- Boost - lists.boost.org

uBLAS tests moved to test subdirectory (confused regression results)
by Michael Stevens 31 Mar '05

31 Mar '05

A couple of weeks ago I moved the uBLAS tests into 'libs/numeric/ublas/test' for consistency with the rest of Boost. Since then the regression test results are filled with a lot of confussing results. I assume this results from the combing of stale result and the new configuration. I assume (and hope) that the next time testers do a 'clean build' things will clear up. This is a lot of work for many testers and there test machines however. Any ideas how to better make such changes in the future to avoid such confussion? Michael -- ___________________________________ Michael Stevens Systems Engineering Navigation Systems, Estimation and Bayesian Filtering http://bayesclasses.sf.net ___________________________________

2 1

Re: Serialization issuses under MS VC++ 7.1 ....
by Klaus Gütter 31 Mar '05

31 Mar '05

Hello Jason, I get the same error message when trying to deserialize this class (MSVC 7.1): class TestVector { private: std::vector<int> v; friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & BOOST_SERIALIZATION_NVP(v); } public: TestVector() { } }; Klaus "Jason Earl" <jason(a)hybd.net> schrieb im Newsbeitrag news:<004001c533fc$48724270$861aa8c0@system2600>... > If you have look at my original code, it does use > BOOST_SERIALIZATION_SPLIT_MEMBER() inside the class (sorry the code doesn't > intend with tabs nicely). > > I also have tried including both... > > #include <boost/serialization/split_member.hpp> > #include <boost/serialization/split_free.hpp> > > but neither work. I assume that the split free is obviously used by > the list > data type as it's not intrusive, and the member one is what I should include > and have been including. > > I'll look running the tests later when I get round to figuring out how > to > run them etc. I'll need to RTFM first though. From the regression thingy, it > said MSVC++7.1 worked OK though. > > I think this could be a little bug in the library (I won't speak too > loudly > here as I'm still a C++ beginner, and therefore there's a good chance I'm > still doing something very dumb), maybe someone else could confirm they get > the same problem as me when trying to serialize an class member varaible > that is a list of objects using load / save methods that are class members > under VS.NET 2003, like in my example.... > > anotherwords I have class A with separate load/save members, and then > I have > class B which has a member inside it that is of type std::list<A*>. > > Sorry in advance if I break the threaded reply structure on this list. > I > didn't have the list mail me replies. I'm new to all this mailing list > stuff. > > Jason > > _______________________________________________ > Unsubscribe & other changes: > http://lists.boost.org/mailman/listinfo.cgi/boost >

2 1

[Boost-bugs] [ boost-Bugs-1173937 ] Bad XML format outpout
by SourceForge.net 31 Mar '05

31 Mar '05

Bugs item #1173937, was opened at 2005-03-31 01:21 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=1173937&group_… Category: test Group: None Status: Open Resolution: None Priority: 5 Submitted By: Nobody/Anonymous (nobody) Assigned to: Beman Dawes (beman_dawes) Summary: Bad XML format outpout Initial Comment: IN FILE : boost_1_32_0\libs\test\src\supplied_log_formatters.cpp IN FUNCTION : xml_log_formatter::track_test_case_scope Missing accodance in test void xml_log_formatter::track_test_case_scope( std::ostream& output, test_case const& tc, bool in_out ) { if( !in_out ) m_indent -= 2; print_indent( output ); output << (in_out ? "<" : "</") << ( tc.p_type ? "TestCase" : "TestSuite" ); // HERE MISSING ORIGINAL CODE : MISSING { !! //if( in_out ) // output << " name"; print_attr_value( output, tc.p_name ); // Patch if( in_out ) { output << " name"; print_attr_value( output, tc.p_name ); } output << ">"; if( in_out ) m_indent += 2; } ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1173937&group_… ------------------------------------------------------- This SF.net email is sponsored by Demarc: A global provider of Threat Management Solutions. Download our HomeAdmin security software for free today! http://www.demarc.com/Info/Sentarus/hamr30 _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

RE: [boost] [format] conversion to string
by Martin Bonner 31 Mar '05

31 Mar '05

> While it's generally considered a bad practice to provide implicit > conversions, I think it would be good to provide such a conversion > for the boost::format class. [snip] Having been bit far too many times by bad implicit conversions, I would vote against such a change. I don't see why one shouldn't pass in a format object to a function (possibly for the function to add one argument to the format object). I still think that providing a member function with(), rather than an operator % would have been better. Then one could write: string s = format("%1%").with(10).str(); which seems does seem neater than: string s = (format("%1%") % 10).str(); If we are to change boost::format, I would prefer to add such a member function than an implicit conversion. -- Martin Bonner Martin.Bonner(a)Pitechnology.com Pi Technology, Milton Hall, Ely Road, Milton, Cambridge, CB4 6WZ, ENGLAND Tel: +44 (0)1223 441434

3 4

[date_time] MinGW no reentrant std functions patch
by Daniel Schlyder 31 Mar '05

31 Mar '05

MinGW doesn't seem to support reentrant std functions, so I think the attached patch should be applied. Best regards, -- Daniel Schlyder http://bitblaze.com/

2 4

Lock free shared_ptr on gcc, help needed
by Peter Dimov 31 Mar '05

31 Mar '05

So, shared_ptr is now lock-free on Windows, but not on the other platforms. We need a portable atomic primitive for that. The only alternative are the internal libstdc++ helpers __atomic_add and __exchange_and_add. But in order to use them correctly, we need to know what their exact semantics are with respect to memory synchronization. To the best of my knowledge, this is not documented anywhere. What do we do now? -- Peter Dimov http://www.pdimov.com

7 8

memory leak in iostreams library-zlib
by George M. Garner Jr. 31 Mar '05

31 Mar '05

Johnathan, I'm getting a memory leak when using the gzip_decompressor due to a failure to call inflateEnd/deflateEnd during object destruction. I have inserted the following line into zlib.cpp tp correct the problem: void zlib_base::reset(bool compress) { z_stream* s = static_cast<z_stream*>(stream_); zlib_error::check(compress ? deflateReset(s) : inflateReset(s)); zlib_error::check(compress ? deflateEnd(s) : inflateEnd(s)); // gmg cleanup stream->state and embedded pointers } There may be a better place for this, for example in the zlib destructor, since it may be desirable on occasion to reset a stream without destroying the state (??). Or perhaps there should be a new function to encapsulate the call to deflateEnd/inflateEnd which is called in the destructor. Regards, George.

2 1

typeof and Visual C++
by Arkadiy Vertleyb 31 Mar '05

31 Mar '05

HI all, God only knows what else can be found inside this compiler if one is willing to dig real deep. For example, Igor' Chesnokov from RSDN (Russian Software Development Network) has found a way to implement typeof() that does not require registration, and probably has compile-time performance of a native typeof. How? Apparently, some weird "feature" of Visual C++ allowed him to twick a template body at the moment of instantiation, when additional context is available, thus "registering" classes on the fly, at the moment of taking typeof(). Microsoft-specific "bugfeatures" are generally not in the area of my interests. However I do realize that, on the Microsoft compiler, this might look much more attractive then anything Peder and I have implemented. And even though I realize that this might be a serious competition, I would feel really bad not to mention this here: http://rsdn.ru/Forum/?mid=1094305 Thoughts? In particular, how do people feel about implementing someting for some particular compiler using specific "features" of this compiler? Should this be acceptable? Regards, Arkadiy -- Regards, Arkadiy

5 15

boost/threads/condition.hpp documentation patch
by Larry Lansing 30 Mar '05

30 Mar '05

Attached please find a patch for boost/libs/thread/doc/condition-ref.xml which re-adds a warning about spurious wakeups that appeared in older versions of the boost threads documentation. Compare and contrast: http://www-eleves-isia.cma.fr/documentation/BoostDoc/boost_1_29_0/libs/thre… vs. http://www.boost.org/doc/html/condition.html The newer boost.org version omits this paragraph: *** Danger: This version should always be used within a loop checking that the state logically associated with the condition has become true. Without the loop, race conditions can ensue due to possible "spurious wake ups". The second version encapsulates this loop idiom internally and is generally the preferred method. *** Unless there was an important reason that this warning was removed, it seems like the spurious wakeup issue should definitely be mentioned in the wait() and timed_wait() interfaces. Unless I've screwed it up, the attached patch should put this paragraph in the right places. Thank you for your time. -- Larry Lansing Index: condition-ref.xml =================================================================== RCS file: /cvsroot/boost/boost/libs/thread/doc/condition-ref.xml,v retrieving revision 1.3 diff -d -u -r1.3 condition-ref.xml --- condition-ref.xml 17 Jul 2004 04:33:59 -0000 1.3 +++ condition-ref.xml 15 Mar 2005 19:18:42 -0000 @@ -97,6 +97,13 @@ lock.</simpara></effects> <throws><simpara><classname>lock_error</classname> if <code>!lock.locked()</code></simpara></throws> + <notes><simpara>This version should always be used within a + loop checking that the state logically associated with the + <classname>condition</classname> has become true. Without + the loop, race conditions can ensue due to possible + "spurious wake ups". The second version encapsulates this + loop idiom internally and is generally the preferred + method.</simpara></notes> </method> <method name="wait"> @@ -153,6 +160,13 @@ otherwise <code>true</code>.</simpara></returns> <throws><simpara><classname>lock_error</classname> if <code>!lock.locked()</code></simpara></throws> + <notes><simpara>This version should always be used within a + loop checking that the state logically associated with the + <classname>condition</classname> has become true. Without + the loop, race conditions can ensue due to possible + "spurious wake ups". The second version encapsulates this + loop idiom internally and is generally the preferred + method.</simpara></notes> </method> <method name="timed_wait">

2 2

IMPORTANT: yahoogroups admin rights problem
by Toon Knapen 30 Mar '05

30 Mar '05

Hello ALL, As many of you might know, currently we are unable to add new members to the ublas-ml. The problem is that only Joerg has the rights to add members to the list but Joerg does not respond anymore. Could some of you try to contact Joerg or give me his telephone number that I talk to him about this. If we do not have a solution by end of November I will create a new project on yahoogroups (and we all have to switch). Additionally we need to contact Joerg and Mathias (again!) about applying the Boost Software License. I currently have the impression that they are not willing to change ublas to the BSL. In that case IMO, we will have to rewrite uBLAS to not get thrown out of boost. Again his phone-number would be handy if you have it available. Please contact the ml or me asap about the above issues or we will need to take drastic actions. toon

4 4