- Boost - lists.boost.org

[Boost-bugs] [ boost-Patches-1438626 ] rational.hpp::gcd returns a negative value sometimes
by SourceForge.net 25 Feb '06

25 Feb '06

Patches item #1438626, was opened at 2006-02-25 09:49 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=307586&aid=1438626&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: None Group: None Status: Open Resolution: None Priority: 5 Submitted By: David Benbennick (dbenbenn) Assigned to: Nobody/Anonymous (nobody) Summary: rational.hpp::gcd returns a negative value sometimes Initial Comment: boost/rational.hpp provides a gcd function, which is supposed to be non-negative. It sometimes returns a negative value, which can cause problems in other parts of rational.hpp. For example, assuming long is a 4-byte type, boost::gcd<long>(6, -2147483648) returns -2. As a result, boost::rational<long>(-1073741821, 6) + boost::rational<long>(-1073741827, 6) produces an invalid rational number, 1073741824/-3, instead of the correct answer -1073741824/3. Here is a small patch to fix the problem. Note that this is how boost/math/common_factor_rt.hpp calculates the greatest common divisor. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=307586&aid=1438626&group_… ------------------------------------------------------- This SF.Net email is sponsored by xPML, a groundbreaking scripting language that extends applications into web and mobile media. Attend the live webcast and join the prime developer group breaking into this new coding territory! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642 _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

[config] BOOST_NO_MEMBER_FUNCTION_SPECIALIZATIONS
by AlisdairM 25 Feb '06

25 Feb '06

AFAICT, the confing macro BOOST_NO_MEMBER_FUNCTION_SPECIALIZATIONS is not used anywhere outside the config tests that test for this feature! Should we retire old workarounds like this, that are no longer tested for? Or do they add value to the config library, supporting 'customer' code relying on config for its own feature detection / workarounds? -- AlisdairM

1 0

Re: [boost] [multi_index] - Thank You
by JOAQUIN LOPEZ MU?Z 25 Feb '06

25 Feb '06

----- Mensaje original ----- De: Joseph Wu <josephclwu(a)yahoo.com> Fecha: Viernes, Febrero 24, 2006 5:34 pm Asunto: [boost] [multi_index] - Thank You > Hi Joaquin, > > I just would like to send a personal note to say thank you for the > wonderfuljob on multi-index. I have used it since the early days, > and found it rather > useful. Having just read the 1.34 release notes, the addition of > random-access index and other constant improvements are really > encouraging. > Thanks again, > > Joseph Hello Joseph, thank you very much for your compliments, it's a rewarding feeling to know the lib is appreciated and used out there. Some proposals of new features have had low response in the past, which made me wonder whether the features were worth including after all. So I'm always eager to receive as much feedback as possible on how to make Boost.MultiIndex better suited to its users' needs. Best regards, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

1 0

spirit typeof tests failing for cw-8_3
by Arkadiy Vertleyb 25 Feb '06

25 Feb '06

I am not currently planning to support emulation mode for this compiler. AFAIU, there are no plans to support this compiler beyond 1.34, so I figured it doesn't worth the effort (and extra code for workarounds), especially since the native mode is there. Please let me know if having emulation mode on cw-8_3 is important for some reason. Regards, Arkadiy

1 0

[serialization] BOOST_CLASS_EXPORT in a separate static library
by Marcin Kalicinski 24 Feb '06

24 Feb '06

I noticed strange behavior of serialization on VC 8.0. I use BOOST_CLASS_EXPORT macro in a .cpp file to register a derived class, so that it can be properly serialized through base class pointer. This works fine, as long the .obj file resulting from compilation of .cpp file which uses BOOST_CLASS_EXPORT macro is linked directly to the final executable. But when I first put this .obj file into a static library, and then link the library to executable, I start getting unregistered_class exceptions - as if there was no BOOST_CLASS_EXPORT invoked. So far I believed that there is no difference between linking to object file and to static library file containing that object file. It seems my beliefs were now shattered. Has anybody else had this issue before? Thank you, Marcin

3 2

[bind][function_types] Why do bind objects not have an ::arity, ... interface ?
by Peter Soetens 24 Feb '06

24 Feb '06

Hi, I want to use a partial template specialisation which is selected on basis of the arity of a bind function object type. Event more, I'd like to access the ::arg1_type, etc types of a bind object type. Why is this not part of the boost::bind or lambda::bind interface ? I thought solving it with the (to be accepted) function_types::function_type_arity<BoundType>::type::value, but the latest version of function_types does not 'support' inspection of bound object types. Other library writers must have felt a need for this interface as well ? Or am I missing something ? With kind regards, Peter Soetens -- Peter Soetens -- FMTC -- <http://www.fmtc.be>

3 2

Re: [boost] [serialization] [multi_index][cw-8_3] help request forregressions in CW 8.3 Windows
by JOAQUIN LOPEZ MU?Z 24 Feb '06

24 Feb '06

----- Mensaje original ----- De: David Abrahams <dave(a)boost-consulting.com> Fecha: Viernes, Febrero 24, 2006 5:38 pm Asunto: [boost] [serialization] [multi_index][cw-8_3] help request forregressions in CW 8.3 Windows > Joaquín Mª López Muñoz <joaquin(a)tid.es> writes: > > > Excuse my insistence, but I have no other way to try to fix current > > regressions in Boost.MultiIndex but to ask for external support. So > > I'm reposting my plea for help: > > > > Hello, > > > > Most current tests for Boost.MultiIndex under CW 8.3 in Windows > > fail at run time with the following: > > > > Running 1 test case... > > unknown location(0): fatal error in "test_main_caller( argc, > argv )": > > unknown type > > > > *** 1 failure detected in test suite "Test Program" > > > > EXIT STATUS: 201 > > > > (see for instance http://tinyurl.com/fqarz) compilation and > building> having gone fine. As this is giving very little info, > I'd be grateful if > > > > someone with local access to CW 8.3 Windows (it is likely that > > Mac will also do) volunteer to do some exploration of the > > problem. If you are such volunteer, please contact me privately. > > I tried it; I'm mostly seeing this: > > cw.compile.c++ ..\..\..\bin.v2\libs\serialization\build\cw- > 8.3\debug\basic_serializer_map.obj ### mwcc Compiler: > # In: ..\..\..\boost\archive\detail\basic_serializer_map.hpp > # From: ..\..\serialization\src\basic_serializer_map.cpp > # --------------------------------------------------------- > # 56: }; > # Error: ^ > # using implicit copy assigment for class with const or > reference member ('boost::archive::detail::basic_serializer_map') > > Errors caused tool to abort. > > which is a problem in serialization. Aside from that all the tests > are passing for me with cw-8. Do you mean that multi_index/test_basic, for instance, passes? What could be then the reason for the discrepancy with http://tinyurl.com/fqarz? Are there maybe upgrades to the CW 8.3 compiler and your version differs from that of metacomm? Thanks for your input, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

[rerepost][multi_index][cw-8_3] help request for regressions in CW 8.3 Windows
by Joaquín Mª López Muñoz 24 Feb '06

24 Feb '06

Excuse my insistence, but I have no other way to try to fix current regressions in Boost.MultiIndex but to ask for external support. So I'm reposting my plea for help: Hello, Most current tests for Boost.MultiIndex under CW 8.3 in Windows fail at run time with the following: Running 1 test case... unknown location(0): fatal error in "test_main_caller( argc, argv )": unknown type *** 1 failure detected in test suite "Test Program" EXIT STATUS: 201 (see for instance http://tinyurl.com/fqarz) compilation and building having gone fine. As this is giving very little info, I'd be grateful if someone with local access to CW 8.3 Windows (it is likely that Mac will also do) volunteer to do some exploration of the problem. If you are such volunteer, please contact me privately. Thank you very much, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

[multi_index] key swap status
by gruenedd＠idmt.fraunhofer.de 24 Feb '06

24 Feb '06

Hello, What is the status of key swapping, issued here[1]? Do you see a chance to have this included sometime? :] Best Regards, David [1] http://lists.boost.org/boost-users/2005/04/11223.php ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

3 2

boost::null
by Raffaele Romito 24 Feb '06

24 Feb '06

What about a null_helper class? This is my approach: namespace boost { static class null_helper { public: template <typename T> inline operator shared_ptr<T>() const { return shared_ptr<T>(); } template <typename T> inline operator T *() const { return 0; } }null; } I find this very useful in samples like this: class node { public: node(boost::shared_ptr<node> parent = boost::null, // instead of boost::shared_ptr<node>() int *value = boost::null); // no difference from ordinary pointers }; Waiting for your feedback ;) ------------- Raffaele Romito

1 0