Boost release 1.40.1 beta 1 is available from SourceForge: http://sourceforge.net/projects/boost/files/ This release contains one new libraries and numerous bug fixes for existing libraries. For details, see http://beta.boost.org/users/news/version_1_41_0for details. To install the beta on your system, see http://beta.boost.org/doc/libs/1_40_0/more/getting_started/index.html Please report any problems with the beta on the developers or users mailing list. Unless unexpected problems arise, the plan is to ship the final 1.40.0 release in a week to ten days. Thanks, --The Boost release team
Beman Dawes wrote:
Boost release 1.40.1 beta 1 is available from SourceForge:
http://sourceforge.net/projects/boost/files/
This release contains one new libraries and numerous bug fixes for existing libraries. For details, see http://beta.boost.org/users/news/version_1_41_0for details.
To install the beta on your system, see http://beta.boost.org/doc/libs/1_40_0/more/getting_started/index.html
Please report any problems with the beta on the developers or users mailing list.
Unless unexpected problems arise, the plan is to ship the final 1.40.0 release in a week to ten days.
Thanks,
You mean 1.41.0, not 1.40.1 or 1.40.0, right? :)
On Wed, Nov 4, 2009 at 8:31 AM, Andrey Semashev <andrey.semashev@gmail.com>wrote:
Beman Dawes wrote:
Boost release 1.40.1 beta 1 is available from SourceForge:
http://sourceforge.net/projects/boost/files/
This release contains one new libraries and numerous bug fixes for existing libraries. For details, see http://beta.boost.org/users/news/version_1_41_0for details.
To install the beta on your system, see http://beta.boost.org/doc/libs/1_40_0/more/getting_started/index.html
Please report any problems with the beta on the developers or users mailing list.
Unless unexpected problems arise, the plan is to ship the final 1.40.0 release in a week to ten days.
Thanks,
You mean 1.41.0, not 1.40.1 or 1.40.0, right? :)
Yep:-) --Beman
Beman Dawes <bdawes <at> acm.org> writes:
On Wed, Nov 4, 2009 at 8:31 AM, Andrey Semashev <andrey.semashev <at> gmail.com>wrote:
Beman Dawes wrote:
Boost release 1.40.1 beta 1 is available from SourceForge: You mean 1.41.0, not 1.40.1 or 1.40.0, right? :)
Yep
Numbering confusions extend to the names of the package files themselves, I think: boost_1_41_1_beta1.7z boost_1_41_1_beta1.zip boost_1_41_1_beta1.tar.bz2 boost_1_41_1_beta1.tar.gz Joaquín M López Muñoz Telefónica, Investigación y Desarrollo
On Wed, Nov 4, 2009 at 9:17 AM, Joaquin M Lopez Munoz <joaquin@tid.es>wrote:
Beman Dawes <bdawes <at> acm.org> writes:
On Wed, Nov 4, 2009 at 8:31 AM, Andrey Semashev <andrey.semashev <at> gmail.com>wrote:
Beman Dawes wrote:
Boost release 1.40.1 beta 1 is available from SourceForge: You mean 1.41.0, not 1.40.1 or 1.40.0, right? :)
Yep
Numbering confusions extend to the names of the package files themselves, I think:
boost_1_41_1_beta1.7z boost_1_41_1_beta1.zip boost_1_41_1_beta1.tar.bz2 boost_1_41_1_beta1.tar.gz
Ouch! Renamed via the SourceForge file manager. It will take an indeterminate length of time before the change is reflected in their release system. Thanks for catching this, --Beman
Beman Dawes wrote:
On Wed, Nov 4, 2009 at 9:17 AM, Joaquin M Lopez Munoz <joaquin@tid.es>wrote:
Beman Dawes <bdawes <at> acm.org> writes:
On Wed, Nov 4, 2009 at 8:31 AM, Andrey Semashev <andrey.semashev <at> gmail.com>wrote:
Beman Dawes wrote:
Boost release 1.40.1 beta 1 is available from SourceForge: You mean 1.41.0, not 1.40.1 or 1.40.0, right? :)
Yep
Numbering confusions extend to the names of the package files themselves, I think:
boost_1_41_1_beta1.7z boost_1_41_1_beta1.zip boost_1_41_1_beta1.tar.bz2 boost_1_41_1_beta1.tar.gz
Ouch! Renamed via the SourceForge file manager. It will take an indeterminate length of time before the change is reflected in their release system.
It seems that file release system is the only feature of SourceForge that we use, and that it's still kinda painful to use. Maybe, we should move? E.g. serving files from the web server or using Google Code (which has scripted upload) or something else? - Volodya
On Wed, Nov 4, 2009 at 10:05 AM, Vladimir Prus <vladimir@codesourcery.com>wrote:
Beman Dawes wrote:
Ouch! Renamed via the SourceForge file manager. It will take an indeterminate length of time before the change is reflected in their release system.
It seems that file release system is the only feature of SourceForge that we use, and that it's still kinda painful to use. Maybe, we should move? E.g. serving files from the web server or using Google Code (which has scripted upload) or something else?
SourceForge has improved their release system, and it is quite a bit easier to use than in the past. I've been updating http://svn.boost.org/trac/boost/wiki/ReleasePractices/ManagerCheckList#Relea... I become familiar with the new SourceForge interface. The name goofs were strictly on my side. The names were wrong when I created the release candidate; see http://boost.cowic.de/rc/ Thanks for the comment, however, since we always need to be on the lookout for ways to improve our process. Speaking of that, I notice that SourceForge has a Thumbs up/Thumbs down rating system now. We need to take a look at the roughly 10% of the Boost ratings that get a "Thumbs down" to see what is bothering users. --Beman
AMDG Beman Dawes wrote:
Speaking of that, I notice that SourceForge has a Thumbs up/Thumbs down rating system now.
We need to take a look at the roughly 10% of the Boost ratings that get a "Thumbs down" to see what is bothering users.
10% = 1 out of 11. I don't think this one is very useful to us: "**Meta programming taken a few steps too far. There are some very usefull libraries here, but the readability of the code is awfull." In Christ, Steven Watanabe
On Wed, Nov 4, 2009 at 10:59 AM, Steven Watanabe <watanabesj@gmail.com>wrote:
AMDG
Beman Dawes wrote:
Speaking of that, I notice that SourceForge has a Thumbs up/Thumbs down rating system now.
We need to take a look at the roughly 10% of the Boost ratings that get a "Thumbs down" to see what is bothering users.
10% = 1 out of 11. I don't think this one is very useful to us:
"**Meta programming taken a few steps too far. There are some very usefull libraries here, but the readability of the code is awfull."
Oh, well. Perhaps we should include a spelling guide, starting with "usefull" and "awfull":-) Thanks for checking, --Beman
Steven Watanabe wrote:
AMDG
Beman Dawes wrote:
Speaking of that, I notice that SourceForge has a Thumbs up/Thumbs down rating system now.
We need to take a look at the roughly 10% of the Boost ratings that get a "Thumbs down" to see what is bothering users.
10% = 1 out of 11. I don't think this one is very useful to us:
"**Meta programming taken a few steps too far. There are some very usefull libraries here, but the readability of the code is awfull."
The spelling of the comment is awful also.
Beman Dawes wrote:
Speaking of that, I notice that SourceForge has a Thumbs up/Thumbs down rating system now.
We need to take a look at the roughly 10% of the Boost ratings that get a "Thumbs down" to see what is bothering users.
I have always wanted a "rating system" on a library by libray basis. Ideally, this would be a web page that could be invoked for any library library: select library Rate 1 (unacceptable) ... 5(great) the following: functionality/utility(trivial)..(indispensable) robustness(flakey),..,(non percieved) percieved performance(slow),...(unperceptible) documentation (useless).(tax code)..(a pleasure to read and use) ease of use(tax code)...(no-brainer) .... response to questions/support(awol)...(right next to me) overall quality comments:.... and a summary page for each library total number of responses results for above questions overall rating. links to comments/pages/ar ratings. Robert Ramey
On Thu, Nov 5, 2009 at 12:47 PM, Robert Ramey <ramey@rrsd.com> wrote:
Beman Dawes wrote:
Speaking of that, I notice that SourceForge has a Thumbs up/Thumbs down rating system now.
We need to take a look at the roughly 10% of the Boost ratings that get a "Thumbs down" to see what is bothering users.
I have always wanted a "rating system" on a library by libray basis. Ideally, this would be a web page that could be invoked for any library
...
Interesting idea. Do you know of any easy yet reliable way to set up such a system? Are there any free web sites that do this sort of thing? --Beman
Beman Dawes wrote:
Boost release 1.40.1 beta 1 is available from SourceForge:
http://sourceforge.net/projects/boost/files/
This release contains one new libraries and numerous bug fixes for existing libraries. For details, see http://beta.boost.org/users/news/version_1_41_0for details.
To install the beta on your system, see http://beta.boost.org/doc/libs/1_40_0/more/getting_started/index.html
Please report any problems with the beta on the developers or users mailing list.
For some reason, the link to Spirit still points to "classic", not 2.1. In my local release directory tree, the redirect for "libs/spirit/index.html" works as expected. Regards, -- Joel de Guzman http://www.boostpro.com http://spirit.sf.net http://www.facebook.com/djowel Meet me at BoostCon http://www.boostcon.com/home http://www.facebook.com/boostcon
On Wed, Nov 4, 2009 at 1:58 PM, Joel de Guzman <joel@boost-consulting.com> wrote:
For some reason, the link to Spirit still points to "classic", not 2.1. In my local release directory tree, the redirect for "libs/spirit/index.html" works as expected.
That's because it's the 1.40 documentation on the site, when the 1.41 documentation is uploaded it'll work okay. Daniel
Beman Dawes <bdawes@acm.org> writes:
http://beta.boost.org/users/news/version_1_41_0 for details.
This page doesn't include the updates to boost.thread. Can you remind me which file I need to modify to fix that? Also, I've made some doc changes on trunk. Is it OK to merge them to release? Anthony -- Author of C++ Concurrency in Action | http://www.manning.com/williams just::thread C++0x thread library | http://www.stdthread.co.uk Just Software Solutions Ltd | http://www.justsoftwaresolutions.co.uk 15 Carrallack Mews, St Just, Cornwall, TR19 7UL, UK. Company No. 5478976
I'm getting errors building signals with MSVC10 beta 2. Build errors follow: boost/iterator/iterator_facade.hpp(528) : error C2039: 'decrement' : is not a member of 'boost::signals::detail::named_slot_map_iterator' boost/signals/detail/named_slot_map.hpp(81) : see declaration of 'boost::signals::detail::named_slot_map_iterator' boost/iterator/iterator_facade.hpp(681) : see reference to function template instantiation 'void boost::iterator_core_access::decrement<Derived>(Facade &)' being compiled with [ Derived=boost::signals::detail::named_slot_map_iterator, Facade=boost::signals::detail::named_slot_map_iterator ] boost/iterator/iterator_facade.hpp(680) : while compiling class template member function 'boost::signals::detail::named_slot_map_iterator &boost::iterator_facade<Derived,Value,CategoryOrTraversal>::operator --(void)' with [ Derived=boost::signals::detail::named_slot_map_iterator, Value=boost::signals::detail::connection_slot_pair, CategoryOrTraversal=boost::forward_traversal_tag ] boost/signals/detail/named_slot_map.hpp(85) : see reference to class template instantiation 'boost::iterator_facade<Derived,Value,CategoryOrTraversal>' being compiled with [ Derived=boost::signals::detail::named_slot_map_iterator, Value=boost::signals::detail::connection_slot_pair, CategoryOrTraversal=boost::forward_traversal_tag ] boost/iterator/iterator_facade.hpp(546) : error C2039: 'advance' : is not a member of 'boost::signals::detail::named_slot_map_iterator' boost/signals/detail/named_slot_map.hpp(81) : see declaration of 'boost::signals::detail::named_slot_map_iterator' boost/iterator/iterator_facade.hpp(694) : see reference to function template instantiation 'void boost::iterator_core_access::advance<Derived>(Facade &,__w64 int)' being compiled with [ Derived=boost::signals::detail::named_slot_map_iterator, Facade=boost::signals::detail::named_slot_map_iterator ] boost/iterator/iterator_facade.hpp(693) : while compiling class template member function 'boost::signals::detail::named_slot_map_iterator &boost::iterator_facade<Derived,Value,CategoryOrTraversal>::operator +=(__w64 int)' with [ Derived=boost::signals::detail::named_slot_map_iterator, Value=boost::signals::detail::connection_slot_pair, CategoryOrTraversal=boost::forward_traversal_tag ]
I'm getting errors building signals with MSVC10 beta 2. Build errors follow:
boost/iterator/iterator_facade.hpp(528) : error C2039: 'decrement' : is not a member of 'boost::signals::detail::named_slot_map_iterator' boost/signals/detail/named_slot_map.hpp(81) : see declaration of 'boost::signals::detail::named_slot_map_iterator'
boost/iterator/iterator_facade.hpp(546) : error C2039: 'advance' : is not a member of 'boost::signals::detail::named_slot_map_iterator' boost/signals/detail/named_slot_map.hpp(81) : see declaration of 'boost::signals::detail::named_slot_map_iterator'
These two errors have been fixed in trunk already. Merging them to the release branch depends on whether we strive to 'support' the beta of this compiler. What's the release managers' verdict? Regards Hartmut ------------------- Meet me at BoostCon http://boostcon.com
Hartmut Kaiser wrote:
I'm getting errors building signals with MSVC10 beta 2. Build errors follow: <snip>
These two errors have been fixed in trunk already.
Merging them to the release branch depends on whether we strive to 'support' the beta of this compiler. What's the release managers' verdict?
Hi Hartmut, If you feel this is a low-risk fix and the tests have cycled on trunk and the results look good, please go ahead and merge to release. -- Eric Niebler BoostPro Computing http://www.boostpro.com
Hartmut Kaiser wrote:
I'm getting errors building signals with MSVC10 beta 2. Build errors follow: <snip>
These two errors have been fixed in trunk already.
Merging them to the release branch depends on whether we strive to 'support' the beta of this compiler. What's the release managers' verdict?
Hi Hartmut,
If you feel this is a low-risk fix and the tests have cycled on trunk and the results look good, please go ahead and merge to release.
Ok, done. Regards Hartmut ------------------- Meet me at BoostCon http://boostcon.com
With MSVC10 beta 2 having added nullptr support, boost/config/compiler/visualc.hpp should probably be updated to #define BOOST_NO_NULLPTR conditionally similarly to BOOST_NO_RVALUE_REFERENCES.
Folks, Boost 1.40.0 beta 1 fail to build on SPARC64 machines. 1.37, however, builds flawlessly. Brief analysis indicates that the issue arises when code of smart_ptr is assembled. The "boost/smart_ptr/detail/sp_counted_base_gcc_sparc.hpp" file contains SPARC-specific assembly for atomic compare-and-swap implementation and assembler rejects the code generated by the compiler out of that function. The generated code contains a "cas" instruction where first operand is a memory reference with complex addressing (register and offset). According to the SPARC v9 manual only register containing stored address may go there. Environment: "uname -a": FreeBSD boost 8.0-BETA4 FreeBSD 8.0-BETA4 #4: Tue Sep 15 14:03:28 UTC 2009 boris@poly-e450-1:/usr/obj/usr/src/sys/E450 sparc64 "c++ --version": c++ (GCC) 4.2.1 20070719 [FreeBSD] "as --version": GNU assembler 2.15 [FreeBSD] 2004-05-23 The attached patch allows the code to build. What is the appropriate way to check that the patched code works correctly both in terms of logic and atomicity? Alexander Churanov maintainer of devel/boost-* ports for FreeBSD
The attached is a patch that seems to be more appropriate. However, it is not yet tested.
Alexander Churanov wrote:
Folks,
Boost 1.40.0 beta 1 fail to build on SPARC64 machines. 1.37, however, builds flawlessly.
Brief analysis indicates that the issue arises when code of smart_ptr is assembled.
...
The attached patch allows the code to build.
Doesn't seem right to me. How about - __asm__ __volatile__( "cas %0, %2, %1" - : "+m" (*dest_), "+r" (swap_) - : "r" (compare_) - : "memory" ); + __asm__ __volatile__( "cas [%1], %2, %0" + : "+r" (swap_) + : "r" (dest_), "r" (compare_) + : "memory" ); ?
What is the appropriate way to check that the patched code works correctly both in terms of logic and atomicity?
You should run the smart_ptr tests, including shared_ptr_mt_test and weak_ptr_mt_test, on a multicore machine.
Peter, Your solution looks like the best one. It's close to the second patch I've send in a separate e-mail. Thank you for instructions I'll try to run the tests. Will boost-1.40.0 release contain the fix? Currently I am going to add the patch to FreeBSD ports collection. I guess other operating systems are still affected in case they are using the same compiler and assembler. Alexander Churanov
Alexander Churanov wrote:
Peter,
Your solution looks like the best one. It's close to the second patch I've send in a separate e-mail. Thank you for instructions I'll try to run the tests.
Will boost-1.40.0 release contain the fix?
This is for the release managers to decide, but I would guess not.
On Fri, Nov 13, 2009 at 2:58 PM, Peter Dimov <pdimov@pdimov.com> wrote:
Alexander Churanov wrote:
Peter,
Your solution looks like the best one. It's close to the second patch I've send in a separate e-mail. Thank you for instructions I'll try to run the tests.
Will boost-1.40.0 release contain the fix?
This is for the release managers to decide, but I would guess not.
Sorry, too late. Assuming not problems surface in tonight's tests, the plan is to start building the release first thing in the morning. --Beman
2009/11/14 Beman Dawes <bdawes@acm.org>
Will boost-1.40.0 release contain the fix? Sorry, too late. Assuming not problems surface in tonight's tests, the
Alexander Churanov wrote: plan is to start building the release first thing in the morning.
No problems, I'll just keep file with the patch for the issue in ports, and users will get fixed Boost when installing on FreeBSD. Alexander Churanov
Alexander Churanov wrote:
2009/11/14 Beman Dawes <bdawes@acm.org>
Will boost-1.40.0 release contain the fix? Sorry, too late. Assuming not problems surface in tonight's tests,
Alexander Churanov wrote: the plan is to start building the release first thing in the morning.
No problems, I'll just keep file with the patch for the issue in ports, and users will get fixed Boost when installing on FreeBSD.
Can you confirm that the tests work with the change?
Beman, Is is correct to assume that there is no changes between beta1 and the 1.41.0 release? If so, it will be not necessary to re-test all applications that depend on boost when release is published. Usually Boost ports were prepared after the release is published and the question had not exist. This time I am going to publish a port for FreeBSD with minimal delay. Even re-building all applications that depend on Boost may take days and I'd like to save some time and efforts. Alexander Churanov
On Mon, Nov 16, 2009 at 7:47 AM, Alexander Churanov <alexanderchuranov@gmail.com> wrote:
Beman,
Is is correct to assume that there is no changes between beta1 and the 1.41.0 release?
Actually, there were quite a lot of changes. Some bug fixes + lots changes to clear compiler warnings. You can run an svn diff --summarize on the two revisions to see what files changed. --Beman
2009/11/16 Beman Dawes <bdawes@acm.org>
Actually, there were quite a lot of changes. Some bug fixes + lots changes to clear compiler warnings.
OK, thank you for information.
You can run an svn diff --summarize on the two revisions to see what files changed.
Do you mean the '/tags/release/Boost_1_41_0_beta1' and '/tags/release/Boost_1_41_0' directories? Alexander Churanov
participants (15)
-
Adam Merz -
Alexander Churanov -
Andrey Semashev -
Anthony Williams -
Beman Dawes -
Daniel James -
Edward Diener -
Eric Niebler -
Hartmut Kaiser -
Joaquin M Lopez Munoz -
Joel de Guzman -
Peter Dimov -
Robert Ramey -
Steven Watanabe -
Vladimir Prus