- Boost - lists.preview.boost.org

[fiber] Suggestions regarding asio
by Tatsuyuki Ishi 22 Sep '16

22 Sep '16

The current example seems non-flexible and very hacky. 1. The timer handler has no problem, since it's expected to be called from any thread. (Description in "Deeper Dive into asio" is meaningless) 2. Possible alternatives: * A post(empty lambda) is a little bit waste but solves the threading problem. It wakes up one thread by doing interrupt of epoll etc. Maybe adding a interrupt method to asio codebase could be more efficient. * One thread doing event polling and the other doing condvar is enough. Since we're not expecting a slow handler (most times, it just make a fiber ready), this should be sufficient. Correct me if I'm wrong.

2 1

[interprocess] Improved shared_memory lifetime using file locks?
by Sebastian Theophil 21 Sep '16

21 Sep '16

Hi, coming from Windows, the filesystem persistence of basic_managed_shared_memory is giving us headaches. When we have N processes accessing the same shared memory, and one of them crashes, the other N-1 are in an undefined state. That’s the same on Windows and POSIX platforms and is ok for us. On Windows, the problem resolves itself when the other N-1 processes are all closed and then restarted, because Windows deletes the shared memory automatically. On POSIX systems, I think we can use file locks to make sure that the shared memory is reinitialized by the first process to access it. I’d appreciate any feedback, especially if it points out the problems with this approach: - basic_managed_shared_memory(boost::interprocess::open_or_create, …) opens the backing file with flags O_CREAT | O_EXLOCK | O_NONBLOCK - if that succeeds: - we are the first - (re-) initialize shared memory - also a good time to unlink/open the named POSIX semaphore that’s hanging around in an invalid state too! - degrade exclusive lock using flock(LOCK_SH) - else - open file with flag O_SHLOCK (open only, and blocking!) - when that succeeds, open shared memory Problems I can see: - Only works when (advisory) shared and exclusive locking and the degradation of one to the other is supported (at least flock not supported on Windows, e.g.) - What about other file systems? NFS? - Somebody could lock all your files. (OTOH, sb could also delete all my backing files all the time. My application would run, but not have any shared memory. Might be an advantage that my application does *not* run when the file is locked.) - What else? Is this worth implementing in boost.interprocess? I appreciate any feedback, Sebastian PS: I’ve googled previous discussions about shared memory lifetime in boost.interprocess. I don’t claim that my proposal gives the same semantics as Windows shared memory. It doesn’t and it seems impossible to achieve. I’m looking for, erm, usable semantics on POSIX systems. Clearly, everybody has to implement some lifetime tracking. It would be nice if boost.interprocess provided one or more solutions. -- Dr. Sebastian Theophil | stheophil(a)think-cell.com Senior Software Engineer We are looking for C++ Developers: https://www.think-cell.com/career think-cell Software GmbH | Chausseestr. 8/E | 10115 Berlin | Germany https://www.think-cell.com | phone +49 30 666473-10 | US phone +1 800 891 8091 Amtsgericht Berlin-Charlottenburg, HRB 180042 Directors: Dr. Markus Hannebauer, Dr. Arno Schödl

1 0

[release] 1.62.0 Master branch is closed
by Rene Rivera 21 Sep '16

21 Sep '16

The master branch is now closed for all changes, except by permission from a release manager. Thanks, The Release Team -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

1 0

[release] Boost 1.62.0. Beta 2
by Rene Rivera 20 Sep '16

20 Sep '16

The packages for 1.62.0 beta 2 are now available: <https://sourceforge.net/projects/boost/files/boost/1.62.0.beta.2/> The SHA256 checksums are as follows: 95feae3620b3e604e93f14d97195fc04a4b67063692859af8df19657d41b82d5 boost_1_62_0_b2.7z 27bc9b328ff73d7dbf80cd9d4b5f79b0878221e5d8e4b6c47444b135edc1080a boost_1_62_0_b2.tar.bz2 3939d5efa683449e87b374d7e92e3eeb6490063ddb2877919f0fad26e3a0f53e boost_1_62_0_b2.tar.gz cff9ab5db0928244c4c836cf3da5140e6e93db7756e2791ba99a21104ad3aa72 boost_1_62_0_b2.zip As always, the release managers would appreciate it if you download the archive of your choice and give building and using it a try. Please report both success and failure, and anything else that is noteworthy. Thanks! -- The release managers -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

4 9

Re: [boost] first steps to submitting - boost :: observers
by Bjorn Reese 19 Sep '16

19 Sep '16

On 09/17/2016 08:10 PM, Robert McInnis wrote: > https://github.com/tiny/boost_observers.git What are the advantages of this library over Boost.Signals2 [1]? Same question with regards to the Boost.Synapse [2] proposal. My impression from a brief look at your repository is that you need to focus more on thread-safety. Is there any way to avoid the overhead of thread synchronization in a single-threaded application? Observers are invoked while the mutex is locked, so observers cannot make calls on the subject. How does the lock-free mutex perform, especially when congested? Why is the lock-free mutex recursive? There are several member variables that are used as if they are atomic but they are not declared as such, e.g. Subject::_block, Numeric<T>::_x (think T == long long.) The Scope template can be replaced by lock_guard if you model the LockFreeMutex after the BasicLockable concept [3]. The thread_self() and thread_pid() functions can be replaced by this_thread::get_id(). Replace the hAtomic macro with a type alias. Use <cstdint> instead of <stdint.h> to reduce polution of the global namespace. [1] http://www.boost.org/doc/html/signals2.html [2] http://zajo.github.io/boost-synapse/ [3] http://en.cppreference.com/w/cpp/concept/BasicLockable

4 6

[density] version 7
by Giuseppe Campana 19 Sep '16

19 Sep '16

Hi to all, I’ve just merged to the master branch a major review of density, an header-only library of heterogeneous dense containers. Unfortunately this update includes some renamings. Anyway the new names are simpler and more explicative: heterogeneous_array, heterogeneous_queue, function_queue, lifo_array. The update consists of fixes and optimizations. The results of the benchmarks are included in the documentation. The container function_queue gives the best results, being many times faster than an std::queue of std::function’s (http://peggysansonetti.it/tech/density/html/func_queue_bench.html) The library is not yet ready for asking a formal review, but hopefully it will be in a few months. Html overview (now a little more schematic) and reference doc at http://peggysansonetti.it/tech/density/html/index.html Repository at https://github.com/giucamp/density/tree/master Giuseppe Campana.

1 0

Fuzzy string search to the Boost String Algortihm
by Зайцев Александр 19 Sep '16

19 Sep '16

Hello. I want to discuss about some proposals. I am working on Levenshtein distance and Damerau-Levenshtein distance algortihms. I hope, that these implementations will be pulled to the Boost String library. But i think that we can add to the Boost String Library some more algorithms for fuzzy searhing. I mean these algortihms: Levenshtein distance Damerau-Levenshtein distance Bitap algorithm with modifications by Wu and Manber Spell-checker method N-gram method Signature hashing method BK-trees What do you think about this idea? If you think that it can be implemented to the Boost String Algo, i will start work on this list immediately. Best regards, Zaitsev Alexander.

2 2

first steps to submitting - boost :: observers
by Robert McInnis 19 Sep '16

19 Sep '16

G'afternoon, This is my first time submitting to a public repo, please be gentle I'd like to submit a series of classes I have been using since '90. The initial set implements a thread safe subject/observer pattern. I have included a handful of example programs, two fairly trivial and the third more in-depth. All examples are single file examples to make compilation trivial. This is the first re-work of my original tools, making them more boost-friendly. I expect to add more default observer templates and observable objects, but that will come with time. WHY? -- The reason for these classes is to help insure the mathematical integrity of your data model in relation to your business rules. If at any time you were to change one of your data elements without updating all those dependent on it, then your data integrity would be compromised. These classes look to solve that. As an example: -- using namespace boost::observers ; Numeric<long> x, y, z ; // rules: y = x + 2; z = y * 3 ; x = 5 ; printf( "%d \n", z ) ; x = 1 ; printf( "%d \n", z ) ; -- Setting the rules properly, the classes allow for 'z' to be correct each time it's printed using code almost exactly like this. This will remove the possibility of accidently forgetting to trigger the rule update. WHERE? -- The repo can be found here: git clone https://github.com/tiny/boost_observers.git Any input on how to continue from this point would also be much appreciated. I have skype and ventrilo servers available if that helps. Thanks in advance, Rob -- Robert McInnis

5 7

[release] Boost 1.62.0 - One week to close.
by Rene Rivera 16 Sep '16

16 Sep '16

Reminder.. Next week, on the 21st, the release will close for all changes. Except for extra ordinary and approved changes. Please exercise caution when merging changes to master branches. Remember that changes need to be well tested on the develop branches before going into master. And that changes at this point should be for critical problems. Thanks, the Release Team. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

2 2

[uuid] PR to remove auto-linking to advapi32 when BOOST_ALL_NO_LIB is defined
by Edward Diener 15 Sep '16

15 Sep '16

I don't know how to resolve a PR for uuid ( https://github.com/boostorg/uuid/pull/11 ). If a library exists only as a DLL and not as a static library should BOOST_ALL_NO_LIB still be considered in effect ? My viewpoint is 'no' but the person who created the PR believes that the BOOST_ALL_NO_LIB config macro should turn off any #pragma auto-linking in code even when their is no other way to use a library other than as a shared library. What I am afraid of is breaking user code even if BOOST_ALL_NO_LIB is defined by producing link errors. But perhaps I am wrong and this should be considered justified when BOOST_ALL_NO_LIB is being used.

5 7

[algorithm] Musser-Nishanov sequence search algorithm
by Jeremy Murphy 15 Sep '16

15 Sep '16

Dear Boost community, most of you are probably familiar with Dave Musser, well known for his work on generic programming, introsort and more. What you maybe didn't know, and I didn't until not that long ago, is that he and Gor Nishanov also designed (based off KMP) and wrote a generic sequence search algorithm to compete with Boyer-Moore, etc, which apart from being competitively fast is ostensibly more generic than the others. You can read for yourself on Dave Musser's website: http://www.cs.rpi.edu/~musser/gp/algorithms.html By "more generic" I mean it has a fallback algorithm if the corpus does not have random-access iterators, and its behaviour can easily be tuned for the alphabet size to dramatic effect. (I haven't looked into the 'tunability' of BM, BMH, KMP, etc, so I'm not sure if this is a unique feature.) But at face value, it appears to have powerful features that the others lack without compromising performance, which is pretty neat. So I've been in contact with the original authors and with their permission am just playing the messenger to bring the algorithm to wider attention with the expectation that it is something we could really benefit from. What I hope for now is to have discussion on whether indeed it is an algorithm that we want, presumably alongside the others in Boost.Algorithm. My slightly more controversial thought is that, if no shortcomings or weaknesses are revealed, this could up being the superior search algorithm. I've opened a pull request on Boost.Algorithm with the code so everyone can play around with it if they want. https://github.com/boostorg/algorithm/pull/25 Looking forward to hearing what everyone thinks. Cheers. Jeremy

1 0

[uuid][rational] Pushed bug fixes to 'master'
by Edward Diener 15 Sep '16

15 Sep '16

I pushed some bug fixes from 'develop' to 'master' for the uuid and rational libraries. The fixes had been tested on the 'develop' branch for a long time and the results showed that the fixes were fine. While none of the fixes were major I felt that the fixes were important enough to go in the upcoming Boost release if they are going to be picked up. I, and probably other members of CMT, should have been more proactive in pushing these fixes to 'master' sooner, once they showed green on the develop test matrix.

2 3

Final Call for Extended Abstracts and Posters: International Conference on Information Society (i-Society 2016), Technical Co-Sponsored by IEEE UK/RI Chapter!
by David Brown, i-Society 2016 14 Sep '16

14 Sep '16

1 0

[log] Boost 1.62 cannot compile log with Oracle Developer Studio 12.5
by Aparna Kumta 14 Sep '16

14 Sep '16

Compiling Boost.log with Oracle Developer Studio 12.5, we see the following failure: % "CC" -std=c++11 -library=stdcpp,CrunG3 -xO4 -mt -erroff=%none -m32 -KPIC -DBOOST_ALL_NO_LIB=1 -DBOOST_ATOMIC_DYN_LINK=1 -DBOOST_CHRONO_DYN_LINK=1 -DBOOST_DATE_TIME_DYN_LINK=1 -DBOOST_FILESYSTEM_DYN_LINK=1 -DBOOST_LOG_BUILDING_THE_LIB=1 -DBOOST_LOG_DLL -DBOOST_LOG_HAS_PTHREAD_MUTEX_ROBUST -DBOOST_LOG_WITHOUT_DEBUG_OUTPUT -DBOOST_LOG_WITHOUT_EVENT_LOG -DBOOST_SPIRIT_USE_PHOENIX_V3=1 -DBOOST_SYSTEM_DYN_LINK=1 -DBOOST_SYSTEM_NO_DEPRECATED -DBOOST_THREAD_BUILD_DLL=1 -DBOOST_THREAD_DONT_USE_CHRONO=1 -DBOOST_THREAD_POSIX -DBOOST_THREAD_USE_DLL=1 -DDATE_TIME_INLINE -DNDEBUG -D_XOPEN_SOURCE=500 -D__EXTENSIONS__ -D__STDC_CONSTANT_MACROS -D__typeof__=__typeof__ -I"../../.." -I"../../../libs/log/src" -c -o "../../../bin.v2/libs/log/build/sun/release/stdlib-gnu/threading-multi/posix/ipc_reliable_message_queue.o" "../../../libs/log/src/posix/ipc_reliable_message_queue.cpp" "../../../libs/log/src/posix/ipc_reliable_message_queue.cpp", line 61: Error: #error Boost.Log: Native 32-bit atomic operations are required but not supported by Boost.Atomic on the target platform. This seems to be a configuration issue. The following changes seems to work: 1) In boost/atomic/detail/platform/platform.hpp % !diff diff ./platform.hpp ./platform.hpp_orig 46c46 < #elif (defined(__GNUC__) || (__SUNPRO_CC >= 0x5130)) && (defined(__i386__) || defined(__x86_64__)) --- > #elif defined(__GNUC__) && (defined(__i386__) || defined(__x86_64__)) 68c68 < #elif (defined(__GNUC__) || (__SUNPRO_CC >= 0x5130)) && defined(__sparc) --- > #elif defined(__GNUC__) && defined(__sparc_v9__) The above change causes the compilation to proceed further, however, it fails again with another configuration issue causing failure to build atomic library "CC" -std=c++11 -library=stdcpp,CrunG3 -xO4 -mt -erroff=%none -m32 -KPIC -DBOOST_ALL_NO_LIB=1 -DBOOST_ATOMIC_DYN_LINK=1 -DBOOST_ATOMIC_SOURCE -DNDEBUG -I"../../.." -c -o "../../../bin.v2/libs/atomic/build/sun/release/stdlib-gnu/threading-multi/lockpool.o" "../../../libs/atomic/src/lockpool.cpp" "../../../boost/atomic/detail/ops_gcc_sparc.hpp", line 69: Error: The function "fence_before_store" must have a prototype. In boost/atomic/detail/ops_gcc_sparc.hpp, the following change seems to resolve the issue % !diff % diff ./ops_gcc_sparc.hpp ./ops_gcc_sparc.hpp_orig 69c69 < fence_before(order); --- > fence_before_store(order); 71c71 < fence_after(order); --- > fence_after_store(order); 110c110 < fence_before(order); --- > base_type::fence_before(order); 118c118 < fence_after(order); --- > base_type::fence_after(order); I will submit a PR for this change. Thanks, Aparna

1 0

[filesystem] How to remove specific files from a directory?
by Andrzej Krzemienski 14 Sep '16

14 Sep '16

Hi, I wonder what is boost::filesystem recommendation for solving the following problem. I want to remove all files from a given directory that satisfy a certain predicate, e.g., only these whose names start with letter "s". It is my understanding that the calling filesystem::remove may invalidate the iterator, and therefore the following solution is incorrect: fsys::directory_iterator it{path}, itEnd{}; for ( ; it != itEnd ; ++it ) { if (predicate(*it)) fsys::remove(it->path()); } But, is the following guaranteed to work?: fsys::directory_iterator it{path}, itEnd{}; for ( ; it != itEnd ; ) { if (predicate(*it)) fsys::remove((it++)->path()); else ++it; } If not, does there exist a dedicated solution for solving this problem that would not require of me N traversals through the directory? Thanks, &rzej;

6 11

[release] Boost 1.62.0 Beta 1
by Rene Rivera 14 Sep '16

14 Sep '16

The first beta packages for 1.62.0 beta 1 are now available: <https://sourceforge.net/projects/boost/files/boost/1.61.0.beta.1/> The SHA256 checksums are as follows: f7cc83e6a3b03cd7aa8822992aa3e36422cc2214644d69277ef55a4db51f44bd boost_1_62_0_b1.7z 4057993811e40b1fd4719675e44c2550b6c2379546192da2299c17fec9a3dddf boost_1_62_0_b1.tar.bz2 fb0c10b625c5ec47cc837b54740540e400c59571b2eabf427c00f32bffc6ca70 boost_1_62_0_b1.tar.gz b38ef21c43c5ff814a2144bc23871c134164e7f51c71672d0f49a44636fb50b5 boost_1_62_0_b1.zip As always, the release managers would appreciate it if you download the archive of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. Thanks! -- The release managers -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

4 4

Re: [boost] [Boost-users] [release] Boost 1.62.0 Beta 1
by Rene Rivera 13 Sep '16

13 Sep '16

On Mon, Sep 12, 2016 at 12:44 AM, Oliver Kowalke <oliver.kowalke(a)gmail.com> wrote: > 2016-09-12 0:00 GMT+02:00 Tom Kent <lists(a)teeks99.com>: > >> Since the fiber build was a failure, >> > > should be fixed > Could someone try a build of the master snapshot to see if it's really fixed? I'm still in favor of a beta 2 build to make sure that gets flushed out >> before going to RCs. >> > > +1 > We can certainly do that. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

2 1

[math] Undefined reference to promote_args
by Florian Lindner 13 Sep '16

13 Sep '16

Hello, when compiling a file that uses float_distance from #include <boost/math/special_functions/relative_difference.hpp> I get a linking error: undefined reference to `boost::math::tools::promote_args<double, double, float, float, float, float>::type boost::math::float_distance<double, double>(double const&, double const&)' I try to add all boost_math libs I could find, but none helped: -lboost_system -lboost_filesystem -lboost_math_tr1 -lboost_math_c99 -lboost_math_c99 Since float_distance is not a C99 / TR1 function, shouldn't boost.math work as header only? What is the problem here? Thanks, Florian

2 1

[qi][endian]BOOST_ENDIAN_FORCE_PODNESS is defined in Qi
by Takatoshi Kondo 12 Sep '16

12 Sep '16

Hi, I'm using Boost.Endian with Boost.Qi. It seems that BOOST_ENDIAN_FORCE_PODNESS is defined at boost/spirit/home/support/detail/endian.hpp http://www.boost.org/doc/libs/1_61_0/boost/spirit/home/support/detail/endia… And boost/spirit/include/qi.hpp includes boost/spirit/home/support/detail/endian.hpp According to the Boost.Endian document, endian_buffer's constructors are disappeared by BOOST_ENDIAN_FORCE_PODNESS. http://www.boost.org/doc/libs/1_61_0/libs/endian/doc/buffers.html I'd like to use both endian_buffer's constructors and Boost.Qi. Is there any good way to do that? I don't know much abour Boost.Qi, but defining BOOST_ENDIAN_FORCE_PODNESS in the header file makes big side effects. For my workaround, I changed my include order. Sucessfully compiled http://melpon.org/wandbox/permlink/gSijxeBFJorWpzSd Error ocuured http://melpon.org/wandbox/permlink/FVffqcw8JEKGFXU0 --- Takatoshi Kondo

1 0

[property_tree] arrays in info files ?
by alainm 11 Sep '16

11 Sep '16

Hi, Is there a simple way to specify an array value with the info format, something like arr { 42 "" 43 "" 44 "" } but without the empty strings. thanks

1 0

docca: C++ Documentation XSLT library
by Vinnie Falco 09 Sep '16

09 Sep '16

Hi. When I developed Beast for Boost.Asio (https://github.com/vinniefalco/Beast) I lifted the xsl transforms from Boost.Asio (thanks Chris!) and adapted them. These scripts take the output of Doxygen and transform them into Boost.Book code (or QuickBook code or whatever its called). In other words, it creates a file called reference.qbk. This file is included in the documentation, which gets processed through a Jamfile to produce the HTML documentation that you see here: http://vinniefalco.github.io/beast/index.html This is great, and I am using the same system for my new key/value database engine (https://github.com/vinniefalco/NuDB/) The problem is that now I want to start another project and use the same documentation, but there is a considerable amount of copying going on here. The entire reference.xsl file is copied and placed in the new project: https://github.com/vinniefalco/NuDB/blob/master/doc/reference.xsl This file has to be edited significantly because it has references to project-specific namespaces, prefixes, and various other annoying minutiae. Therefore, I have started a new project called "docca:" https://github.com/vinniefalco/docca This project aims to refactor my variant of Boost.Asio's reference.xsl file so that it can be re-used in other projects. The procedure would be to create your own empty .xsl file, and include docca's xsl file after defining some variables. For example: <?xml version="1.0" encoding="utf-8"?> <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0"> <xsl:variable name="namespace" select="'ripple::'"/> <xsl:include href="../include/docca/doxygen.xsl"/> </xsl:stylesheet> Above is the file you provide, while include/docca/doxygen.xsl is provided by docca (This is very much a work in progress). By this point many of you have fallen asleep, XSLT is a great cure for insomnia. If you're still awake you're probably very excited at the prospects of a turn-key solution for making great looking docs out of Javadoc comments in your C++ headers so I'll move right on to the questions: 1. Is there already a system for doing this? 2. If not, is there any interest in such a system? Is this project worth the effort? 3. Is this approach viable? Or is there a different approach? 4. Are these the right tools for the job? 5. Is there anyone out there who is fairly knowledgeable with XSLT and Boost.Book / QuickBook? Disclaimer: I am by no means an expert at either Boost.Book, Quick.Book, XSLT, or bjam. My only experience with these tools is recently, to put the Beast repository together. -- Follow me on Github: https://github.com/vinniefalco

2 1

Header-only singletons (was Re: [config] RFC PR 82)
by Vinnie Falco 09 Sep '16

09 Sep '16

On Mon, Aug 1, 2016 at 4:13 PM, Vinícius dos Santos Oliveira >...I want my library header only and Boost.System requires the category to > be a single static object and it'll even use address comparison to test if the > categories are the same: > ... > Is it guaranteed the above code will return the same object even if this > code was included in different libraries used by the same program? Sorry for the late reply. This header-only idiom works: inline boost::system::error_category const& my_category() { struct my_error_category : boost::system::error_category { ... }; static my_error_category cat; return cat; } This idiom also works: template<class = void> boost::system::error_category const& my_category() { struct my_error_category : boost::system::error_category { ... }; static my_error_category cat; return cat; } Justification: >From the C++ standard (see https://isocpp.org/std/the-standard) 7.1.6 describes inline functions. It says they must exist in every TU that they are used in, and must be the same in all TUs (no diagnostic required). 7.1.6 p6 also states that the address of an (extern) inline function shall be the same in all TUs. Elsewhere, the standard specifies that inline functions are extern by default. Empirical evidence also supports these claims, both of the aforementioned idioms have been used extensively in my projects under gcc, clang, and msvc. This likely upcoming feature (maybe its already in?) further cements support for global variables in header-only libraries: http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2015/n4424.pdf However, even in the absence of additional language features, the ability to have global variables in header-only libraries exists today, with nothing more than C++11 support. Hope this helps!

1 0

[config] recent change to suffix.hpp causing several failures for Oracle Developer Studio 12.5
by Aparna Kumta 09 Sep '16

09 Sep '16

A recent change to suffix.hpp seems to cause several tests to fail when compiling with Oracle Developer Studio12.5. The error is: "CC" -std=c++11 -xO4 -mt -erroff=%none -m32 -DBOOST_ALL_NO_LIB=1 -DBOOST_CHRONO_STATIC_LINK=1 -DBOOST_SYSTEM_NO_DEPRECATED -DBOOST_SYSTEM_STATIC_LINK=1 -DBOOST_TEST_NO_AUTO_LINK=1 -DBOOST_TIMER_STATIC_LINK=1 -DNDEBUG -D__typeof__=__typeof__ -I".." -c -o ./count.o" "../libs/accumulators/test/count.cpp" "../boost/config/suffix.hpp", line 628: Error: Badly formed constant expression. <snipped> See: http://www.boost.org/development/tests/develop/developer/output/oracle-spar… % diff ./suffix.hpp ./suffix.hpp_orig 626a627,630 > # elif defined(__has_cpp_attribute) > # if __has_cpp_attribute(noreturn) > # define BOOST_NORETURN [[noreturn]] > # endif % Thanks, Aparna

2 1

[test] Test failures for gcc 4.6
by Andrey Semashev 09 Sep '16

09 Sep '16

Hi, I'm seeing these test failures for gcc 4.6 (MinGW): http://www.boost.org/development/tests/develop/developer/output/igaztanaga-… Looks like the error originates from Boost.Test, and I can also see lots of failures in Boost.Test itself for that compiler. Is this a known problem? Any fixes planned?

2 5

Gmane going away
by Tom Kent 07 Sep '16

07 Sep '16

So it looks like Gmane is under attack and the guy who runs it is thinking about just turning it off (that would be sad). https://lars.ingebrigtsen.no/2016/07/28/the-end-of-gmane/ Anyway, I know a lot of boost users use is as the gnu mailman archive pages are really hard to use. I would like to propose that we (specifically the list owner(s)) mirror the boost mailing lists to the mail-archive.com site. It has good search features and decent thread browse-ability. Thoughts? Tom

8 9