August 2006 - Boost - lists.boost.org

Re: [boost] Exception Visitor
by Ames Andreas 12 Aug '06

12 Aug '06

> -----Original Message----- > From: boost-bounces(a)lists.boost.org > [mailto:boost-bounces@lists.boost.org] On Behalf Of Alexander Nasonov > Sent: Friday, July 28, 2006 12:54 PM > Subject: Re: [boost] Exception Visitor > > Please, NO! > It's my experience that one catch(...) in a program may be a > source of many annoying glitches that users report > periodically but developers don't understand what's going on. > Removing this often helps. Users start reporting crashes. > Most difficult at this stage is to find a user who is able to > run Dr. Watson, check 'Create Crash Dump File' and send a > dump to developers :) > > Also worth noting that SEH is not standard C/C++. FWIW, I'd like to strongly second that. When catching SEH exceptions (like catch(...) does), predictable behaviour is only possible if you compile for 'asynchronous exception handling', which according to the docs increases the generated code size (I don't know about the runtime performance cost). If you compile for 'synchronous exception handling' (the more usual thing, I believe), you should either be prepared to live with bad cleanup of the stack or you install a global SEH-exception handler in each thread you start, which does nothing but rethrow and therefore crash the app. IMHO, that's nothing a library should impose on its apps. Anyway, I fail to see what's the benefit of catching segfaults _as the default behaviour_ (that's not to say there aren't applications that need to do). FWIW again, I won't use a solution which includes catch(...) anytime soon (and esp. not if it's done in a pervasive manner). cheers, aa -- Andreas Ames | Programmer | Comergo GmbH | Voice: +49 69 7505 3213 | ames AT avaya DOT com

5 6

Re: [boost] Proposed library Boost Exception updated
by Emil Dotchevski 12 Aug '06

12 Aug '06

Pavel, > "Emil Dotchevski" wrote: > > > http://www.revergestudios.com/boost-exception/boost-exception.html > > > > It is also possible to extend the interface to support enumeration of > > the > > exception info stored in boost::exception objects, but my personal > > preference is to first collect more feedback from people using the > > library, > > because I'm not sure if this enumeration is needed in practice. Brad > > King > > illustrated this idea in his earlier post: > > > > http://article.gmane.org/gmane.comp.lib.boost.devel/146322 > > > > I believe this functionality should be present > and will be used extensively at the top level > of an application. Could you provide an example of using this functionality at the top level of an application? > There should be also ability to support > visitation of exception objects. > Imagine situation: > > try { > .... > } catch (boost::exception& e) { > e << ....; > > // now some exceptions could be handled > // and some need to be passed up > ???? > > > The ??? could be typeswitch or it could > be a visitor object passed into the exception. > The visitor would be able to throw original > or new exception or handle the situation. > The visitor could be separated, reused > and is more dynamic than a typeswitch. I am not sure I understand what you mean by typeswitch. Can you illustrate your point with more code? --Emil

3 10

[random][lognormal_distribution] compilation error
by Roman Yakovenko 11 Aug '06

11 Aug '06

boost(a)lists.boost.org Hi. Instantiation of class lognormal_distribution<double> could not be compiled. Error message: D:\boost_cvs\boost\random\lognormal_distribution.hpp(63) : error C2440: 'return' : cannot convert from 'const double' to 'double &' Conversion loses qualifiers D:\boost_cvs\boost\random\lognormal_distribution.hpp(63) : while compiling class-template member function 'double &boost::lognormal_distribution<>::mean(void) const' dev\boost_random\random_export.hpp(323) : see reference to class template instantiation 'boost::lognormal_distribution<>' being compiled I think that "mean" and "sigma" functions should have return type RealType instead of RealType&. RealType& mean() const { return _mean; } ==> RealType mean() const { return _mean; } RealType& sigma() const { return _sigma; } ==> RealType sigma() const { return _sigma; } Thanks -- Roman Yakovenko C++ Python language binding http://www.language-binding.net/

2 1

Re: [boost] Boost inspection notification (2006-08-09/RC_1_34_0) *X*
by Alexander 11 Aug '06

11 Aug '06

It's uneasy to find new Boost inspection notification at my client. I spend near 5 minutes just to find this notification. May be making special brunch for notifications not such a good idea? Regards, Alexander

2 3

[test] Buglet in floating point comparison code.
by John Maddock 11 Aug '06

11 Aug '06

I'm using Boost.Test, specifically BOOST_CHECK_CLOSE with non-built-in floating point types. In particular I have a type "real_concept" that models the concepts I want an arbitrary real number type to adhere to. This type does *not* specialise std::numeric_limits: deliberately so because I want to support types with whose precision varies at runtime. This causes a problem inside safe_fpt_division which uses std::numeric_limits without the check to see if it's specialised or not. In turn this causes almost all comparisons using real_concept to succeed because safe_fpt_division will almost always return 0 for real_concept. The obvious fix is simply to place the numeric_limits checks inside an if(std::numeric_limits<FPT>::is_specialized) block, and to perform no safety checks in this case and just go ahead and perform the division regardless (there are no checks that can be performed in this case anyway). Thanks, John.

3 12

[utility] nonconstructible
by Chris Weed 11 Aug '06

11 Aug '06

Hi, I think it would be nice to have a "nonconstructible" class to derive from similar to noncopyable. Recently I was burned by a traits class that was getting constructed. I think this should also be used for template metafunctions. I envision something like the following: struct nonconstructible { private: nonconstructible() {} } template<typename T> struct my_traits : nonconstructible { typedef int my_type; }; Chris

3 4

[Boost-bugs] [ boost-Bugs-1538741 ] wstring_rvalue_from_python crash on zero length PyObject
by SourceForge.net 11 Aug '06

11 Aug '06

Bugs item #1538741, was opened at 2006-08-11 14:30 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1538741&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Python Group: None Status: Open Resolution: None Priority: 5 Submitted By: Goncalo Nuno M. de Carvalho (glslang) Assigned to: David Abrahams (david_abrahams) Summary: wstring_rvalue_from_python crash on zero length PyObject Initial Comment: Using Python 2.3 (OSX native) and OSX 10.4.7 (powerpc-apple-darwin8- gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build 5341)) Boost version is 1_33_1 and was compiled from source. When PyObject_Length(intermediate) is zero in wstring_rvalue_from_python::extract (when referring to an empty string) a crash will occur in PyUnicode_AsWideChar, most probably because 0 is then passed to it. Doing a if( result.empty( ) ) return std::wstring( ) right after the result string is constructed seems to solve this and produce correct results on my code. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1538741&group_… ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

Whatever happened to boost::circular_buffer?
by Will Wilson 11 Aug '06

11 Aug '06

Hi All, I was just wondering what happened to this very handy container, as I notice it still hasn't made it into the boost release. This is despite it being formally reviewed over two years ago. According to http://boost.org/more/formal_review_schedule.html it's listed as "Accepted -- not added yet". Is there anything preventing the addition of the circular_buffer? If not, can I just +1 including it in the next release. Cheers, Will Wilson.

4 3

[Boost-bugs] [ boost-Bugs-1538697 ] Possible problem with condition releasing
by SourceForge.net 11 Aug '06

11 Aug '06

Bugs item #1538697, was opened at 2006-08-11 13:32 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1538697&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: threads Group: None Status: Open Resolution: None Priority: 5 Submitted By: David Vest (davevest) Assigned to: Mike Glassford (glassfordm) Summary: Possible problem with condition releasing Initial Comment: I think this is a bug but since I have only just started using the threads library it could be a misunderstanding on my part. Since the actual application is quite complex I've created an imaginary scenario that shows the problem. Scenario: A single 'throttle' manager that only lets three threads at a time perform some work. When a thread wants to work it must successfully addRef and when it has finished it releases. Problem: If all available 'slots' are used I am using a condition in the addRef to wait for a release before continuing. Therefore I would expect that after the wait I can successfully assert that a slot is available. However, sometimes this does not appear to be true and I can only assume that sometimes the condition is not working properly. Work around: In CThrottle::release remove the 'unlock' so that the notify_one occurs while the method still has a lock on the m_mutexChangeState. Have I misunderstood how condition is supposed to work? Thanks, Dave. The code: #include <boost\thread\thread.hpp> #include <boost\thread\recursive_mutex.hpp> #include <boost/thread/condition.hpp> #include <boost/shared_ptr.hpp> #include <windows.h> #include <crtdbg.h> #include <vector> class CThrottle { int m_nThrottle; public: CThrottle() : m_nThrottle(0) {} void addRef() { boost::mutex::scoped_lock lockIncrement ( m_mutexIncrement ); boost::recursive_mutex::scoped_lock lockChangeState( m_mutexChangeState ); if ( !isAllowed() ) m_condition.wait( lockChangeState ); _ASSERT( isAllowed() ); // PROBLEM ++m_nThrottle; } void release() { boost::recursive_mutex::scoped_lock lockChangeState( m_mutexChangeState ); --m_nThrottle; lockChangeState.unlock(); m_condition.notify_one(); } bool isAllowed() { boost::recursive_mutex::scoped_lock lockChangeState( m_mutexChangeState ); return m_nThrottle < 3; } private: boost::mutex m_mutexIncrement; boost::recursive_mutex m_mutexChangeState; boost::condition m_condition; }; class CTestThread { public: CTestThread( CThrottle * pThrottle ) : m_pThrottle( pThrottle ) {} void operator()() { m_pThrottle->addRef(); std::cout << "Do something\n"; m_pThrottle->release(); } CThrottle * m_pThrottle; }; typedef boost::shared_ptr<CTestThread> _TestThreadPtr; int main(int argc, _TCHAR* argv[]) { CThrottle * pThrottle = new CThrottle(); boost::thread_group groupThreads; std::vector<_TestThreadPtr> list; for( int n=0; n < 10000; ++n ) { _TestThreadPtr ptrTest( new CTestThread ( pThrottle ) ); groupThreads.create_thread( *ptrTest.get() ); list.push_back( ptrTest ); } groupThreads.join_all(); delete pThrottle; return 0; } ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1538697&group_… ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

boost wanted libraries
by shams 11 Aug '06

11 Aug '06

Hi, Does anyone know about the following libraries: boost::logging (required much in offically boost) boost::smart_assert boost::output_formatters if these libraries are available in the sandbox cvs? Does anyone know if they are ever going going to make it into boost? Thanks Shams

2 2