[1.55.0] Release candidate 2 available

Marshall Clow

3 Nov 2013 3 Nov '13

9:15 p.m.

RC2 files for 1.55.0 are available at http://boost.cowic.de/rc/ 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. This helps ensure the candidates build OK before we push them out to SourceForge. The files (and associated md5s) are: 93780777cfbf999a600f62883bd54b17 *boost_1_55_0_rc2.tar.gz d6eef4b4cacb2183f2bf265a5a03a354 *boost_1_55_0_rc2.tar.bz2 8aca361a4713a1f491b0a5e33fee0f1f *boost_1_55_0_rc2.zip 4e5bbc15fc8c80df8be428f8a5b5a823 *boost_1_55_0_rc2.7z Thanks! -- The release managers === Changes to release since rc1 === ---- Author: bemandawes Date: 2013-10-28 13:16:50 EDT (Mon, 28 Oct 2013) New Revision: 86503 URL: http://svn.boost.org/trac/boost/changeset/86503 Log: Merge 86426 from trunk. This should fix msvc test failures. ---- Author: bemandawes Date: 2013-10-29 09:15:03 EDT (Tue, 29 Oct 2013) New Revision: 86514 URL: http://svn.boost.org/trac/boost/changeset/86514 Log: Merge 84443 from trunk, to fix mpl assert failures for msvc in many libraries. ---- Author: eric_niebler Date: 2013-10-29 12:53:08 EDT (Tue, 29 Oct 2013) New Revision: 86516 URL: http://svn.boost.org/trac/boost/changeset/86516 Log: merge [86460] from trunk, fix bootstrap on mingw --- Author: jhunold Date: 2013-10-29 13:19:33 EDT (Tue, 29 Oct 2013) New Revision: 86519 URL: http://svn.boost.org/trac/boost/changeset/86519 Log: Merge 84402 from ^/trunk ------------------------------------------------------------------------ r84402 | jhunold | 2013-05-21 14:23:20 +0200 (Di, 21 Mai 2013) | 4 lines Fix: cache file loading and saving must use native paths Fixes caching at least on Windows platform ------------------------------------------------------------------------ --- Author: johnmaddock Date: 2013-11-01 05:32:55 EDT (Fri, 01 Nov 2013) New Revision: 86532 URL: http://svn.boost.org/trac/boost/changeset/86532 Log: Merge fix for dependency tracking. --- Author: danieljames Date: 2013-11-03 14:20:09 EST (Sun, 03 Nov 2013) New Revision: 86551 URL: http://svn.boost.org/trac/boost/changeset/86551 Log: Merge Visual C++ 12 fix for Unordered. --- Author: viboes Date: 2013-11-03 15:43:42 EST (Sun, 03 Nov 2013) New Revision: 86552 URL: http://svn.boost.org/trac/boost/changeset/86552 Log: Thread: merge [86540] to fix #9319.

Show replies by date

Tom Kent

3 Nov 3 Nov

10:25 p.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

On Sun, Nov 3, 2013 at 3:15 PM, Marshall Clow wrote:

...

RC2 files for 1.55.0 are available at http://boost.cowic.de/rc/

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.

This helps ensure the candidates build OK before we push them out to SourceForge.

The files (and associated md5s) are: 93780777cfbf999a600f62883bd54b17 *boost_1_55_0_rc2.tar.gz d6eef4b4cacb2183f2bf265a5a03a354 *boost_1_55_0_rc2.tar.bz2 8aca361a4713a1f491b0a5e33fee0f1f *boost_1_55_0_rc2.zip 4e5bbc15fc8c80df8be428f8a5b5a823 *boost_1_55_0_rc2.7z

Thanks!

-- The release managers

When is the release expected? (How long do we have to try building?) Tom

Marshall Clow

11:05 p.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

On Nov 3, 2013, at 2:25 PM, Tom Kent wrote:

...

On Sun, Nov 3, 2013 at 3:15 PM, Marshall Clow wrote: RC2 files for 1.55.0 are available at http://boost.cowic.de/rc/

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.

This helps ensure the candidates build OK before we push them out to SourceForge.

The files (and associated md5s) are: 93780777cfbf999a600f62883bd54b17 *boost_1_55_0_rc2.tar.gz d6eef4b4cacb2183f2bf265a5a03a354 *boost_1_55_0_rc2.tar.bz2 8aca361a4713a1f491b0a5e33fee0f1f *boost_1_55_0_rc2.zip 4e5bbc15fc8c80df8be428f8a5b5a823 *boost_1_55_0_rc2.7z

Thanks!

-- The release managers

When is the release expected? (How long do we have to try building?)

We try to ship an RC on Wednesday, then release on Monday. If there are no problems found in this RC, I expect we'll release on Friday or Saturday. -- Marshall Marshall Clow Idio Software mailto:mclow.lists@gmail.com A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait). -- Yu Suzuki

Brian Ravnsgaard Riis

4 Nov 4 Nov

9:28 a.m.

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, On 2013-11-03 22:15, Marshall Clow wrote:

...

RC2 files for 1.55.0 are available at http://boost.cowic.de/rc/

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.

This helps ensure the candidates build OK before we push them out to SourceForge.

The files (and associated md5s) are: 93780777cfbf999a600f62883bd54b17 *boost_1_55_0_rc2.tar.gz d6eef4b4cacb2183f2bf265a5a03a354 *boost_1_55_0_rc2.tar.bz2 8aca361a4713a1f491b0a5e33fee0f1f *boost_1_55_0_rc2.zip 4e5bbc15fc8c80df8be428f8a5b5a823 *boost_1_55_0_rc2.7z

Thanks!

-- The release managers

Built successfully on Linux Mint 15 with clang in c++11 mode (using libstdc++, not libc++). Regards, /Brian Ravnsgaard Riis -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/ iQEcBAEBAgAGBQJSd2jCAAoJEFES4N8QrEod4SwIAJ7TUhuqYaIl1Jf4WMc3+4tX mnlHA0srh+SQGuWWo3LTdBIo/lOCBHISD/9gv176AA98w9OffaYocw+9vyTbusxl pu/21RtwvxJnsyfWqZgDbJwAsgbbU8TZZMjhe1Gaxu89wJaiPcXrWllNIVoCe0TV CJkiBHlDi5N/N6IuCucrH27QjjVXfsQcj41VzTV5c92vD9PfZcBwzztOpGaNNMrW L65odzXhBtBjoLc3Pnim/W4xIECm4iB43XGeDD015gdiX+G9creO9jLBdLtVQo3n 9zI1tfGjLj7pzMUVZH4uoHI7cx99H7JYtURlfb1kJWloYjJZrqFN9rFFNIKIa2E= =qypE -----END PGP SIGNATURE-----

Frédéric Bron

5 Nov 5 Nov

11:45 a.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

Hi, When I build the library (1.55.0 rc2) on x86_64 linux using a gcc cross-compiler to target i686-w64-mingw32 or x86_64-w64-mingw32, I get this error in the 'log' library: gcc.compile.c++ bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/event_log_backend.o "i686-w64-mingw32-g++-4.8.2" -ftemplate-depth-128 -O3 -finline-functions -Wno-inline -Wall -mthreads -m32 -fno-strict-aliasing -ftemplate-depth-1024 -DBOOST_ALL_NO_LIB=1 -DBOOST_CHRONO_STATIC_LINK=1 -DBOOST_FILESYSTEM_STATIC_LINK=1 -DBOOST_LOG_BUILDING_THE_LIB=1 -DBOOST_LOG_USE_AVX2 -DBOOST_LOG_USE_SSSE3 -DBOOST_SPIRIT_USE_PHOENIX_V3=1 -DBOOST_SYSTEM_NO_DEPRECATED -DBOOST_SYSTEM_STATIC_LINK=1 -DBOOST_THREAD_BUILD_LIB=1 -DBOOST_THREAD_DONT_USE_CHRONO=1 -DBOOST_THREAD_USE_LIB=1 -DDATE_TIME_INLINE -DNDEBUG -I"." -I"bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi" -c -o "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/event_log_backend.o" "libs/log/src/event_log_backend.cpp" libs/log/src/event_log_backend.cpp:36:30: fatal error: simple_event_log.h: No such file or directory #include "simple_event_log.h" ^ compilation terminated. ...failed gcc.compile.c++ bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/event_log_backend.o... Here are my commands: $ CC="i686-w64-mingw32-gcc-4.8.2" CFLAGS="-O2 -DNDEBUG -ffloat-store" CXX="i686-w64-mingw32-g++-4.8.2" CXXFLAGS="-O2 -DNDEBUG -ffloat-store -std=c++98 -DBOOST_THREAD_USE_LIB" AR="i686-w64-mingw32-ar" AS="i686-w64-mingw32-as" RANLIB="i686-w64-mingw32-ranlib" ./bootstrap.sh --without-icu --prefix=/cm/shared/apps/soft/bal/usr/mingw32-4.8.2-c++98-release/boost --with-toolset=gcc $ ./b2 -d2 -j16 -sBZIP2_INCLUDE=/path/to/mingw32-4.8.2-c++98-release/bzip2/include -sBZIP2_LIBPATH=/path/to/mingw32-4.8.2-c++98-release/bzip2/lib -sZLIB_INCLUDE=/path/to/mingw32-4.8.2-c++98-release/zlib/include -sZLIB_LIBPATH=/path/to/mingw32-4.8.2-c++98-release/zlib/lib --layout=system --user-config=user-config.jam --without-mpi --without-python --without-context --without-coroutine link=static threading=multi runtime-link=shared target-os=windows threadapi=win32 address-model=32 variant=release install

Andrey Semashev

11:50 a.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

On Tue, Nov 5, 2013 at 3:45 PM, Frédéric Bron wrote:

...

Hi,

When I build the library (1.55.0 rc2) on x86_64 linux using a gcc cross-compiler to target i686-w64-mingw32 or x86_64-w64-mingw32, I get this error in the 'log' library:

gcc.compile.c++ bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/event_log_backend.o

"i686-w64-mingw32-g++-4.8.2" -ftemplate-depth-128 -O3 -finline-functions -Wno-inline -Wall -mthreads -m32 -fno-strict-aliasing -ftemplate-depth-1024 -DBOOST_ALL_NO_LIB=1 -DBOOST_CHRONO_STATIC_LINK=1 -DBOOST_FILESYSTEM_STATIC_LINK=1 -DBOOST_LOG_BUILDING_THE_LIB=1 -DBOOST_LOG_USE_AVX2 -DBOOST_LOG_USE_SSSE3 -DBOOST_SPIRIT_USE_PHOENIX_V3=1 -DBOOST_SYSTEM_NO_DEPRECATED -DBOOST_SYSTEM_STATIC_LINK=1 -DBOOST_THREAD_BUILD_LIB=1 -DBOOST_THREAD_DONT_USE_CHRONO=1 -DBOOST_THREAD_USE_LIB=1 -DDATE_TIME_INLINE -DNDEBUG -I"." -I"bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi" -c -o "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/event_log_backend.o" "libs/log/src/event_log_backend.cpp"

libs/log/src/event_log_backend.cpp:36:30: fatal error: simple_event_log.h: No such file or directory #include "simple_event_log.h" ^ compilation terminated. ...failed gcc.compile.c++ bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/event_log_backend.o...

Here are my commands: $ CC="i686-w64-mingw32-gcc-4.8.2" CFLAGS="-O2 -DNDEBUG -ffloat-store" CXX="i686-w64-mingw32-g++-4.8.2" CXXFLAGS="-O2 -DNDEBUG -ffloat-store -std=c++98 -DBOOST_THREAD_USE_LIB" AR="i686-w64-mingw32-ar" AS="i686-w64-mingw32-as" RANLIB="i686-w64-mingw32-ranlib" ./bootstrap.sh --without-icu --prefix=/cm/shared/apps/soft/bal/usr/mingw32-4.8.2-c++98-release/boost --with-toolset=gcc $ ./b2 -d2 -j16 -sBZIP2_INCLUDE=/path/to/mingw32-4.8.2-c++98-release/bzip2/include -sBZIP2_LIBPATH=/path/to/mingw32-4.8.2-c++98-release/bzip2/lib -sZLIB_INCLUDE=/path/to/mingw32-4.8.2-c++98-release/zlib/include -sZLIB_LIBPATH=/path/to/mingw32-4.8.2-c++98-release/zlib/lib --layout=system --user-config=user-config.jam --without-mpi --without-python --without-context --without-coroutine link=static threading=multi runtime-link=shared target-os=windows threadapi=win32 address-model=32 variant=release install

Do you have message compiler (mc.exe) installed? Is it found by bjam (see the beginning of the build log)? Are there errors when compiling the mc file in the build log?

Frédéric Bron

11:58 a.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

...

...
When I build the library (1.55.0 rc2) on x86_64 linux using a gcc cross-compiler to target i686-w64-mingw32 or x86_64-w64-mingw32, I get this error in the 'log' library:

libs/log/src/event_log_backend.cpp:36:30: fatal error: simple_event_log.h: No such file or directory

Do you have message compiler (mc.exe) installed? Is it found by bjam (see the beginning of the build log)? Are there errors when compiling the mc file in the build log?

good point: it tries to call mc but that calls midnight commander on linux (see below). What should I do? Frédéric mc.compile bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/simple_event_log.h mc -h "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi" -r "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi" "libs/log/src/simple_event_log.mc" Usage: mc [OPTION...] [+number] [this_dir] [other_panel_dir] +number - Set initial line number for the internal editor GNU Midnight Commander 4.7.0.2 Help Options: -h, --help Show help options --help-all Show all help options --help-terminal Terminal options --help-color Color options ... compile.resource.null bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/simple_event_log_res.o as /dev/null -o "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/simple_event_log_res.o"

Andrey Semashev

12:03 p.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

On Tue, Nov 5, 2013 at 3:58 PM, Frédéric Bron wrote:

...

...
...
When I build the library (1.55.0 rc2) on x86_64 linux using a gcc cross-compiler to target i686-w64-mingw32 or x86_64-w64-mingw32, I get this error in the 'log' library:

libs/log/src/event_log_backend.cpp:36:30: fatal error: simple_event_log.h: No such file or directory

Do you have message compiler (mc.exe) installed? Is it found by bjam (see the beginning of the build log)? Are there errors when compiling the mc file in the build log?

good point: it tries to call mc but that calls midnight commander on linux (see below). What should I do?

Frédéric

mc.compile bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/simple_event_log.h

mc -h "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi" -r "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi" "libs/log/src/simple_event_log.mc"

Usage: mc [OPTION...] [+number] [this_dir] [other_panel_dir] +number - Set initial line number for the internal editor

GNU Midnight Commander 4.7.0.2

Help Options: -h, --help Show help options --help-all Show all help options --help-terminal Terminal options --help-color Color options ...

compile.resource.null bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/simple_event_log_res.o

as /dev/null -o "bin.v2/libs/log/build/gcc-mingw-4.8.2/release/address-model-32/link-static/log-api-winnt/target-os-windows/threadapi-win32/threading-multi/simple_event_log_res.o"

I suppose, the only solution right now is to disable event log support. Add define=BOOST_LOG_WITHOUT_EVENT_LOG to b2 command line and rebuild. I'll probably need to make the config check more robust so that it checks that mc generates the needed files. I thought b2 would do this automatically.

Frédéric Bron

12:18 p.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

...

I suppose, the only solution right now is to disable event log support. Add define=BOOST_LOG_WITHOUT_EVENT_LOG to b2 command line and rebuild.

This works fine. I see now that it is in the doc. However automatically disabling this feature if unsupported would be great. Thanks, Frédéric

Frédéric Bron

12:23 p.m.

New subject: [Boost-users] [1.55.0] Release candidate 2 available

1.55.0 rc2 seems to build correctly but like others said there is a strange message just after typing ./b2 command: link.jam: No such file or directory building on linux x86_64 targetting linux x86_64 or i686-w64-mingw32 or x86_64-w64-mingw32 (same error message). Frédéric

Jeremiah Willcock

10:05 p.m.

On Sun, 3 Nov 2013, Marshall Clow wrote:

...

RC2 files for 1.55.0 are available at http://boost.cowic.de/rc/

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.

This helps ensure the candidates build OK before we push them out to SourceForge.

The files (and associated md5s) are: 93780777cfbf999a600f62883bd54b17 *boost_1_55_0_rc2.tar.gz d6eef4b4cacb2183f2bf265a5a03a354 *boost_1_55_0_rc2.tar.bz2 8aca361a4713a1f491b0a5e33fee0f1f *boost_1_55_0_rc2.zip 4e5bbc15fc8c80df8be428f8a5b5a823 *boost_1_55_0_rc2.7z

Thanks!

I tried building in C++03 mode on the 20131013 snapshot of GCC 4.9 with Boost.MPI enabled on Linux (RHEL 6.4), and still had some errors (and many, many warnings): 1. There is still the message about link.jam being missing that I posted about with RC1. 2. I get a number of failures; a few are due to warnings-as-errors with more warnings enabled in 4.9, but many seem to be real problems. 3. qi_match_manip1 seems to enter an infinite loop intermittently (but most of the time) and use a large amount of memory. A partial test log (just of the failures up until the infinite loop in #3) is at URL:http://www.crest.iu.edu/~jewillco/gcc-49-build-errors.txt. I used the -w option to try to hide most of the warnings, and this was run after some other builds so it does not show the parts of the test suite that worked. -- Jeremiah Willcock

Steven Watanabe

6 Nov 6 Nov

4:36 p.m.

AMDG On 11/05/2013 02:05 PM, Jeremiah Willcock wrote:

...

1. There is still the message about link.jam being missing that I posted about with RC1.

This is harmless. Someone got overly excited about merging the changes for git. In Christ, Steven Watanabe

Tobias Loew

9:43 a.m.

There are still compilation errors on msvc-12.0 when using Boost.Container. Could somebody familiar with Boost.Container/Boost.Intrusive have a look at Ticket #9332, please. thanks -- View this message in context: http://boost.2283326.n4.nabble.com/1-55-0-Release-candidate-2-available-tp46... Sent from the Boost - Dev mailing list archive at Nabble.com.

Ion Gaztañaga

9:23 p.m.

El 06/11/2013 10:43, Tobias Loew escribió:

...

There are still compilation errors on msvc-12.0 when using Boost.Container. Could somebody familiar with Boost.Container/Boost.Intrusive have a look at Ticket #9332, please. thanks

I don't think we support that compiler for Boost 1.55. It was released 15 days ago and it might require lot of changes depending on compiler limitations/bugs. It's too risky IMHO. Ion

Antony Polukhin

7 Nov 7 Nov

7:40 a.m.

2013/11/7 Ion Gaztañaga

...

El 06/11/2013 10:43, Tobias Loew escribió:

There are still compilation errors on msvc-12.0 when using

...
Boost.Container. Could somebody familiar with Boost.Container/Boost.Intrusive have a look at Ticket #9332, please. thanks

I don't think we support that compiler for Boost 1.55. It was released 15 days ago and it might require lot of changes depending on compiler limitations/bugs. It's too risky IMHO.

+1 IMO, fussily adding support for suddenly released (and untested during development stage) compilers is not what must be done during the final release stage. May be a better solution would be to release a 1.56 sooner with less new features but with fixes for msvc-12.0? -- Best regards, Antony Polukhin

Tobias Loew

7:58 a.m.

...

May be a better solution would be to release a 1.56 sooner with less new features but with fixes for msvc-12.0?

+1 (I'm keen on switching to msvc-12 since it has significantly improved on the new C++11 features) -- View this message in context: http://boost.2283326.n4.nabble.com/1-55-0-Release-candidate-2-available-tp46... Sent from the Boost - Dev mailing list archive at Nabble.com.

Sergey Cheban

9:42 a.m.

On 07.11.2013 11:40, Antony Polukhin wrote:

...

+1 IMO, fussily adding support for suddenly released (and untested during development stage) compilers is not what must be done during the final release stage. I would propose fixing the compile-time bugs with VC12. Then, users will be able to test the 1.55.0 with their code and find the runtime bugs.

...

May be a better solution would be to release a 1.56 sooner with less new features but with fixes for msvc-12.0?

If there was no git migration, I would suggest releasing 1.55.1 with "experimental" VC12 support. -- Sergey Cheban

4021

Age (days ago)

4025

Last active (days ago)

List overview

Download

16 comments

11 participants

participants (11)

Andrey Semashev
Antony Polukhin
Brian Ravnsgaard Riis
Frédéric Bron
Ion Gaztañaga
Jeremiah Willcock
Marshall Clow
Sergey Cheban
Steven Watanabe
Tobias Loew
Tom Kent