first, and report any problems. It is probably a bit surer to follow the full clone instructions at https://svn.boost.org/trac/boost/wiki/TryModBoost rather than trying to make use of a previous clone. Also note that those instructions have been corrected several times, the most recent correction just a minute ago. We will be setting up write permissions, but that may take a day or so. Thanks, --Beman

1 0

[conversion] Progress Report
by Beman Dawes 02 Dec '13

02 Dec '13

We are very close to being able to go live with boostorg at GitHub. * Following the instructions at https://svn.boost.org/trac/boost/wiki/TryModBoost, the "master" branch seems ready to go. * A couple of libraries on the "develop" branch may still have end-of-line issues. If so, those should be resolved first thing in the morning. So please be patient one more day, and then we should be able to get to work using git and modular boost. --Beman

8 12

circular buffer patch
by Serge A. 02 Dec '13

02 Dec '13

It appears that the circular_buffer no longer has a maintainer. I have a patch that fixes a problem introduced in the latest versions of the circular buffer that it can not correctly work with a custom memory allocator. The patch is provided here: https://svn.boost.org/trac/boost/ticket/9334 The former maintainer suggested to submit a request to this list, so that the patch is included in the daily build and the next release. Thanks! Serge

2 1

[COUNTERTREE + SUBALLOCATORS] Errors corrected, .New Lock system, manycore strategies
by Francisco José Tapia 02 Dec '13

02 Dec '13

I have corrected several errors ( thanks to the people who sent me comment, suggestions and bugs). The lock mechanism ( 1 Writer / N Readers) have been redesigned in order to correct previous errors and increment the speed. The actual version is built over spinlocks and yield functions. It provide an alternative to the upgrade mutex. You have a detailed description with diagrams and examples in https://dl.dropboxusercontent.com/u/8437476/works/countertree_2.0/doc/singl… *LIBRARY DESCRIPTION* This library provide a concurrent and thread safe implementation of the *tree based data structures* (set, multiset, map and multimap, plus a unordered tree named vector_tree) with *random access iterators* and *access by position like a vector*. This permit to *distribute the elements* stored in the trees, between an *arbitrary number of threads*, identically as do with the vectors. Over it you can use many algorithms designed for vectors ( as quicksort over an unsorted tree) and *making easy the design of parallel algorithms* applied to the tree based data structures , and their use by the multicore development tools.. [image: Imágenes integradas 1] These data structures are based on a kind of augmented red-black trees, where each node have a counter with with the number of nodes under which permit the access by position, like in a vector, to the elements stored in the tree, and random access iterators. I named this tree of counters *countertree* Until now, in many practical problems, the binary trees was not used , by the difficult for to distribute the elements stored between an arbitrary number of cores, and the difficult for to develop parallel algorithms. This limitation force to use only 1 thread with these data structures. And with big problems, usually it is not a good solution. With the countertrees, perhaps it would be interested reevaluate some algorithms, and make an comparison between the present algorithms and the possible options, in order to see the performance, the memory used and the flexibility. If you are interested, you have a brief document ( 2 pages) about the manycore strategies<https://dl.dropboxusercontent.com/u/8437476/works/doc/Manycore_Strategies.p…>with this library, and other brief project description<https://dl.dropboxusercontent.com/u/8437476/works/doc/Brief_project_descrip…>(2 pages) where provide a global vision of the project and their parts. I think is a good starting point in order to examine the library. *You can find the code and the documentation in my web pages in dropbox* *https://dl.dropboxusercontent.com/u/8437476/works/countertree_2.0/index.html <https://dl.dropboxusercontent.com/u/8437476/works/countertree_2.0/index.html>* *or if you prefer in a git format* *https://github.com/fjtapia/countertree_2.0 <https://github.com/fjtapia/countertree_2.0>* Please, if you have any idea , any comment or see any fail, any lack, please say me in order to improve the library, and provide between all ,a fast, robust and reliable tool to the C++ programmers. If you want more information, need something, please, say me, and I will try to do. Thanks by your interest. Yours Francisco Jose Tapia e-mail : fjtapia(a)gmail.com

1 0

Subversion repository closed for all commits on all branches
by Beman Dawes 02 Dec '13

02 Dec '13

The Subversion repository is now closed for all commits on all branches. An attempt to commit will fail a pre-commit hook, resulting in the following error message: Command: Commit Modified: C:\boost\trunk\libs\system\doc\index.html Sending content: C:\boost\trunk\libs\system\doc\index.html Error: Commit failed (details follow): Error: Commit blocked by pre-commit hook (exit code 1) with output: Error: This Subversion repository is no longer in active use. Error: Error: It has been superseded by a Git super-project repository Error: and individual Git repositories for each Boost library. Error: Error: See: https://github.com/boostorg Completed!: The next step in the conversion to Git and Modular Boost is to run the conversion script one last time, and then turn it off. We will let you know when that is completed. -- The Release Managers

3 3

[git-help][eclipse] Fetching the full repo?
by Rene Rivera 02 Dec '13

02 Dec '13

First, I'm asking about using git from within Eclipse since that's where I do everything.. I've had a Boost master clone for a long time now. And having a very small amount of time I decided to start updating and looking at the latest state of the repo. Unfortunately everything I attempt to get my local checkout to match the head state of the repo is failing. In particular the subrepos are not being fetched or refreshed. And some of the subrepos are missing and hence show up as pending delete changes. Does anyone know how to get the Eclipse git to fetch/reset to the head fully? Rene. -- -- -- 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

1 1

[boost.random] Dependencies/modularization
by Philip Bennefall 01 Dec '13

01 Dec '13

Hi all, I know that I have brought up this subject before and that work is being done to improve the situation, but I just wanted to ask about a specific case that I came across today while attempting to extract a subset of Boost. I want to use mersenne_twister.hpp, uniform_int_distribution.hpp and uniform_real_distribution.hpp from boost.random as part of an open source library that I am developing. Since the only part of Boost that I am using in that library is the Mersenne Twister prng, I figured I would ship the relevant headers along with my own source distribution. I used bcp to scan my code which includes these three headers, and I get a directory which is 2.89 MB in size and which contains 304 files. I looked at what is being included, and in order of size the worst offenders are as follows: preprocessor - 1.82 MB type_traits - 604 KB config - 250 KB The other libraries that are included are tiny, and I would not consider config to be a problem either. Clearly, preprocessor is the worst offender by far here in terms of bloat. I wonder if there is an easy way to get rid of the dependency on that library without hacking Boost too much? As for type_traits, it seems a bit large for what I gather it is being used for in this case but is not the end of the world either. Does anyone have any suggestions? Will the modularization of Boost reduce these dependencies in the not too distant future? I know that a huge effort is being spent moving Boost to Git, but I am curious to know whether there is a roughly estimated timeline for the modularization work? Kind regards, Philip Bennefall

1 0

[type_traits] Why is there no total extent?
by Mostafa 01 Dec '13

01 Dec '13

I was wondering why there is no total_extent in the TypeTraits library. total_extent would represent the total number of elements in a complete built-in array type, else it would be zero. I would think that this would be a natural supplement to the existing extent metafunction. (There is a very similar functionality in the smart pointers library, under the name of array_total here: smart_ptr/detail/array_traits.hpp) Mostafa

1 0

ZeroMQ integration with Boost-v1.54
by Rahul Mathur 01 Dec '13

01 Dec '13

All, Looking to have ZeroMQ integration with Boost-v1.54. I am familiar with Boost.ASIO client and Server model. BTW, How do I proceed to have ZeroMQ integrated with Boost-v1.54 with GCC on Linux platform?

1 0