- Boost - lists.boost.org

Re: [boost] [Review] Review of the Accumulators library begins today Jan 29
by Steven Watanabe 31 Jan '07

31 Jan '07

AMDG Eric Niebler wrote: > Hi, Steven, > > Steven Watanabe wrote: > >> accumulators.qbk lines 35-37 >> [def _mpl_ [@../../libs/mpl MPL]] >> [def _mpl_lambda_expression_ >> [@../../libs/mpl/doc/refmanual/lambda-expression.html MPL Lambda >> Expression]] >> [def _parameters_ [@../../libs/parameters Boost.Parameters]] >> these links are bad >> > > > Thanks. I think I was overly-optimistic when I wrote that. The links > would work if Accumulators were in it's rightful (a-hem) place in the > Boost tree. ;-) > unless I missed something libs/accumulators/doc/../../libs/ == libs/libs/ Which I think is not what you intend. Somehow in the user_s_guide.html this becomes libs/accumulators/libs/ I'm not quite sure whether this is a quickbook bug or the file was built as though it were in libs/accumulators/doc/html/ > >> post_construct() >> Unless for some obscure reason that I am missing it is just >> too difficult, I would strongly prefer to make sure the constructors >> are called in the correct order. >> >> The whole dance with fusion::vector is unnecessary. >> It would be simpler to use >> struct accumulator_tuple_base { >> template<class Args> >> accumulator_tuple(const Args&) {} >> } >> template<class T, class Base> >> struct accumulator_tuple : T, Base { >> template<class Args> >> accumulator_tuple(const Args& args) : T(args), Base(args) {} >> }; >> Since every accumulator is put before all the >> accumulators that depend on it in make_accumulator_tuple >> you can write >> typedef >> typename mpl::reverse_fold< >> sorted_accumulator_vector >> , accumulator_tuple_base >> , accumulator_tuple<mpl::_2, mpl::_1> >> >::type >> type; >> Then every accumulator is constructed after any >> others that it depends on. As a bonus the only >> limit to the number of accumulators is whatever >> the compiler can handle. >> > > > > This would be a valid design choice, but it would involve reinventing a > lot of Fusion. I use Fusion both to generate the accumulator set, to > process the accumulators, and to find accumulators that match Fusion > predicates. Code reuse is good. > I'm not suggesting that you reinvent the fusion algorithms. I just think that fusion::vector is not necessarily a good sequence to use when the constructor parameters are all identical. Also if you write the sequence then you can guarantee the order of initialization. template<class From, class To> struct match_cv : mpl::if_<is_const<From>, const To, To> {}; struct accumulator_set_tag {}; struct accumulator_set_impl_base { typedef accumulator_set_tag fusion_tag; template<class Args> accumulator_set_impl(const Args&) {} } template<class T, class Base> struct accumulator_set_impl : T, Base { typedef T value_type; typedef Base next_type; accumulator_set_impl(const accumulator_set_impl& other) : T(static_cast<const T&>(other)), Base(static_cast<Base&>(other)) {} template<class Args> accumulator_set_impl(const Args& args) : T(args), Base(args) {} }; struct accumulator_set_iterator_tag {}; template<class Set> struct accumulator_set_iterator { typedef accumulator_set_iterator_tag fusion_tag; accumulator_set_iterator(Set& s) : set(s) {} Set& set; }; template<> struct boost::fusion::extension::begin_impl<accumulator_set_tag> { template<class Set> struct apply { typedef accumulator_set_iterator<Set> type; type call(Set& set) { return(type(set)); } }; }; template<> struct boost::fusion::extension::end_impl<accumulator_set_tag> { template<class Set> struct apply { typedef accumulator_set_iterator<typename match_cv<Set, accumulator_set_impl_base>::type> type; type call(Set& set) { return(type(set)); } }; }; template<> struct boost::fusion::extension::next_impl<accumulator_set_iterator_tag> { template<class Iterator> struct apply { typedef typename T::set_type set_type; typedef accumulator_set_iterator< typename match_cv<set_type, typename set_type::next_type>::type& > type; type call(Set& set) { return(set.set); } }; }; template<> struct boost::fusion::extension::deref_impl<accumulator_set_iterator_tag> { template<class Iterator> struct apply { typedef typename T::set_type set_type; typedef typename match_cv<set_type, typename set_type::value_type>::type& type; type call(Set& set) { return(static_cast<type>(set.set)); } }; }; > Regarding post_process(), I decided at the time, after discussing this > with Joel de Guzman (of Fusion) and Dave Abrahams, that relying on the > construction order of accumulators within a set would be a mistake. It > would rule out size optimizations like compressed_pair, where empty > accumulators are optimized away via the Empty Base Optimization. Fusion > doesn't do this optimization today, but it may in the future. > But EBCO can't help you anyway. Two distinct sub objects of the same type (accumulator_base) cannot share the same address. I assumed that you didn't want to rely on fusion initializing the elements in order, that was part of the reason I suggested rolling your own fusion sequence. > > >> What happens if you freeze a feature that another feature depends on? >> #include <iostream> >> #include <boost/accumulators/accumulators.hpp> >> #include <boost/accumulators/statistics/mean.hpp> >> #include <boost/accumulators/statistics/sum.hpp> >> >> using namespace boost::accumulators; >> >> int main() { >> accumulator_set<double, features<tag::mean, droppable<tag::sum> >> > > accum_set; >> accum_set(1.0); >> accum_set(2.0); >> accum_set.drop<tag::sum>(); >> accum_set(3.0); >> std::cout << mean(accum_set) << std::endl; //prints 1 >> } >> This behavior can easily cause surprises. It must >> be fixed to print 2. I'm not quite sure how to implement >> it correctly. I think that probably you need to store the >> result at the point that drop is called for external lookup >> and continue to collect data for internal usage. Of course, >> this only applies in the face of dependencies. >> > > > The code is doing something a little unusual, but the result looks > correct to me. What do you think the result should be, and why? > The result should be 2 because 1.0 + 2.0 + 3.0 / 3 = 2.0 Users of particular accumulators should not have to care about the internal dependencies. The mean should return the mean of all the values that have been added regardless of whether count/sum have been dropped. > >> accumulators_fwd.hpp #define's >> FUSION_MAX_VECTOR_SIZE/FUSION_MAX_TUPLE_SIZE >> This is a very bad idea as it might make >> make the value smaller than it would otherwise >> have been. Please just check FUSION_MAX_VECTOR_SIZE >> and issue a #error directive if it is too small. >> > > > I had it that way at first, but changed it due to user feedback. The > idea is to make it so that you can just #define > BOOST_ACCUMULATORS_MAX_FEATURES and not have to worry about configuring > Fusion. > > > Consider. file a.cpp #include <boost/accumulators/accumulators.hpp> #include <boost/fusion/sequence/container/vector.hpp> file b.cpp #include <boost/fusion/sequence/container/vector.hpp> #include <boost/accumulators/accumulators.hpp> I'm not absolutely certain that this actually causes a problem with ODR, but even if it doesn't now it is very fragile. > >> The docs need to state clearly what happens when >> more than one accumulator maps to the same feature. >> > > > Right. The first accumulator with a particular feature is the one that > gets used. Any others with the same feature are ignored. > And the ones that are specified in the sequence are processed before those found indirectly. In Christ, Steven Watanabe

3 2

[graph] subgraph problems
by Hailin Jin 31 Jan '07

31 Jan '07

I compiled and ran the subgraph.cpp which comes with the library (in example directory). I got results which are different from subgraph.expected. Here are my outputs: G0: 0 --> 1 1 --> 2 3 2 --> 5 3 --> 4 --> 1 5 5 --> 3 0(0,1) 1(1,2) 2(1,3) 6(2,5) 3(4,1) 4(4,5) 5(5,3) G1: 0 --> 2 1 --> 2 2 --> 6(0,2) 4(1,2) G2: 0 --> 1 1 --> 0(0,1) I also have a question: in the output, the indices for the both subgraphs start from 0 which makes sense to me. However, why don't the edge index start from 0, as well? It looks like the edge index matches their corresponding indices in the main graph. I don't think it makes sense. Moreover, it creates problems when I try to attach properties to edges because the maximum edge index for a subgraph is larger than the number of edges in that subgraph. Thanks. Hailin

1 0

Re: [boost] [Review] Review of the Accumulators library beginstoday Jan 29
by Steven Watanabe 31 Jan '07

31 Jan '07

AMDG "John Maddock" wrote: > Right: and for some accumulators the addition of two accumulators may not be > possible at all. But can't we make this a conceptual requirement that if > you try and combine two accumulator sets, then each accumulator must have a > "combine" method (otherwise you get a compile time error). The "combine" > methods for min/max/sum/count etc are all trivial aren't they - and could be > added fairly easily ? But Eric would know the details :-) > > John. Why not let accumulator_set support a fusion sequence interface? Then combine/reset could be implemented using fusion::for_each, right? In Christ, Steven Watanabe

1 0

Re: [boost] [Review] Review of the Accumulators library begins today Jan 29
by Steven Watanabe 31 Jan '07

31 Jan '07

AMDG John Phillips wrote: > The formal review of the Boost.Accumulators library, submitted by Eric > Neibler begins today and ends on Wednesday, February 7. The library is > available from > > http://boost-consulting.com/vault/index.php?directory=Math%20-%20Numerics > > From the documentation: > > Boost.Accumulators is both a library for incremental statistical > computation as well as an extensible framework for incremental > calculation in general. The library deals primarily with the concept of > an accumulator, which is a primitive computational entity that accepts > data one sample at a time and maintains some internal state. These > accumulators may offload some of their computations on other > accumulators, on which they depend. Accumulators are grouped within an > accumulator set. Boost.Accumulators resolves the inter-dependencies > between accumulators in a set and ensures that accumulators are > processed in the proper order. > > Your comments may be brief or lengthy. If you identify problems along > the way, please note if they are minor, serious, or showstoppers. > > Here are some questions you might want to answer in your review: > ? What is your evaluation of the design? > ? What is your evaluation of the implementation? > ? What is your evaluation of the documentation? > ? What is your evaluation of the potential usefulness of > the library? > ? Did you try to use the library? With what compiler? Did > you have any problems? > ? How much effort did you put into your evaluation? A glance? A > quick reading? In-depth study? > ? Are you knowledgeable about the problem domain? > > And finally, every review should answer this question: > ? Do you think the library should be accepted as a Boost library? > Be sure to say this explicitly so that your other comments don't > obscure your overall opinion. > > All interested parties are encouraged to submit a review. These can be > sent to the developer list, the user list, or if you don't want to share > your review with the general public it can be sent directly to me. > > Thanks to all for the work you will do on this review. > > > John Phillips > Review Manager user_s_guide.html#id450311 line 363 accumulators.qbk line 189 for some reason external<> is not a link in the html accumulators.qbk lines 35-37 [def _mpl_ [@../../libs/mpl MPL]] [def _mpl_lambda_expression_ [@../../libs/mpl/doc/refmanual/lambda-expression.html MPL Lambda Expression]] [def _parameters_ [@../../libs/parameters Boost.Parameters]] these links are bad post_construct() Unless for some obscure reason that I am missing it is just too difficult, I would strongly prefer to make sure the constructors are called in the correct order. The whole dance with fusion::vector is unnecessary. It would be simpler to use struct accumulator_tuple_base { template<class Args> accumulator_tuple(const Args&) {} } template<class T, class Base> struct accumulator_tuple : T, Base { template<class Args> accumulator_tuple(const Args& args) : T(args), Base(args) {} }; Since every accumulator is put before all the accumulators that depend on it in make_accumulator_tuple you can write typedef typename mpl::reverse_fold< sorted_accumulator_vector , accumulator_tuple_base , accumulator_tuple<mpl::_2, mpl::_1> >::type type; Then every accumulator is constructed after any others that it depends on. As a bonus the only limit to the number of accumulators is whatever the compiler can handle. What happens if you freeze a feature that another feature depends on? #include <iostream> #include <boost/accumulators/accumulators.hpp> #include <boost/accumulators/statistics/mean.hpp> #include <boost/accumulators/statistics/sum.hpp> using namespace boost::accumulators; int main() { accumulator_set<double, features<tag::mean, droppable<tag::sum> > > accum_set; accum_set(1.0); accum_set(2.0); accum_set.drop<tag::sum>(); accum_set(3.0); std::cout << mean(accum_set) << std::endl; //prints 1 } This behavior can easily cause surprises. It must be fixed to print 2. I'm not quite sure how to implement it correctly. I think that probably you need to store the result at the point that drop is called for external lookup and continue to collect data for internal usage. Of course, this only applies in the face of dependencies. The user guide needs to specify what header each component is in. BOOST_PARAMETER_NESTED_KEYWORD.html There is no description of what this macro is for. accumulator_set.hpp line 149 corresponds to -- What? accumulators_fwd.hpp #define's FUSION_MAX_VECTOR_SIZE/FUSION_MAX_TUPLE_SIZE This is a very bad idea as it might make make the value smaller than it would otherwise have been. Please just check FUSION_MAX_VECTOR_SIZE and issue a #error directive if it is too small. For simple cases at least the library seems to cause little to no overhead with msvc compared to hard coding everything. The docs need to state clearly what happens when more than one accumulator maps to the same feature. Overall the users guide is quite good. I had no difficulty understanding it. The reference manual is pretty sketchy though. For the most part the interface is clean and easy to use. I definitely think Accumulators should be included in boost. In Christ, Steven Watanabe

5 6

Regression results of JustSoftwareSolutions
by Roland Schwarz 31 Jan '07

31 Jan '07

@Anthony: The below result of your latest run let me believe, there still is something wrong with your settings: call "c:\Program Files\Microsoft Visual Studio\VC98\Bin\VCVARS32.BAT" >nul cl /Zm800 -nologo @"g:\boost\rc-1-34\results\boost\bin.v2\libs\serialization\build\msvc-6.5\debug\link-static\threading-multi\xml_wgrammar.obj.rsp" xml_wgrammar.cpp ..\boost/spirit/core/config.hpp(56) : fatal error C1189: #error : "Compiler not supported. See note in <boost/spirit/core/config.hpp>" Could you please post the content of: \boost\rc-1-34\results\boost\bin.v2\libs\serialization\build\msvc-6.5\debug\link-static\threading-multi\xml_wgrammar.obj.rsp and the content of your: user-config.jam ? Thank you, Roland

2 3

[Report] 1052 regressions on RC_1_34_0 (2007-01-31)
by Douglas Gregor 31 Jan '07

31 Jan '07

Boost Regression test failures Report time: 2007-01-31T13:22:54Z This report lists all regression test failures on release platforms. Detailed report: http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/i… The following platforms have a large number of failures: borland-5.8.2 msvc-6.5 1052 failures in 17 libraries (28 are from non-broken platforms) algorithm/string (1) bind (1) date_time (0 of 7 failures are from non-broken platforms) dynamic_bitset (0 of 1 failures are from non-broken platforms) filesystem (2) function (1) iterator (2) math (1) multi_index (0 of 1 failures are from non-broken platforms) numeric/interval (7) python (1) regex (2 of 3 failures are from non-broken platforms) serialization (0 of 996 failures are from non-broken platforms) spirit (2) test (2 of 20 failures are from non-broken platforms) thread (1) wave (5) Test failures marked with a (*) represent tests that failed on platforms that are considered broken. They are likely caused by misconfiguration by the regression tester or a failure in a core library such as Test or Config. |algorithm/string| replace: msvc-8.0 |bind| mem_fn_eq_test: msvc-8.0 |date_time| testgreg_serialize: borland-5.8.2* testgreg_serialize_dll: borland-5.8.2* testgreg_serialize_xml: borland-5.8.2* testtime_serialize: borland-5.8.2* testtime_serialize_std_config: borland-5.8.2* testtime_serialize_xml: borland-5.8.2* testtime_serialize_xml_std_config: borland-5.8.2* |dynamic_bitset| dyn_bitset_unit_tests2: borland-5.8.2* |filesystem| path_test: msvc-8.0 path_test_dll: msvc-8.0 |function| contains_test: msvc-8.0 |iterator| filter_iterator_test: msvc-8.0 lvalue_concept_fail: msvc-8.0 |math| complex_test: msvc-8.0 |multi_index| test_serialization: msvc-6.5* |numeric/interval| det: msvc-8.0 integer: msvc-8.0 mul: msvc-8.0 overflow: msvc-8.0 pi: msvc-8.0 pow: msvc-8.0 test_float: msvc-8.0 |python| exec: msvc-8.0 |regex| regex_config_info: borland-5.8.2* regex_regress_threaded: msvc-8.0 test_grep: sun-5.8 |serialization| test_array_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_array_binary_archive_dll: borland-5.8.2* msvc-6.5* test_array_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_array_text_archive_dll: borland-5.8.2* msvc-6.5* test_array_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_array_text_warchive_dll: borland-5.8.2* msvc-6.5* test_array_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_array_xml_archive_dll: borland-5.8.2* msvc-6.5* test_array_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_array_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_binary_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_binary_binary_archive_dll: borland-5.8.2* msvc-6.5* test_binary_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_binary_text_archive_dll: borland-5.8.2* msvc-6.5* test_binary_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_binary_text_warchive_dll: borland-5.8.2* msvc-6.5* test_binary_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_binary_xml_archive_dll: borland-5.8.2* msvc-6.5* test_binary_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_binary_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_class_info_load_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_load_binary_archive_dll: borland-5.8.2* msvc-6.5* test_class_info_load_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_load_text_archive_dll: borland-5.8.2* msvc-6.5* test_class_info_load_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_load_text_warchive_dll: borland-5.8.2* test_class_info_load_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_load_xml_archive_dll: borland-5.8.2* msvc-6.5* test_class_info_load_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_load_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_class_info_save_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_save_binary_archive_dll: borland-5.8.2* msvc-6.5* test_class_info_save_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_save_text_archive_dll: borland-5.8.2* msvc-6.5* test_class_info_save_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_save_text_warchive_dll: borland-5.8.2* msvc-6.5* test_class_info_save_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_save_xml_archive_dll: borland-5.8.2* msvc-6.5* test_class_info_save_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_class_info_save_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_contained_class_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_contained_class_binary_archive_dll: borland-5.8.2* msvc-6.5* test_contained_class_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_contained_class_text_archive_dll: borland-5.8.2* msvc-6.5* test_contained_class_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_contained_class_text_warchive_dll: borland-5.8.2* msvc-6.5* test_contained_class_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_contained_class_xml_archive_dll: borland-5.8.2* msvc-6.5* test_contained_class_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_contained_class_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_cyclic_ptrs_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_cyclic_ptrs_binary_archive_dll: borland-5.8.2* msvc-6.5* test_cyclic_ptrs_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_cyclic_ptrs_text_archive_dll: borland-5.8.2* msvc-6.5* test_cyclic_ptrs_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_cyclic_ptrs_text_warchive_dll: borland-5.8.2* msvc-6.5* test_cyclic_ptrs_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_cyclic_ptrs_xml_archive_dll: borland-5.8.2* msvc-6.5* test_cyclic_ptrs_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_cyclic_ptrs_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_delete_pointer_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_delete_pointer_binary_archive_dll: borland-5.8.2* msvc-6.5* test_delete_pointer_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_delete_pointer_text_archive_dll: borland-5.8.2* msvc-6.5* test_delete_pointer_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_delete_pointer_text_warchive_dll: borland-5.8.2* msvc-6.5* test_delete_pointer_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_delete_pointer_xml_archive_dll: borland-5.8.2* msvc-6.5* test_delete_pointer_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_delete_pointer_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_demo: borland-5.8.2* borland-5.8.2* test_demo_auto_ptr: borland-5.8.2* borland-5.8.2* msvc-6.5* test_demo_auto_ptr_dll: borland-5.8.2* msvc-6.5* test_demo_dll: borland-5.8.2* test_demo_exception: borland-5.8.2* borland-5.8.2* test_demo_exception_dll: borland-5.8.2* test_demo_fast_archive: borland-5.8.2* borland-5.8.2* test_demo_fast_archive_dll: borland-5.8.2* test_demo_pimpl: borland-5.8.2* borland-5.8.2* test_demo_pimpl_dll: borland-5.8.2* test_demo_polymorphic: borland-5.8.2* borland-5.8.2* msvc-6.5* test_demo_polymorphic_dll: borland-5.8.2* msvc-6.5* test_demo_portable_archive: borland-5.8.2* borland-5.8.2* test_demo_portable_archive_dll: borland-5.8.2* test_demo_shared_ptr: borland-5.8.2* borland-5.8.2* test_demo_shared_ptr_dll: borland-5.8.2* test_demo_xml: borland-5.8.2* borland-5.8.2* msvc-6.5* test_demo_xml_dll: borland-5.8.2* msvc-6.5* test_demo_xml_load: borland-5.8.2* borland-5.8.2* test_demo_xml_load_dll: borland-5.8.2* test_demo_xml_save: borland-5.8.2* borland-5.8.2* test_demo_xml_save_dll: borland-5.8.2* test_deque_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_deque_binary_archive_dll: borland-5.8.2* msvc-6.5* test_deque_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_deque_text_archive_dll: borland-5.8.2* msvc-6.5* test_deque_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_deque_text_warchive_dll: borland-5.8.2* msvc-6.5* test_deque_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_deque_xml_archive_dll: borland-5.8.2* msvc-6.5* test_deque_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_deque_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_derived_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_binary_archive_dll: borland-5.8.2* msvc-6.5* test_derived_class_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_binary_archive_dll: borland-5.8.2* msvc-6.5* test_derived_class_ptr_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_ptr_binary_archive_dll: borland-5.8.2* msvc-6.5* test_derived_class_ptr_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_ptr_text_archive_dll: borland-5.8.2* msvc-6.5* test_derived_class_ptr_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_ptr_text_warchive_dll: borland-5.8.2* msvc-6.5* test_derived_class_ptr_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_ptr_xml_archive_dll: borland-5.8.2* msvc-6.5* test_derived_class_ptr_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_ptr_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_derived_class_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_text_archive_dll: borland-5.8.2* msvc-6.5* test_derived_class_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_text_warchive_dll: borland-5.8.2* msvc-6.5* test_derived_class_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_xml_archive_dll: borland-5.8.2* msvc-6.5* test_derived_class_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_class_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_derived_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_text_archive_dll: borland-5.8.2* msvc-6.5* test_derived_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_text_warchive_dll: borland-5.8.2* msvc-6.5* test_derived_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_xml_archive_dll: borland-5.8.2* msvc-6.5* test_derived_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_derived_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_diamond_binary_archive: borland-5.8.2* borland-5.8.2* test_diamond_binary_archive_dll: borland-5.8.2* test_diamond_text_archive: borland-5.8.2* borland-5.8.2* test_diamond_text_archive_dll: borland-5.8.2* test_diamond_text_warchive: borland-5.8.2* borland-5.8.2* test_diamond_text_warchive_dll: borland-5.8.2* test_diamond_xml_archive: borland-5.8.2* borland-5.8.2* test_diamond_xml_archive_dll: borland-5.8.2* test_diamond_xml_warchive: borland-5.8.2* borland-5.8.2* test_diamond_xml_warchive_dll: borland-5.8.2* test_exported_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_exported_binary_archive_dll: borland-5.8.2* msvc-6.5* test_exported_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_exported_text_archive_dll: borland-5.8.2* msvc-6.5* test_exported_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_exported_text_warchive_dll: borland-5.8.2* msvc-6.5* test_exported_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_exported_xml_archive_dll: borland-5.8.2* msvc-6.5* test_exported_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_exported_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_list_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_binary_archive_dll: borland-5.8.2* msvc-6.5* test_list_ptrs_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_ptrs_binary_archive_dll: borland-5.8.2* msvc-6.5* test_list_ptrs_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_ptrs_text_archive_dll: borland-5.8.2* msvc-6.5* test_list_ptrs_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_ptrs_text_warchive_dll: borland-5.8.2* msvc-6.5* test_list_ptrs_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_ptrs_xml_archive_dll: borland-5.8.2* msvc-6.5* test_list_ptrs_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_ptrs_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_list_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_text_archive_dll: borland-5.8.2* msvc-6.5* test_list_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_text_warchive_dll: borland-5.8.2* msvc-6.5* test_list_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_xml_archive_dll: borland-5.8.2* msvc-6.5* test_list_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_list_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_map_binary_archive: borland-5.8.2* borland-5.8.2* test_map_binary_archive_dll: borland-5.8.2* test_map_text_archive: borland-5.8.2* borland-5.8.2* test_map_text_archive_dll: borland-5.8.2* test_map_text_warchive: borland-5.8.2* borland-5.8.2* test_map_text_warchive_dll: borland-5.8.2* test_map_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_map_xml_archive_dll: borland-5.8.2* msvc-6.5* test_map_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_map_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_mi_binary_archive: borland-5.8.2* borland-5.8.2* test_mi_binary_archive_dll: borland-5.8.2* test_mi_text_archive: borland-5.8.2* borland-5.8.2* test_mi_text_archive_dll: borland-5.8.2* test_mi_text_warchive: borland-5.8.2* borland-5.8.2* test_mi_text_warchive_dll: borland-5.8.2* test_mi_xml_archive: borland-5.8.2* borland-5.8.2* test_mi_xml_archive_dll: borland-5.8.2* test_mi_xml_warchive: borland-5.8.2* borland-5.8.2* test_mi_xml_warchive_dll: borland-5.8.2* test_mult_archive_types: borland-5.8.2* borland-5.8.2* test_mult_archive_types_dll: borland-5.8.2* test_multiple_ptrs_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_multiple_ptrs_binary_archive_dll: borland-5.8.2* msvc-6.5* test_multiple_ptrs_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_multiple_ptrs_text_archive_dll: borland-5.8.2* msvc-6.5* test_multiple_ptrs_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_multiple_ptrs_text_warchive_dll: borland-5.8.2* msvc-6.5* test_multiple_ptrs_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_multiple_ptrs_xml_archive_dll: borland-5.8.2* msvc-6.5* test_multiple_ptrs_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_multiple_ptrs_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_no_rtti_binary_archive: msvc-6.5* test_no_rtti_binary_archive_dll: msvc-6.5* test_no_rtti_text_archive: msvc-6.5* test_no_rtti_text_archive_dll: msvc-6.5* test_no_rtti_text_warchive: msvc-6.5* test_no_rtti_text_warchive_dll: msvc-6.5* test_no_rtti_xml_archive: msvc-6.5* test_no_rtti_xml_archive_dll: msvc-6.5* test_no_rtti_xml_warchive: msvc-6.5* test_no_rtti_xml_warchive_dll: msvc-6.5* test_non_default_ctor2_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor2_binary_archive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor2_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor2_text_archive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor2_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor2_text_warchive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor2_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor2_xml_archive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor2_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor2_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor_binary_archive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor_text_archive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor_text_warchive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor_xml_archive_dll: borland-5.8.2* msvc-6.5* test_non_default_ctor_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_default_ctor_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_non_intrusive_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_intrusive_binary_archive_dll: borland-5.8.2* msvc-6.5* test_non_intrusive_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_intrusive_text_archive_dll: borland-5.8.2* msvc-6.5* test_non_intrusive_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_intrusive_text_warchive_dll: borland-5.8.2* msvc-6.5* test_non_intrusive_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_intrusive_xml_archive_dll: borland-5.8.2* msvc-6.5* test_non_intrusive_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_non_intrusive_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_null_ptr_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_null_ptr_binary_archive_dll: borland-5.8.2* msvc-6.5* test_null_ptr_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_null_ptr_text_archive_dll: borland-5.8.2* msvc-6.5* test_null_ptr_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_null_ptr_text_warchive_dll: borland-5.8.2* msvc-6.5* test_null_ptr_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_null_ptr_xml_archive_dll: borland-5.8.2* msvc-6.5* test_null_ptr_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_null_ptr_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_nvp_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_nvp_binary_archive_dll: borland-5.8.2* msvc-6.5* test_nvp_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_nvp_text_archive_dll: borland-5.8.2* msvc-6.5* test_nvp_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_nvp_text_warchive_dll: borland-5.8.2* msvc-6.5* test_nvp_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_nvp_xml_archive_dll: borland-5.8.2* msvc-6.5* test_nvp_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_nvp_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_object_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_object_binary_archive_dll: borland-5.8.2* msvc-6.5* test_object_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_object_text_archive_dll: borland-5.8.2* msvc-6.5* test_object_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_object_text_warchive_dll: borland-5.8.2* msvc-6.5* test_object_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_object_xml_archive_dll: borland-5.8.2* msvc-6.5* test_object_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_object_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_optional_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_optional_binary_archive_dll: borland-5.8.2* msvc-6.5* test_optional_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_optional_text_archive_dll: borland-5.8.2* msvc-6.5* test_optional_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_optional_text_warchive_dll: borland-5.8.2* msvc-6.5* test_optional_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_optional_xml_archive_dll: borland-5.8.2* msvc-6.5* test_optional_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_optional_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_polymorphic_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_polymorphic_binary_archive_dll: borland-5.8.2* msvc-6.5* test_polymorphic_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_polymorphic_text_archive_dll: borland-5.8.2* msvc-6.5* test_polymorphic_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_polymorphic_text_warchive_dll: borland-5.8.2* msvc-6.5* test_polymorphic_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_polymorphic_xml_archive_dll: borland-5.8.2* msvc-6.5* test_polymorphic_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_polymorphic_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_primitive_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_primitive_binary_archive_dll: borland-5.8.2* msvc-6.5* test_primitive_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_primitive_text_archive_dll: borland-5.8.2* msvc-6.5* test_primitive_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_primitive_text_warchive_dll: borland-5.8.2* msvc-6.5* test_primitive_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_primitive_xml_archive_dll: borland-5.8.2* msvc-6.5* test_primitive_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_primitive_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_private_ctor: borland-5.8.2* borland-5.8.2* msvc-6.5* test_private_ctor_dll: borland-5.8.2* msvc-6.5* test_recursion_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_recursion_binary_archive_dll: borland-5.8.2* msvc-6.5* test_recursion_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_recursion_text_archive_dll: borland-5.8.2* msvc-6.5* test_recursion_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_recursion_text_warchive_dll: borland-5.8.2* msvc-6.5* test_recursion_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_recursion_xml_archive_dll: borland-5.8.2* msvc-6.5* test_recursion_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_recursion_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_registered_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_registered_binary_archive_dll: borland-5.8.2* msvc-6.5* test_registered_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_registered_text_archive_dll: borland-5.8.2* msvc-6.5* test_registered_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_registered_text_warchive_dll: borland-5.8.2* msvc-6.5* test_registered_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_registered_xml_archive_dll: borland-5.8.2* msvc-6.5* test_registered_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_registered_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_reset_object_address: borland-5.8.2* borland-5.8.2* test_reset_object_address_dll: borland-5.8.2* test_set_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_set_binary_archive_dll: borland-5.8.2* msvc-6.5* test_set_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_set_text_archive_dll: borland-5.8.2* msvc-6.5* test_set_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_set_text_warchive_dll: borland-5.8.2* msvc-6.5* test_set_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_set_xml_archive_dll: borland-5.8.2* msvc-6.5* test_set_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_set_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_132_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_132_binary_archive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_132_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_132_text_archive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_132_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_132_text_warchive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_132_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_132_xml_archive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_132_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_132_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_binary_archive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_text_archive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_text_warchive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_xml_archive_dll: borland-5.8.2* msvc-6.5* test_shared_ptr_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_shared_ptr_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_simple_class_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_binary_archive_dll: borland-5.8.2* msvc-6.5* test_simple_class_ptr_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_ptr_binary_archive_dll: borland-5.8.2* msvc-6.5* test_simple_class_ptr_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_ptr_text_archive_dll: borland-5.8.2* msvc-6.5* test_simple_class_ptr_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_ptr_text_warchive_dll: borland-5.8.2* msvc-6.5* test_simple_class_ptr_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_ptr_xml_archive_dll: borland-5.8.2* msvc-6.5* test_simple_class_ptr_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_ptr_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_simple_class_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_text_archive_dll: borland-5.8.2* msvc-6.5* test_simple_class_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_text_warchive_dll: borland-5.8.2* msvc-6.5* test_simple_class_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_xml_archive_dll: borland-5.8.2* msvc-6.5* test_simple_class_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_simple_class_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_split_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_split_binary_archive_dll: borland-5.8.2* msvc-6.5* test_split_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_split_text_archive_dll: borland-5.8.2* msvc-6.5* test_split_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_split_text_warchive_dll: borland-5.8.2* msvc-6.5* test_split_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_split_xml_archive_dll: borland-5.8.2* msvc-6.5* test_split_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_split_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_tracking_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_tracking_binary_archive_dll: borland-5.8.2* msvc-6.5* test_tracking_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_tracking_text_archive_dll: borland-5.8.2* msvc-6.5* test_tracking_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_tracking_text_warchive_dll: borland-5.8.2* msvc-6.5* test_tracking_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_tracking_xml_archive_dll: borland-5.8.2* msvc-6.5* test_tracking_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_tracking_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_unregistered_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_unregistered_binary_archive_dll: borland-5.8.2* msvc-6.5* test_unregistered_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_unregistered_text_archive_dll: borland-5.8.2* msvc-6.5* test_unregistered_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_unregistered_text_warchive_dll: borland-5.8.2* msvc-6.5* test_unregistered_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_unregistered_xml_archive_dll: borland-5.8.2* msvc-6.5* test_unregistered_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_unregistered_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_variant_binary_archive: msvc-6.5* test_variant_binary_archive_dll: msvc-6.5* test_variant_text_archive: msvc-6.5* test_variant_text_archive_dll: msvc-6.5* test_variant_text_warchive: msvc-6.5* test_variant_text_warchive_dll: msvc-6.5* test_variant_xml_archive: msvc-6.5* test_variant_xml_archive_dll: msvc-6.5* test_variant_xml_warchive: msvc-6.5* test_variant_xml_warchive_dll: msvc-6.5* test_vector_binary_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_vector_binary_archive_dll: borland-5.8.2* msvc-6.5* test_vector_text_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_vector_text_archive_dll: borland-5.8.2* msvc-6.5* test_vector_text_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_vector_text_warchive_dll: borland-5.8.2* msvc-6.5* test_vector_xml_archive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_vector_xml_archive_dll: borland-5.8.2* msvc-6.5* test_vector_xml_warchive: borland-5.8.2* borland-5.8.2* msvc-6.5* test_vector_xml_warchive_dll: borland-5.8.2* msvc-6.5* test_void_cast: borland-5.8.2* borland-5.8.2* test_void_cast_dll: borland-5.8.2* |spirit| symbols_add_null: msvc-8.0 symbols_add_null_debug: msvc-8.0 |test| boost_check_equal_str: msvc-6.5* msvc-6.5* errors_handling_test: msvc-6.5* msvc-6.5* fixed_mapping_test: msvc-6.5* msvc-6.5* ifstream_line_iterator_test: msvc-6.5* msvc-6.5* output_test_stream_test: msvc-6.5* msvc-6.5* parameterized_test_test: msvc-6.5* msvc-6.5* prg_exec_fail3: cw-9.4 msvc-8.0 result_report_test: msvc-6.5* msvc-6.5* test_case_template_test: msvc-6.5* msvc-6.5* test_tools_test: msvc-6.5* msvc-6.5* |thread| test_tss_lib: msvc-8.0 |wave| test_re2c_lexer: hp_cxx-71_006_tru64 test_slex_lexer: hp_cxx-71_006_tru64 testwave: msvc-8.0 testwave_dll: hp_cxx-71_006_tru64 sun-5.8

1 0

[Boost-bugs] [ boost-Bugs-1649025 ] BOOST_IOSTREAMS_HAS_LSEEK64 on Mac OS X
by SourceForge.net 31 Jan '07

31 Jan '07

Bugs item #1649025, was opened at 2007-01-31 16:35 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=1649025&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 Private: No Submitted By: Gareth Sylvester-Bradley (gmsb) Assigned to: Nobody/Anonymous (nobody) Summary: BOOST_IOSTREAMS_HAS_LSEEK64 on Mac OS X Initial Comment: Mac OS X has 64-bit file offsets but uses the standard lseek call rather than lseek64. The following is one fix for file_descriptor::seek() but there's no doubt a better way... Change lines 203 and 218 from: 203 #ifndef BOOST_IOSTREAMS_HAS_LSEEK64 204 if ( off > integer_traits<long>::const_max || 205 off < integer_traits<long>::const_min ) 206 { 207 throw BOOST_IOSTREAMS_FAILURE("bad offset"); 208 } 209 #endif 210 211 stream_offset result = 212 #ifdef BOOST_IOSTREAMS_HAS_LSEEK64 213 lseek64 214 #else 215 lseek 216 #endif 217 ( pimpl_->fd_, 218 #ifdef BOOST_IOSTREAMS_HAS_LSEEK64 219 off, 220 #else 221 static_cast<long>(off), 222 #endif 223 way == BOOST_IOS::beg ? 224 SEEK_SET : 225 way == BOOST_IOS::cur ? 226 SEEK_CUR : 227 SEEK_END ); To: 203 #if !defined(BOOST_IOSTREAMS_HAS_LSEEK64) && !defined(__MACH__) ... 218 #if defined(BOOST_IOSTREAMS_HAS_LSEEK64) || defined(__MACH__) ... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1649025&group_… ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

[Boost-bugs] [ boost-Patches-1452698 ] iostreams // file_descriptor::seek BUG on files > 4 GB
by SourceForge.net 31 Jan '07

31 Jan '07

Patches item #1452698, was opened at 2006-03-17 20:08 Message generated for change (Comment added) made by gmsb You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=307586&aid=1452698&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 Private: No Submitted By: Nobody/Anonymous (nobody) Assigned to: Jonathan Turkanis (turkanis) Summary: iostreams // file_descriptor::seek BUG on files > 4 GB Initial Comment: Boost 1.33.1 iostreams library file_descriptor::seek() method file_descriptor.cpp, line 198 Win32, Visual C++ 2005 The method seek() can return a wrong value of a newly set offset when the offset is further than 4 GB. //--------------- return offset_to_position((lDistanceToMoveHigh << 32) + dwResultLow); //--------------- The code above has a bug. On 32-bit systems the expression (lDistanceToMoveHigh << 32) equals to lDistanceToMoveHigh. The correct code should first cast the lDistanceToMoveHigh variable to 64-bit and then shift it. //--------------- return (static_cast<boost::intmax_t> lDistanceToMoveHigh) << 32) + dwResultLow; //--------------- Sergey Kolodkin ---------------------------------------------------------------------- Comment By: Gareth Sylvester-Bradley (gmsb) Date: 2007-01-31 16:27 Message: Logged In: YES user_id=271167 Originator: NO There's also a bug 3 lines earlier. Windows GetLastError() shouldn't be checked unless dwResultLow is INVALID_SET_FILE_POINTER. Replace: 195 if (::GetLastError() != NO_ERROR) { 196 throw detail::bad_seek(); 197 } else { 198 return offset_to_position((lDistanceToMoveHigh << 32) + dwResultLow); 199 } 200 } With: 195 if (INVALID_SET_FILE_POINTER == dwResultLow && ::GetLastError() != NO_ERROR) { 196 throw detail::bad_seek(); 197 } else { 198 return offset_to_position((static_cast<stream_offset>(lDistanceToMoveHigh) << 32) + dwResultLow); 199 } 200 } ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=307586&aid=1452698&group_… ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

[Boost-bugs] [ boost-Bugs-1649008 ] Seekable file_descriptor_source /sink?
by SourceForge.net 31 Jan '07

31 Jan '07

Bugs item #1649008, was opened at 2007-01-31 16:17 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=1649008&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 Private: No Submitted By: Gareth Sylvester-Bradley (gmsb) Assigned to: Nobody/Anonymous (nobody) Summary: Seekable file_descriptor_source /sink? Initial Comment: The supplied file_descriptor_source and file_descriptor_sink are not seekable. Is this an oversight? For file_descriptor_source, all that appears to be needed is changing from: struct category : public source_tag, closable_tag { }; to: struct category : input_seekable, device_tag, closable_tag { }; and adding: using file_descriptor::seek; Similarly for file_descriptor_sink. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1649008&group_… ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

[Boost-bugs] [ boost-Bugs-1649001 ] Incorrect usage of bad_write(), bad_seek() etc.
by SourceForge.net 31 Jan '07

31 Jan '07

Bugs item #1649001, was opened at 2007-01-31 16:09 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=1649001&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 Private: No Submitted By: Gareth Sylvester-Bradley (gmsb) Assigned to: Nobody/Anonymous (nobody) Summary: Incorrect usage of bad_write(), bad_seek() etc. Initial Comment: Incorrect usage of bad_write(), bad_seek() etc. defined in <boost/iostreams/detail/error.hpp>. In several places in the rest of Boost.Iostreams where these functions are used to construct exceptions, the exceptions aren't then thrown. This is probably unintended. E.g. restrict.hpp, direct_adapter.hpp and range_adapter.hpp. P.S. Should there be a tracker category of "iostreams"? ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1649001&group_… ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0