December 2012 - Boost - lists.preview.boost.org

Interest in serialization library
by iwg molw5 28 Dec '12

28 Dec '12

I recently wrote a small C++11 template library to unify the specification and construction of a serializable object. The library introduces a simple template markup language used to specify the layout of data on the wire, using that specification to construct the type in question much like a named tuple. A sample of the syntax used by the library is provided below: struct Object : serializable <Object, value <NAME("Field 1"), little_endian <uint32_t>>, value <NAME("Field 2"), big_endian <uint32_t>>> { }; The library is available under the MIT license at: http://www.github.com/molw5/framework with associated documentation located at: http://molw5.github.com/framework Is there any interest in including this library in Boost?

3 10

boost-commit list down?
by Marshall Clow 28 Dec '12

28 Dec '12

The last message I got from boost-commits was: Author: apolukhin Date: 2012-12-27 13:54:28 EST (Thu, 27 Dec 2012) New Revision: 82224 URL: http://svn.boost.org/trac/boost/changeset/82224 The archives at http://lists.boost.org/boost-commit/2012/12/date.php only go through revision 82223. -- Marshall

1 0

[odeint] Typo in Boost.Odeint
by Michel Morin 28 Dec '12

28 Dec '12

Boost.Odeint contains nearly 200 typos! Patch attached in #7835 https://svn.boost.org/trac/boost/ticket/7835 (I don't know why but the owner of odeint is empty in the ticket.) Regards, Michel

2 1

Re: [boost] [GitHelp] Sandbox library to GitHub
by Brian Wood 28 Dec '12

28 Dec '12

Edward Diener: > >On 12/26/2012 5:10 AM, Daniel Pfeifer wrote: >> 2012/12/26 Andrey Semashev <andrey.semashev(a)gmail.com>: >>> >>> Perhaps you didn't set up a ssh public key? >>> >>> https://help.github.com/articles/generating-ssh-keys#platform-windows >>> >>> You will also have to add the key to TortoiseGit/PuTTY. >> >> If you push over HTTPS, you don't need an ssh key. > > That sounds like what I want to do. The whole low-level business of > generating public/private keys, when I already have signed up for GitHub > with a username/password, seems silly to me. I know the Linux gurus all > love this kind of intricate stuff, but I do not. It's one command: ssh-keygen -t rsa -C "username(a)email.net" If you have Cygwin installed on Windows you can use the same command there. -- Brian Ebenezer Enterprises -- in G-d we trust. http://webEbenezer.net (651) 251-9384

2 1

[thread] to std:: or not to std::
by Michael Caisse 27 Dec '12

27 Dec '12

There seems to be a trend to migrate boost::thread to std::thread. Has there been consideration that perhaps the user community may not want this? There are differing opinions on threads (obviously) and recent breaking changes concern me. For example, the thread destructor now calling terminate instead of detach has made the upgrade to 1.52 more troublesome than it should have been. Will the next move be to remove interruption points? We have purposely selected Boost.Thread on a few projects that are also using C++11 because we prefer the Boost.Thread behaviour and feature set. The purpose of this email isn't to debate the merit of the Boost.Thread feature set but instead to understand the future direction of the library. Can Vicente or someone provide the intended road map or general plan for the library? Thanks! michael -- Michael Caisse ciere consulting ciere.com

4 9

Patch for functional/hash
by Thomas Heller 27 Dec '12

27 Dec '12

Hi, Attached is a patch for functional/hash. The first chunk essentially gets rid of some UB (the same technique was already applied for the remainder so this looks like an oversight). I don't remember why exactly i added the second chunk, it helps on some compilers though (I think gcc on Android). Save to apply that patch? Regards, Thomas

2 2

[proposal] safe_ref<T>
by Fernando Pelliccioni 27 Dec '12

27 Dec '12

Hi all, The motivation is to avoid code like the following... //---------------------------------------------------------------------------- #include <iostream> #include <string> struct A { void do_something( std::string s ) const { std::cout << "do something with " << s << " and " << i_ << "\n"; } int i_ = 99; }; // unsafe traditional code struct Unsafe { void set_ref( A& a ) { a_ptr_ = std::addressof(a); } void use_ref( std::string s ) const { if (a_ptr_ != nullptr) { a_ptr_->do_something(s); } } void use_ref_2( std::string s ) const { //forgot to check for nullptr a_ptr_->do_something(s); } A* a_ptr_ = nullptr; //I can't use A& nor reference_wrapper<A>, need to be nullable. }; void f() { A a; Unsafe unsafe; unsafe.use_ref( "hello" ); //null reference, do nothing, hope! unsafe.set_ref(a); unsafe.use_ref( "hello world!" ); //not-null reference, Ok! Unsafe unsafe2; unsafe2.use_ref( "hello" ); //null reference, do nothing, hope! unsafe2.use_ref_2( "hello" ); //null reference, Ops! } //---------------------------------------------------------------------------- The proposed solution: //---------------------------------------------------------------------------- // Proposed (draft) //---------------------------------------------------------------------------- template <typename T> class safe_ref { public: typedef T type; explicit safe_ref() : t_( nullptr ) {} explicit safe_ref(T& t) : t_( std::addressof(t) ) {} template <typename Func> void use ( Func f ) const { if ( is_initialized() ) { f( *t_ ); } } private: bool is_initialized() const { return t_ != nullptr; } T* t_; }; template <typename T> inline safe_ref<T> const ref(T& t) { return safe_ref<T>(t); } template <typename T> inline safe_ref<T const> const cref(T const& t) { return safe_ref<T const>(t); } //---------------------------------------------------------------------------- // Usage //---------------------------------------------------------------------------- #include <iostream> #include <string> struct A { void do_something( std::string s ) const { std::cout << "A::do_something with " << s << " and " << i_ << "\n"; } int i_ = 99; }; struct Safe { void set_ref( A& a ) { a_ref_ = ref(a); } void use_ref( std::string s ) const { a_ref_.use( [&s] ( A& a ) { a.do_something(s); }); } safe_ref<A> a_ref_; }; void usage1() { A a; Safe safe; safe.use_ref( "hello" ); //null reference, do nothing safe.set_ref(a); safe.use_ref( "hello world!" ); //safe reference, Ok! } void usage2() { safe_ref<int> si; si.use( [] (int& i ) { std::cout << i << std::endl; }); int temp = 15; safe_ref<int> si2(temp); si2.use( [] (int& i ) { std::cout << i << std::endl; }); si = ref(temp); si.use( [] (int& i ) { std::cout << i << std::endl; }); } //---------------------------------------------------------------------------- It is similar to optional<T>, but in this case the idea is to not allow access to internal data if the pointer is null, only allowing a safe usage There is interest to add something like this to Boost? Any comments or remarks? Thanks and regards, Fernando Pelliccioni.

5 6

[GitHelp] Sandbox library to GitHub
by Edward Diener 27 Dec '12

27 Dec '12

Under Windows following the instructions at https://svn.boost.org/trac/boost/wiki/ExtractSandbox for moving an SVN sandbox library to GitHub I am successful in the first step of cloning SVN sandbox library to a temporary local directory. But when I go to that local directory and try: git push --mirror git@github.com:eldiener/variadic_macro_data.git I receive a PuTTY message of: "Disconnected: No supported authentication methods available (server sent: publickey)" I have no idea how PuTTY gets into this. I am using TortoiseGit and Git for Windows and evidently getting Git to work under Windows is difficult in some way. Anybody have any idea how to straighten this out ?

4 6

[git] What about unmaintained libraries?
by Andrey Semashev 26 Dec '12

26 Dec '12

Hi, I have little experience with git and none with GitHub, so apologies if my questions sound silly. We have a number of unmaintained libraries in Boost. With svn anyone with commit rights can apply fixes and even take over these libs. In the end of the day, I think, this is a good thing - community support is better than no support. Even with maintained libs, it is sometimes simpler to just apply the fix instead of filing a ticket and waiting for the maintainer to get to it. Also, remember the bug fixing weeks we had a while ago - anyone could choose a library and resolve tickets against it. I'm not sure how this will work after transition to git. IIUC, every lib will be extracted into a separate git repository, and presumably only the library maintainer will have the commit rights to that repo. Essentially this means that unmaintained libs will become immutable after the transition. Also, every minor fix will have to go through the library maintainer, however busy he may be. I suppose the problem with unmaintained libs could be solved by forking but what if noone is ready to become the official maintainer but occasionally people commit fixes and improvements to the lib? I believe, at least Boost.DateTime is currently in such situation. Also, forking does not solve the problem for minor fixes to maintained libs. I would really like it if there was some kind of a user group (of all Boost maintainers) that could be used to grant commit rights to the libraries. Whatever the git term is, by "commit" here I mean the ability to publish the changes to the repository on GitHub, so that they are visible to other developers, including monolithic Boost release scripts. Is this possible? If yes, I think all repositories produced as a result of the transition to git should grant commit rights to this group by default, so that the access rights are the same as with svn now.

4 7

[move] uninitialized_move
by Adam Wulkiewicz 26 Dec '12

26 Dec '12

Hi, I recently saw how boost::uninitialized_move is implemented, it doesn't check for exceptions thrown by move ctor and don't delete already constructed objects. Is this intentional? Regards, Adam

2 2

Boost SIMD beta release
by joel falcou 25 Dec '12

25 Dec '12

We're rolling a beta release of Boost.SIMD along with a beta release of NT2. This version comes very late because we were busy with other things but hopefully we'll now do rolling releases frequently. Documentation is quite lacking but code is relatively stable and works. * Release Files can be found here: http://nt2.metascale.org/downloads/ * Additional material about the library can be found here : - Boost.SIMD PACT paper : http://www.lri.fr/~falcou/pub/pact-2012.pdf - Boost.SIMD MeetingC++ slides : http://meetingcpp.com/tl_files/mcpp/slides/12/simd.pdf * We welcome comments, questions and random requests/rants/bug reports as to polish the library before actual submission to the boost review. Such comments can be sent there, on our nt2-dev Google groups or in the nt2 issue tracker. Regards

13 57

automatic SVN mirror down?
by Robin Eckert 24 Dec '12

24 Dec '12

Hello, The latest commit on the live boost SVN mirror (git://github.com/ryppl/boost-svn.git) is 11 days old. Is the automatic mirror jammed right now or is the mirror probably shut down due to the modularization effort? Regards, Robin

2 2

[GitHelp] Help desk ready to assist
by Beman Dawes 24 Dec '12

24 Dec '12

See https://svn.boost.org/trac/boost/wiki/GitHelp for details. --Beman

1 0

[Git] ** Attention All Boost Committers **
by Dave Abrahams 24 Dec '12

24 Dec '12

If you have *ever* checked *anything* (e.g. code or documentation) into the Boost Subversion repository, this message applies to you. Git uses email addresses to identify authors and committers. In order to accurately convert Subversion history to Git, we need to map all (or as many as possible) Subversion user IDs to working email addresses. I have posted a provisional UserID mapping table at https://github.com/boost-lib/boost/wiki/AuthorMapping (with a simple mangling of '@'=>'<>' and '.'=>'~'). If your email address is already in the table, you've been Bcc'd on this message, which may explain the appearance of multiple copies in your mailbox. I request that you make corrections to this table. To edit the table directly, you'll need to be logged into GitHub. Direct correction is by far the preferable approach, but if you don't have---and don't want to acquire---a free GitHub account, you can send edits directly to me. Important notes (these also appear on the page with the table) * You may have several userids (e.g. I appear to have "dave," "abrahams," and "david_abrahams" --- probably left over from the transition from SourceForge CVS to our own SVN). Please look carefully to be sure that you have accounted for all your userids, and please use the same name and email address for each of them. Please *make sure that all non-duplicate userids stay in the table*. For example, all three of my userids are still there. * A given userid may appear several times in the table. If that userid belongs to you, please delete all but one line. If you don't, we will just use the first entry with that userid. * If you have a GitHub account, please make sure the email address you use in the table is associated with that account at https://github.com/settings/emails Thanks for your cooperation. -- Dave Abrahams BoostPro Computing Software Development Training http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost

3 2

[TypeErasure] Similar simple class.
by Arman Hunanyan 23 Dec '12

23 Dec '12

Hi All, I have implemented xany class https://sourceforge.net/projects/extendableany/?source=directory. Then I noticed that there is a Boost.TypeErasure library which allows to do same as xany does and something more. Implementation of class xany has 1005 lines and it is implemented and tested in a week. Of course xany isn't comparable with Boost.TypeErasure, but it may be interesting to look. Below is a definition of same method for Boost.TypeErasure and for xany. // definition of push_back for Boost.TypeErasure template<class C, class T> struct push_back { static void apply(C& cont, const T& arg) { cont.push_back(arg); } }; namespace boost { namespace type_erasure { template<class C, class T, class Base> struct concept_interface< ::push_back<C, T>, Base, C> : Base { void push_back(typename rebind_any<Base, const T&>::type arg) { call(::push_back<C, T>(), *this, arg); } }; } } any<push_back<_self, int>, _self&> c; // definition of push_back for xany template <typename Xany, typename T> struct push_back_method { typedef void (Xany::* signature) (const T&); template <typename B> struct wrapper : public B { void push_back(const T& e) { this->call(push_back_method(), e); } }; struct implementation { template <typename Container> void operator() (Container& c, const T& e) { c.push_back(e); } }; }; xany<boost::mpl::vector<push_back_method<boost::mpl::_1, int> > > c; xany requires additional signature type to be specified. This allows to use dispatching in implementation. struct implementation { void operator() (OtherCodingStyleContainer& c, const T& e) { c.PushBack(e); } template <typename Container> void operator() (Container& c, const T& e) { c.push_back(e); } }; Documentation contains more examples.

1 0

[Inspect] hadn't run since... August
by Artyom Beilis 23 Dec '12

23 Dec '12

Boost.Inspect hadn't run on trunk or release sinse august: From: http://beta.boost.org/development/testing.html trunk - http://boost.cowic.de/rc/inspect-trunk.html release - http://boost.cowic.de/rc/inspect-snapshot.html Artyom Beilis -------------- CppCMS - C++ Web Framework: http://cppcms.com/ CppDB - C++ SQL Connectivity: http://cppcms.com/sql/cppdb/

2 1

[atomic, x86] need another pair of eyes
by Tim Blechmann 23 Dec '12

23 Dec '12

hi all, i need another pair of eyes regarding boost.atomic on x86: the implementation of memory barrier is merely a compiler barrier, but not a CPU barrier, as it is using code like: __asm__ __volatile__ ("" ::: "memory"); afaict, one should use a `real' CPU barrier like "mfence" or "lock; addl $0,0(%%esp)". is this correct? apart from that, i've seen that compare_exchange is using explicit memory barriers before/after "cmpxchg" instructions. i somehow though that cmpxchg and the 8b/16b variants implicitly issue a memory barrier, so the resulting code would generate multiple memory barriers. can someone with some insights in the x86 architecture confirm this? thanks, tim

6 35

Re: [boost] Boost SIMD beta release
by Brian Wood 23 Dec '12

23 Dec '12

From: Mathias Gaunard: On 20/12/2012 20:34, Dave Abrahams wrote: >> Yes; there's a runtime test. > > Let's put a runtime test at every step of our critical loop, sounds like > a good idea. > I actually somewhat looked into this, and in this case segmented > iterators don't allow any optimization compared to normal iterators. I think less sarcasm would be helpful. >> I'm afraid you have not been specific enough about the >> problem you claim to see for me to be able to evaluate >> whether it exists or not. > > I'm afraid you're the one trying to make everything fit into this > model you've somehow taken a liking to. I'd prefer to focus on results rather than motives. Shalom Brian Ebenezer Enterprises http://webEbenezer.net (651) 251-9384

1 0

[BTREE] Get BTREE library into boost - Market Research
by Joel Young 22 Dec '12

22 Dec '12

My organization is investigating contracting for enhancements to the boost::btree library developed by Beman Dawes with all enhancements to remain under the boost license and author(s) retaining all copyright. No travel is expected to be required as part of this project. We are considering the following enhancements: 1. Develop detailed developer and tutorial documentation in customary boost formats 2. Enhancing the interface to provide more explicit preload and caching control to include the amount of RAM to reserve, perhaps with the ability to specify posix alignment to maximize benefit of linux anonymous huge page support. 3. Make the library C++ 11 ready 4. Finalize data layout and support for variable length payloads 5. Provide convenient interface for storing std::string (or the contents thereof) in payloads 6. Prepare the library for submission for formal review Any comments and suggestions on this are greatly appreciated. Of particular interest are discussions of technical approaches and estimates of hours required. Please do not send formal proposals. In addition to replying on this list (preferred) I can be reached at the telephone numbers and email below. Comments sent by 21 December will be most valuable. Discussion after that is still very welcome! Please don’t include any proprietary data. Also note that I am not authorized to commit the government in any way--this is simply market research. Thanks! Joel Dr. JOEL D. YOUNG Assistant Professor Computer Science Department Naval Postgraduate School Glasgow East Rm 338 jdyoung(a)nps.edu (831) 656-3518

6 11

Security bugs policy
by Artyom Beilis 22 Dec '12

22 Dec '12

Hello, I would like to know if there is any policy about security bug fixes, updates or releases. What I'm talking about. Consider a security related bug discovered in boost library. For example: For example recently reported bug in Boost.Locale that will be fixed in 1.53 https://svn.boost.org/trac/boost/ticket/7743 I consider it security bug as an application that validates the input (checks if it is UTF-8) may accept an invalid sequence. Or very critical bug in UUID (fixed in 1.43): https://svn.boost.org/trac/boost/ticket/3971 That any application that used them may be target to UUID guessing. These kind of bugs in general should be considered critical and somehow published, it may be also important to backport the fixes to stable releases, because not everybody can move forward and they may need to stick with older releases (as Boost is neither ABI nor API backward compatible even between minor releases). Annoucements of such bugs may be critical for example for stable Linux distributions that ship some older Boost versions and likely need to backport such bug fixes to their own boost tree because Boost doesn't do it. Questions: - Do we have any policy about it? - If not we probably need one including backporting of security bug fixes to stable releases. What brings me to other issues... Should Boost support older releases with critical bug fixes/security bug fixes? For how long time? Artyom Beilis -------------- CppCMS - C++ Web Framework: http://cppcms.com/ CppDB - C++ SQL Connectivity: http://cppcms.com/sql/cppdb/

2 1

Question about test failures
by Andrey Semashev 22 Dec '12

22 Dec '12

Hi, I've been looking through Boost.Atomic test results [1] and saw that all tests on IBM_Canada_Ltd are marked as failed. The test result [2] refer to Boost.Thread build results [3] which I think is supposed to mean that the test failed because the dependent library failed to build. However, the Boost.Thread build report is marked as succeeded. I'm not sure how to interpret these results. Can someone clarify this? Thank you. [1] http://www.boost.org/development/tests/trunk/developer/atomic.html [2] http://www.boost.org/development/tests/trunk/developer/output/IBM_Canada_Lt… [3] http://www.boost.org/development/tests/trunk/output/IBM_Canada_Ltd-boost-bi…

1 0

[release] add comments/changes
by Oliver Kowalke 21 Dec '12

21 Dec '12

Which files do I've to use in order to add notes about changes and new features of the next release? thx, Oliver

2 1

[Git] Extracting libraries from the svn sandbox
by Beman Dawes 20 Dec '12

20 Dec '12

One of the issues that came up in the documentation discussion was the desire to extract still useful sub-trees from the old svn sandbox, and to move them to GitHub. I've been able to extract my proposed endian library from the sandbox, history and all. Here was the process: 1) Created a empty repository named "endian" in my GitHub account via the usual web interface. 2) From the command line, in a temporary directory: git svn clone https://svn.boost.org/svn/boost/sandbox/endian (Got a "Error validating server certificate...", and gave "p" as a response to the "(R)eject, accept (t)emporarily or accept (p)ermanently?" query. The process ran for perhaps 10 minutes, but that was fine because it was successful at recovering history. And it is a one-time process.) cd endian git branch -d git-svn git push --mirror git@github.com:Beman/endian.git That's it! The git branch command deletes the git-svn branch, which is an artifact of the conversion process. Finish off by cloning the library in its permanent local location: cd permanent-local-location git clone git@github.com:Beman/endian.git Comments? Is there a better way to do this? Once comments have come in, I'll create a doc page for doing this, with a link from the FAQ. --Beman

7 15

[intrusive] boost/intrusive/detail/has_member_function_callable_with.hpp: inheriting from a const class
by Krzysztof Czainski 20 Dec '12

20 Dec '12

Hello, I tried to use a boost::container::vector with a Texas Instruments compiler, and I got an error pointing to line 88 of <boost/intrusive/detail/has_member_function_callable_with.hpp> "Invalid base class", with Type=const std::allocator<something>" Here's the context: template <typename Type> class BOOST_PP_CAT(has_member_function_named_, BOOST_INTRUSIVE_HAS_MEMBER_FUNCTION_CALLABLE_WITH_FUNCNAME) { struct BaseMixin { void BOOST_INTRUSIVE_HAS_MEMBER_FUNCTION_CALLABLE_WITH_FUNCNAME(); }; struct Base : public Type, public BaseMixin { Base(); }; // <-- line 88 The problem seems to be the constness of Type. Here's a simplified example: struct X {}; template < class T > struct Y : T { Y() {} }; int main() { Y<const X> y; } The above example compiles fine with Linux gcc version 4.5.4, and MinGW 4.5.0, but fails with my TI compiler. My question is: is this a bug of the TI compiler, or is that compiler actually right about this, and should we fix this in Boost? For now, my workaround is: struct Base : public ::boost::remove_cv<Type>::type, public BaseMixin { Base(); }; // <-- line 88 Regards, Kris

3 4

[GIT] What happens to the Trac?
by John Maddock 20 Dec '12

20 Dec '12

Currently the issue tracker is somewhat nicely integrated with SVN, what are we doing for issue tracking short/medium/long term? Apologies if this has been covered already, John.

14 24