August 2017 - Boost - lists.preview.boost.org

boost_check_library.py: correct handling of header-only libs?
by Vinnie Falco 10 Aug '17

10 Aug '17

This script fails for me: <https://github.com/boostorg/boost/blob/master/status/boost_check_library.py…> Because there is no libs/beast/build. But Beast is header-only and has no src/ directory, is this correct? Apologies for posting to the list, but there is no Issues tab on the boostorg/boost repo. Thanks

3 2

Boost 1.65.0 Release Candidate 1
by Marshall Clow 09 Aug '17

09 Aug '17

The release candidates for the 1.65.0 release are now available at: https://dl.bintray.com/boostorg/release/1.65.0/source/ The SHA256 checksums are as follows: 0f1352a7738f24b39b564263b66a8c38bc7d984b57406396400474602e961ad3 boost_1_65_0_rc1.7z c2a022d4d6706668ea94b8e799a76e38929e0e34db884ef451fa9d5db6103238 boost_1_65_0_rc1.tar.bz2 7e84a27276792209d48357a0671348768dbf4dd1d4eadbf01b855274208f148d boost_1_65_0_rc1.tar.gz 537c42771b2351b4ea3eb8e9a13a3423edcd155323d2c14c4606cd23fc22d09f boost_1_65_0_rc1.zip As always, the release managers would appreciate it if you download the candidate 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

7 12

[filesystem] Pull request needs attention
by Robert Dailey 09 Aug '17

09 Aug '17

I have fixed a crash I found due to static initialization ordering in boost filesystem. I have a proposed fix here: https://github.com/boostorg/filesystem/pull/47 It hasn't gotten any attention yet, so I'm hoping someone can offer some advice and/or merge the PR for me. Thanks in advance.

3 3

Boost UUID Maintainer
by James E. King, III 08 Aug '17

08 Aug '17

Greetings! I'm a long-time boost user and over the years I have posted defects into boost trac for the components I have used. I was reviewing the backlog on boost::uuid and in trac and it needs a bit of grooming. I started submitting some PRs to resolve some of these. Edward D. in the CMT group was reviewing some of them and mentioned that there is no active maintainer for the library, and invited me to introduce myself here. I am currently a committer on the Apache Thrift project and I have 20+ years development experience with C++. I have good experience at backlog grooming, and I am very focused on quality. I believe firmly in testing all code branches, and using coverage tools to prove it. I also assisted the glibc team in getting started with Coverity Scan which helped lead to resolution of 138 defects at last count. My current position is a Software Architect at HPE SimpliVity where I have been for the last ~7 years (I'm easy to find on LinkedIn, hopefully). I live in Westford, MA, USA. Given there is no maintainer for boost::uuid currently, I would like to petition for that responsibility. It is a relatively small library in the portfolio and seems like a good introduction to being a maintainer in such a widely used project. Thanks for your consideration, Jim King

5 7

[stacktrace] 1.65.0 RC1 fails to build for MSYS2/mingw-w64
by Stephan Lavavej 07 Aug '17

07 Aug '17

Hi, I just attempted to build Boost 1.65.0 RC1 with MSYS2/mingw-w64. (My build script for 1.64.0 is available at https://github.com/StephanTLavavej/mingw-distro/blob/6e5976af8c63de93bec761… ; I just had to change the filenames for 1.65.0 RC1.) The stacktrace build failed with: gcc.compile.c++ bin.v2\libs\stacktrace\build\gcc-mingw-7.1.0\release\link-static\runtime-link-static\threading-multi\windbg_cached.o In file included from libs\stacktrace\build\..\src\windbg_cached.cpp:10:0: ./boost/stacktrace/detail/frame_msvc.ipp:167:72: warning: missing terminating ' character # error Your compiler does not support C++11 thread_local storage. It's impossible to build with BOOST_STACKTRACE_USE_WINDBG_CACHED. ^ ./boost/stacktrace/detail/frame_msvc.ipp:167:5: error: #error Your compiler does not support C++11 thread_local storage. It's impossible to build with BOOST_STACKTRACE_USE_WINDBG_CACHED. # error Your compiler does not support C++11 thread_local storage. It's impossible to build with BOOST_STACKTRACE_USE_WINDBG_CACHED. ^~~~~ In file included from ./boost/detail/winapi/file_management.hpp:17:0, from ./boost/stacktrace/detail/safe_dump_win.ipp:20, from ./boost/stacktrace/safe_dump_to.hpp:206, from ./boost/stacktrace/frame.hpp:20, from ./boost/stacktrace/detail/frame_msvc.ipp:15, from libs\stacktrace\build\..\src\windbg_cached.cpp:10: ./boost/detail/winapi/overlapped.hpp:39:9: warning: ISO C++ prohibits anonymous structs [-Wpedantic] }; ^ "g++" -O3 -finline-functions -Wno-inline -Wall -pedantic -mthreads -m64 -DBOOST_ALL_NO_LIB=1 -DNDEBUG -I"." -c -o "bin.v2\libs\stacktrace\build\gcc-mingw-7.1.0\release\link-static\runtime-link-static\threading-multi\windbg_cached.o" "libs\stacktrace\build\..\src\windbg_cached.cpp" ...failed gcc.compile.c++ bin.v2\libs\stacktrace\build\gcc-mingw-7.1.0\release\link-static\runtime-link-static\threading-multi\windbg_cached.o... It appears that frame_msvc.ipp is intended to work with both MSVC and MinGW, but it's not correctly recognizing mingw-w64 GCC 7.1.0's behavior. This compiler supports thread_local, although it doesn't support C++11's multithreading headers (without the pthreads wrapper, which I don't build). Either this library should correctly detect mingw-w64, or it should be disabled for mingw-w64. Thanks, STL

2 2

[config] detect std::unique_ptr
by Artyom Beilis 07 Aug '17

07 Aug '17

Hello, I'm searching for a macro to detect reliably existence of std::unique_ptr I didn't found one in boost.config Artyom

2 2

[C++ committee] auto_ptr removal in C++17 - request to comittie
by Artyom Beilis 07 Aug '17

07 Aug '17

Hello, If somebody on committee is on this list. I really want to do something to prevent auto_ptr removal. It is exceptionally useful smart pointer in C++2003 and it is needed by the code that needs to be compatible for both 2003 and 2017. It is needed by libraries that aim to be compatible with C++2003. In the real world there is a huge codebase that uses C++2003 why to break something that works - especially since it works quite well. It isn't "gets" that can't be used securely. C++ committee please stop breaking existing code even if std::unique_ptr much better it does not exist in C++2003. Artyom

3 2

Microsoft vs The Boost License
by Phil Bouchard 06 Aug '17

06 Aug '17

Hi, Just to put everybody in context: 1) Microsoft had invested into Corel so that they get rid of Corel Linux back in 2000 when I was working there: https://www.forbes.com/2000/10/03/1003corel.html 2) Microsoft once again copied my Fornux PowerCalc after I had presented it to them using some web interface: http://www.fornux.com/ With their Microsoft PowerToys: https://en.wikipedia.org/wiki/Microsoft_PowerToys#Included_applications 3) And now here I am with Microsoft trying to copy the idea behind root_ptr's node_proxy: https://github.com/philippeb8/root_ptr With their "deferred_ptr heap": https://github.com/hsutter/gcpp So for some reason I do not trust Microsoft. And I was wondering if the Boost license protects us from an idea behind a library we wrote. Otherwise this makes the Boost license not very useful. Thanks a lot for your help! -Phil

9 20

[BB++] Documentary - Take 3
by Phil Bouchard 06 Aug '17

06 Aug '17

In the following video, I am explaining how objects of a lower scope are being promoted to a higher scope (please fast forward to 10:15): https://youtu.be/OBnpFUVEVe4 I added 5 slides explaining how objects are being promoted and I've renarrated everything after 6:21. I'm keeping the beginning because I am targeting newbies on the subject as well. I hope that's a good explanation because sharing this information publicly is a point of no return for me. Sincerely, -Phil

1 0

Submodule name mismatches between master and develop
by Peter Dimov 05 Aug '17

05 Aug '17

Several submodules have different names on master and develop: - compute (libs/compute on develop) - couroutine2 (libs/coroutine2 on develop) - dll (libs/dll on develop) - qvm (libs/qvm on develop) - vmd (libs/vmd on develop) Should we normalize the names on develop? I think that after fixing .gitmodules, these submodules will appear as-if deinitialized, and would need a `git submodule update --init` to reappear under the new names. There are also libs/hana and libs/metaparse, which have consistent (although "wrong") names on both master and develop. Not sure if we need to bother with them.

6 8