July 2006 - Boost - lists.preview.boost.org

foo(bar, a, b) written as bar.foo(a, b) [was SuperString v2 in the vault]
by me22 11 Jul '06

11 Jul '06

On 7/11/06, Andy Little <andy(a)servocomm.freeserve.co.uk> wrote: > "me22" <me22.ca(a)gmail.com> wrote > > If foo(bar,a,b) could always be written as bar.foo(a,b) > > > That would break some of my code FWIW > In what way? Would it still break with a new overload resolution rule that a member function called as a member function is a better match than a non-member function called as a member function? ~ Scott McMurray

2 1

Error in boost::function doc ?
by Philippe Vaucher 11 Jul '06

11 Jul '06

http://boost.org/doc/html/function/faq.html#id2699448 It looks like the code example of point 3 is wrong, as it's twice the same... Philippe

1 0

[addressof] reference to function?
by Shunsuke Sogame 11 Jul '06

11 Jul '06

Hi, void foo() { } void test() { boost::addressof(::foo); } doesn't compile under GCC 3.4.4. What happens? -- Shunsuke Sogame

2 2

[system] - system_error
by Oliver.Kowalke＠qimonda.com 11 Jul '06

11 Jul '06

Hi, shouldn't system_error::const char * what() const contain throw ()? regards, Oliver

6 8

Re: [boost] [Review][PQS] Review deadline
by Michael Fawcett 11 Jul '06

11 Jul '06

On 6/8/06, Andy Little <andy(a)servocomm.freeserve.co.uk> wrote: > Hi Michael, > > > -----Original Message----- > > From: Michael Fawcett [mailto:michael.fawcett@gmail.com] > > Sent: 08 June 2006 15:49 > > To: andy(a)servocomm.freeserve.co.uk > > Subject: Re: [boost] [Review][PQS] Review deadline > > > > A goal of pqs is to be useable in that sort of situation. To be > > really effective > > > it will need a lot of supporting classes ( matrix, vector, quat) though. > > > > Did you plan on using Boost::uBLAS for this, along with > > Boost::quaternion? This seemed tangential enough to the PQS > > discussions that I didn't want to clutter up the mailing list, but I > > am defeinitely curious. > > FWIW Its extremely important. I think it would be worth asking on the list. > I dont think many people realise! I have discussed it in various replies but > no one has asked me the direct question: > > Neither of those libraries above will work with other than numeric types > AFAIK. IOW where T is a type (simplifying and ignoring e.g floating point > promotion etc), it must satisfy: > > T result = T() + T(); > T result = T() * T(); > > (among others) > > If T is a physical quantity, the second requirement never can be true. > Quantity Compatible library must be written something like: > > Where Op is +,-,*,/, ==, etc ... > > typeof( T1() Op T1() ) result = T1() Op T2(); This was a private e-mail discussion but Andy thought it important to be bruoght up on the list. FWIW, my own current matrix/vector/quaternion code does what you are asking for (with boost's help), but I had been looking at changing all my stuff over to uBLAS/quaternion for some time. As an example: blas::vec3<float> a(1, 0, 0); blas::vec3<double> b(0, 1, 0); blas::vec3<double> result = cross_product(a, b); // returns vec3<double>! I accomplished this using what I think never became a boost library (from the sandbox maybe? Did MPL::transform consume this functionality?) result_of which was used something like: template <typename LHS, typename RHS> typename boost::result_of_plus<LHS, RHS>::type operator +(const LHS &lhs, const RHS &rhs) { typename boost::result_of_plus<LHS, RHS>::type nrv(lhs); nrv += rhs; return nrv; } I would hope some solution could be found where you didn't have to implement all new matrix/vector/quaternion classes and their corresponding functionality, rather, just ensure the return type from the operation follows normal promotion rules (this is actually what I thought the promotion traits library was going to do). --Michael Fawcett

24 132

[Boost-bugs] [ boost-Bugs-1520489 ] significantly different timer class behaviour on Win and Lin
by SourceForge.net 11 Jul '06

11 Jul '06

Bugs item #1520489, was opened at 2006-07-11 15:41 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=1520489&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: timer Group: None Status: Open Resolution: None Priority: 5 Submitted By: Sukhonosenko Kirill (skv) Assigned to: Beman Dawes (beman_dawes) Summary: significantly different timer class behaviour on Win and Lin Initial Comment: clock() function used in implementation of boost::timer behaves differently on Windows (MS XP Pro, MSVC 7.1) and Linux (FC3, gcc 3.4.2). Windows: clock measures elapsed time (wall clock). And this comply with boost::timer interface description. Example: boost::timer t; double sec = 1; ::Sleep(sec*1000); double d = t.elapsed(); assert(abs(d-sec) < 0.1); Linux: clock() function returnes process CPU time. The above code will fail. This difference makes boost::timer a bad candidate for cross-platform development. I can propose alternative implementation with simalar right behaviour. ------------------------------------------ #ifndef _Timer_h_dlfkgh9384ytdflg9045uhy0823475mgd385uyn0v68uy c0n4c5 #define _Timer_h_dlfkgh9384ytdflg9045uhy0823475mgd385uyn0v68uy c0n4c5 #include <time.h> namespace Timer { /* Measures intervals of wall clock time, but not process CPU time. Resolution & accuracy depends on time function abilities. */ class timer { public: // setup reference time point timer() { restart(); } // reset reference time point void restart() { ::time(&m_ref_time); } // report difference between current time and reference point double elapsed() const { time_t curr_time; ::time(&curr_time); double diff = difftime (curr_time, m_ref_time); return diff; } // forbidden private: timer(const timer &); timer & operator=(const timer &); private: time_t m_ref_time; }; } // namespace Timer #endif //_Timer_h_dlfkgh9384ytdflg9045uhy0823475mgd385 uyn0v68uyc0n4c5 ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1520489&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

Re: [boost] [BGL] A Question
by Wang Weiwei 11 Jul '06

11 Jul '06

>> Hello everyone, > >Hello. > >> I'm starting to make use of the BGL in one of my >> projects. It was learnt from the documentation >> that there were basically 2 forms of >> representations of a graph: >> - adjacency-list, and >> - adjacency-matrix >> The latter is normally more suitable for denser >> graphs. >> >> My question is, Is ther any function that can >> shift the graph from between these two forms of >> data structure? > >Try boost::copy_graph. ><http://www.boost.org/libs/graph/doc/copy_graph.html> > > HTH, > Cromwell D. Enage > > Thanks.

1 0

[BGL] A Question
by Wang Weiwei 11 Jul '06

11 Jul '06

Hello everyone, I'm starting to make use of the BGL in one of my projects. It was learnt from the documentation that there were basically 2 forms of representations of a graph: - adjacency-list, and - adjacency-matrix The latter is normally more suitable for denser graphs. My question is, Is ther any function that can shift the graph from between these two forms of data structure? Thanks Max

2 1

Boost.Test: automatic test cases inside a static library
by Joe Van Dyk 11 Jul '06

11 Jul '06

I'm trying to get something like the following to work: test_case1.cpp and test_cast2.cpp are compiled into a static library (libtest.a). They use BOOST_AUTO_TEST_CASE for test cases. test_main.cpp contains: #define BOOST_AUTO_TEST_MAIN #include <boost/test/auto_unit_test.hpp> However, when I compile test_main.cpp and link libtest.a into it, none of the test cases run. The test cases do run if test_main.cpp #includes test_case1.cpp and test_case2.cpp, but I'd rather not do that if it's not necessary. Any ideas? Thanks! Joe Van Dyk

3 8

Re: [boost] Please don't define a fake "operator <" justfororderedcontainers
by Sohail Somani 11 Jul '06

11 Jul '06

> -----Original Message----- > From: boost-bounces(a)lists.boost.org > [mailto:boost-bounces@lists.boost.org] On Behalf Of me22 > Sent: Monday, July 10, 2006 5:07 PM > To: boost(a)lists.boost.org > Subject: Re: [boost] Please don't define a fake "operator <" > justfororderedcontainers > > On 7/10/06, Sohail Somani <s.somani(a)fincad.com> wrote: > > Isn't shared_ptr<T> is supposed to be like T* which has ordering? In > > this case, it's a non-issue no? > > > My understanding is that < is only well-defined for pointers into the > same array and that < between to arbitrary pointers is otherwise some > sort of UB. > > It's less<T*> that provides a full ordering for arbitrary T*s. (By > specialising std::less.) This is true.

1 0

Re: [boost] Please don't define a fake "operator <" justfor orderedcontainers
by Sohail Somani 11 Jul '06

11 Jul '06

> -----Original Message----- > From: boost-bounces(a)lists.boost.org > [mailto:boost-bounces@lists.boost.org] On Behalf Of Peter Dimov > Sent: Monday, July 10, 2006 2:58 PM > To: boost(a)lists.boost.org > >>> Looking at <http://www.boost.org/libs/smart_ptr/shared_ptr.htm>, > >>> i.e. the Boost version, it seems that the authors fell into the > >>> "fake operator <" trap. If any of those authors are reading this, > >>> can you tell us why you added the fake operator instead of a > >>> custom-named comparison class? > >> > >> shared_ptr defines operator< for map<shared_ptr<>, V> to work. What > >> do you mean by "fake"? > > > > I know what I did, but this still doesn't tell me what "fake" > means in this > context. What objective properties make an operator< "fake"? Isn't shared_ptr<T> is supposed to be like T* which has ordering? In this case, it's a non-issue no?

2 1

Comment on operator <
by Sean Parent 10 Jul '06

10 Jul '06

operator < should be defined anytime an objects of type can be totally ordered and the comparison has average constant complexity with worst case complexity proportional to the area of the object (the complexity guarantees are why we don't provide a less-than- operator for hash containers - we could just as we do for vector, but the cost is prohibitive). In the absence of a better choice, a lexicographical comparison of the members should be provided. The ordering must be total because !(a < b) && !(b < a) => a == b [trichotomy law]. The reason for the allowing comparison functions is containers is to allow for other orderings, including partial ones. Recommended reading - lecture 7 of <http://www.stepanovpapers.com/ notes.pdf>. Alex asserts in his notes that even when the less than operator has no meaning other than to provide a runtime ordering (such as for iterators on linked lists) it should be provided. In the case of linked list iterator - I also think that in the presence of increment that a < b should imply that [a, b) is a valid range and b is reachable from a through increment (which would not be the case using pointer ordering for linked list iterators). Limiting the application of operator < to relationships which survive serialization (that is, ones which do not depend on accidental relationships of object identity) would be my preference. Sean

1 0

Shmem named semaphore issue on linux
by Ryan, Paul L. 10 Jul '06

10 Jul '06

Ion, The named semaphore on linux seems to never increment the initial count when creating. Looking into the posix implementation a sem_init is never called if create is given and it just returns true. bool operator()(const shared_memory::segment_info_t * info, bool created) const { mp_sem = reinterpret_cast<sem_t *>(info->get_user_ptr()); switch(m_create){ case open_only: mp_sem = reinterpret_cast<sem_t *>(info->get_user_ptr()); break; case open_or_create: if(created){ if(sem_init(mp_sem, 1, m_initialCount) == 0){ return true; } else{ mp_sem = 0; return false; } } break; case create_only: return true; break; default: return false; break; } return true; } I think this method should be bool operator()(const shared_memory::segment_info_t * info, bool created) const { mp_sem = reinterpret_cast<sem_t *>(info->get_user_ptr()); switch(m_create){ case open_only: mp_sem = reinterpret_cast<sem_t *>(info->get_user_ptr()); break; case open_or_create: if(created){ if(sem_init(mp_sem, 1, m_initialCount) == 0){ return true; } else{ mp_sem = 0; return false; } } break; case create_only: if(sem_init(mp_sem, 1, m_initialCount) == 0){ return true; } else{ mp_sem = 0; return false; } break; default: return false; break; } return true; } Paul Ryan SAIC - Tactical Systems

2 1

Re: [boost] Boost.Test: automatic test cases inside a static library
by Sohail Somani 10 Jul '06

10 Jul '06

> -----Original Message----- > From: boost-bounces(a)lists.boost.org > [mailto:boost-bounces@lists.boost.org] On Behalf Of Gennadiy Rozental > Sent: Monday, July 10, 2006 2:47 PM > To: boost(a)lists.boost.org > Subject: Re: [boost] Boost.Test: automatic test cases inside > a static library > >> Hm, I've never heard of that option. I'll look into it. > > > > Any chance you could tell me what flags the GNU GCC > compiler would do > > the trick? I can't find anything that resembles "avoid > unused static > > files discard". > > Looks like these may work: > > -fkeep-static-consts > -fno-toplevel-reorder > -fno-unit-at-a-time (?) On Linux, I believe there is also a linker flag -Wl,-whole-archive. There is also an equivalent on solaris but it escapes me...

1 0

Trying to get Boost.Python to build on Solaris 10 using SUN Studio 11
by Timo Geusch 10 Jul '06

10 Jul '06

Good afternoon, I've so far managed to make decent progress on building RC 1.34's Boost.Python on this OS/Compiler combination using jam v1 with the patches I submitted earlier. Unfortunately I now seem to have hit a brick wall in file pickle_support.cpp. The offending lines of code are line 57, which reads: static object result(&instance_reduce); If I replace this line with plain static object result; the file will compile, however the above line produces the following error message from the compiler: sunpro-C++-action bin/boost/libs/python/build/libboost_python.so/sunpro/debug/pickle_support.o "./boost/python/converter/arg_to_python.hpp", line 215: Error: Could not find boost::python::converter::detail::arg_to_python_base::arg_to_python_base(boost::python::tuple(*)(boost::python::api::object), const boost::python::converter::registration) to initialize base class. "./boost/python/object_core.hpp", line 309: Where: While instantiating "boost::python::api::object_base_initializer<boost::python::tuple(*)(boost::python::api::object)>(boost::python::tuple(*const)(boost::python::api::object)&)". "./boost/python/object_core.hpp", line 309: Where: Instantiated from non-template code. 1 Error(s) detected I think I'm missing something obvious here but I'm not familiar enough with the Boost.Python code to work out what it is. I'd appreciate if someone who is would be able to have a closer look as this is the last file that doesn't build here. Regards, Timo

2 2

[function] Internal compiler error on VC 7.1 with default parameter
by Allen Bierbaum 10 Jul '06

10 Jul '06

The following code snippet: #include <boost/function.hpp> class TestClass { public: void doSomethingDefault(boost::function2<void, bool, bool> cb=boost::function2<void,bool,bool>()) {;} }; int main() { TestClass test; test.doSomethingDefault(); return 1; } Gives an internal compiler error on vc 7.1: cl /nologo /Od /Ob0 /Z7 /W1 /MDd /EHsc /GR /TP /Ic:\Software\boost\boost_1_33_1 /DWIN32 /D_WINDOWS /D_DEBUG /c test.cpp /Fotest.obj test.cpp test.cpp(6) : fatal error C1001: INTERNAL COMPILER ERROR (compiler file 'msc1.cpp', line 2701) Please choose the Technical Support command on the Visual C++ Help menu, or open the Technical Support help file for more information scons: *** [test.obj] Error 2 scons: building terminated because of errors. (line 6 is the method declaration with the default value). If I remove the use of the default value, the code compiles with no errors. I know this code compiles and works correctly on gcc 4.0 and I believe I have used it with older versions of gcc. Is there something incorrect with my usage? If not, is this an known issue? I searched through the bug tracker and the mailing lists but did not see anything about this bug. Does anyone have any ideas on how to work around this bug? Thanks, Allen

2 2

Preliminary release of clone_ptr library
by Kevin Spinar 10 Jul '06

10 Jul '06

(this will be the first and last "progress update" message until it comes time to summit for a review to boost) For better or for worse (it seems there's much debate on clone_ptr) I have started on the clone_ptr library and have a working preliminary clone_ptr class. Details can be found at the following URL and I will update the website as my work progresses: http://www.peltkore.net/~alipha/clone_ptr.html If anyone is interested in receiving email updates or would like to collaborate on the project (which could just simply be someone I could bounce ideas off of) email me. To contribute some to the discussion: It has been noted that the Boost.pointer_container library could be used instead of having a container of clone_ptr. However, what if you want to use non-standard containers? Yes, Boost.pointer_container has ptr_*_adapter classes, but they can't be used for containers which don't follow the container requirements set by the C++ standard. Secondly, even if a third party container class follows the container requirements, it is still unnecessary work to create your own pointer container. Consider this generic tree class: http://www.aei.mpg.de/~peekas/tree/ http://www.aei.mpg.de/~peekas/tree/doxygen/html/functions.html There's dozens of functions in the tree class that would have to be wrapped in a ptr_tree class that the boost::ptr_sequence_adapter class doesn't wrap for you. Compare this to: tree<clone_ptr<T> > foo; The clone_ptr solution seems a lot easier. The same argument could be made with boost::graph, etc. Also, I've decided to give clone_ptr clone-on-write semantics instead of clone-on-copy. This should alleviate any concerns about performance when inserting into a std::vector or copying containers (copying clone_ptr would give similar performance to boost::shared_ptr). Kevin Spinar

4 10

Exception library proposal update
by Emil Dotchevski 10 Jul '06

10 Jul '06

Hello, As a result of today's discussion, I added a function get_exception_info to the interface of the proposed Boost Exception library: template <class T> exception_info * get_exception_info( T & ); It can be used in a catch( T & ) block, to check if the exception object has a sub-object of class exception_info. I updated the documentation, source code and tests accordingly, all available at: http://www.revergestudios.com/exception/exception.htm Click "download" to get the code and the tests, complete with boost build jamfiles. Thanks, Emil

5 15

Quickbook :/
by Philippe Vaucher 10 Jul '06

10 Jul '06

Hello, I've received many valuable advices on how to set quickbook however I still can't make it work. My platform is windows, I managed to have bjam build quickbook.exe, but from there I can't build any of the docs like in \boost\libs\type_traits\doc. For convenience's sakes I put quickbook in C:\WINNT so it's easier to read for you, but no matter which path I put bjam can't find quickbook I have user-config.jam with : using xsltproc ; using boostbook : C:/boost/tools/build/v2/docbook-xsl-1.70.1/ : C:/boost/tools/build/v2/docbook-xml-4.2/ ; using doxygen : C:/Progra~1/doxygen/bin/ ; using quickbook : C:/WINNT/quickbook.exe ; I tried to set "using quickbook;" but then it start building teh whole boost, I tried "using quickbook : C:/WINNT/ ;" but same problem. When I run "bjam --v2" in \boost\libs\functional\hash\doc, here's the output : C:\boost\libs\functional\hash\doc>bjam --v2 warning: Python location is not configured warning: the Boost.Python library won't be built Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (and if you don't know what ICU is then you probably don't need it). warning: toolset quickbook initialization: can't find tool C:/WINNT/quickbook.exe warning: initialized from warning: toolset quickbook initialization: can't find tool C:/WINNT/quickbook.exe warning: initialized from ...found 19 targets... ...updating 3 targets... quickbook.quickbook-to-boostbook..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug\hash.xml Generating Output File: C:\boost\libs\functional\hash\doc\hash.xml xslt-xsltproc ..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug\hash.docbook warning: failed to load external entity "../../../../bin.v2/libs/functional/hash/doc/msvc/debug/hash.xml" unable to parse ..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug\hash.xml set XML_CATALOG_FILES=..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug/catalog.xml "xsltproc" --xinclude -o "..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug\hash.docbook" "C:\boost\tools\boostbook\xsl\docbook.xsl" "..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug\hash.xml" ...failed xslt-xsltproc ..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug\hash.docbook... ...skipped <phtml>HTML.manifest for lack of <p..\..\..\..\bin.v2\libs\functional\hash\doc\msvc\debug>hash.docbook... ...failed updating 1 target... ...skipped 1 target... ...updated 1 target... If anyone has a clue about what's going on it'd be nice. Philippe

5 7

[test] regressions for vc-6_5
by Joaquín Mª López Muñoz 10 Jul '06

10 Jul '06

The parameterized_test_test test for Boost.Test is failing for MSVC 6.5, as shown at http://tinyurl.com/lek2w . This can be fixed by just reverting a workaround on boost/test/parameterized_test.hpp which was removed Jan 28th 2006 (v1.6 --> v.1.7) Is it OK to bring the workaround back? If given permission, I can do that myself. Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 2

Re: [boost] [shmem] error while closing the segment
by Ion Gaztañaga 10 Jul '06

10 Jul '06

> yes, the destructor performs two destroy calls in front of the close > call (see below). But it should not give a damn about, destroyed is > destroyed. The call of close should only delete objects that still > exists. I've printed out the name of the named shared object which > triggers the error (it's "shmDefinition"). If I comment out the two > lines which destroy my named shared objects in the destructor, it works. > Hmm, I don't know why it not work when I destroy the objects by hand. > But definitely it should not be done twice. You are right. But this indicates that you somehow have tried to erase the same object (in this case ("shmDefinition")) twice (maybe another process?). It should assert when using a pointer (because otherwise that could corrupt all the segment) when deleting the object (a debug protection like when trying to "operator delete" the same ptr two times). However, when destroying by name, a bool is returned, saying if the object has been destroyed or not (well the only problem is to found the name in the name-object index). So it should return false in the second case, instead of asserting. The only problem is that both name and pointer delete operation uses the same code, and only one should assert. Thanks for finding this! Please, erase that assertion from your code. I will do the same in Sandbox-CVS, while I find some other solution. Regards, Ion

1 0

Minor exception lib changes
by Emil Dotchevski 10 Jul '06

10 Jul '06

Hello, I am contemplating a few minor changes to the exception library and I wanted to get some input before I implement them. First, I think that the function template failed should be renamed. Somehow, boost::failed doesn't feel right. Since English is not native to me, I really need help with this one. Second, I'm thinking of renaming wrap_string/get_string. Those names did make sense in the implementation I've been using in my own code, because the whole thing was in namespace except. Clearly, boost::get_string doesn't suggest that it's part of the exception library. I'm thinking of renaming them both to exception_string. So, stuffing a string at the time of the throw would look like this: throw failed<foo>() << exception_string<tag>("value"); To extract the string from an exception_info object, you'd do this: exception_info & xi; if( std::string const * s = exception_string<tag>(xi) ) .... Similarly, it makes sense to rename both wrap_errno/get_errno to exception_errno. Thanks for everyone's input! --Emil

2 1

boost.shmem bugs
by Dmitry Smirnov 10 Jul '06

10 Jul '06

1. If you hasn't semaphore support in your glibc libraries (in my case it was without TLS) "sem_open" function will return (sem_t*)0 instead of (sem_t*)-1 as it is described in my system header "semaphore.h". Plese correct file "shared_memory.hpp" for using SEM_FAILED macros. Defines in "semaphore.h": /* Value returned if `sem_open' failed. */ #define SEM_FAILED ((sem_t *) 0) /* Maximum value the semaphore can have. */ #define SEM_VALUE_MAX (2147483647) 2. Unclear sinuation. I wrote bugous example program where I write some data in shared_message_queue in infinity cycle without any timeout. If you stop program by Ctrl+C "boost_shmem_shm_global_mutex" stays in locked state so program will not work from second start. If you include some timeout in cycle then program will work properly. Thank you for good job! \o/ Best Regards, Dmitry Smirnov Saint-Petersburg, Russia Slackware 10.2 gcc-3.4.4 kernel 2.6.13 glibc-2.3.6

2 4

[function] Conflict with Boost.TR1
by John Maddock 10 Jul '06

10 Jul '06

Unfortunately another TR1 problem, in boost/function/detail/prolog.hpp we currently have: # include <cassert> # include <algorithm> # include <functional> // unary_function, binary_function The inclusion of <functional> here causes a circular dependency between the TR1 version of <functional> and <boost/function.hpp> leading to a long list of compile errors: simply including <boost/function.hpp> on it's own is enough to trigger these when the TR1 headers are in the include path. The only (ugly) fix I can offer at present is to use <boost/config/no_tr1/functional.hpp> in place of <functional>. Better suggestions, or policies welcomed, this is starting to become a maintenance nightmare.... John.

1 0

tr1/test_function_tricky trickier than needed
by Peter Dimov 10 Jul '06

10 Jul '06

The only reason for test_function_tricky.cpp to not compile with boost::function is the presence of the following two lines: BOOST_STATIC_ASSERT((::boost::is_base_and_derived<std::unary_function<int, long>, std::tr1::function<long (int)> >::value)); BOOST_STATIC_ASSERT((::boost::is_base_and_derived<std::binary_function<int, char, double>, std::tr1::function<double (int,char)> >::value)); that check for the (highly questionable and not particularly relevant) requrement that function<> should derive from unary_function or binary_function in some cases. I think that these two lines need to be moved to a separate test.

2 7