- Boost - lists.boost.org

RE: [boost] Re: Customer Friendlier Boost Installation
by Brian Braatz 06 Jun '05

06 Jun '05

> -----Original Message----- > From: boost-bounces(a)lists.boost.org [mailto:boost-bounces@lists.boost.org] > On Behalf Of David Abrahams > > I have read the MPL book now FIVE TIMES cover to cover. On the fith > > pass, I "saw it". (the mpl::for_each<> is what kicked it in for me) > > Wow, I applaud your fortitude and apologize for writing something that > took you so long to "see." [Brian Braatz Writes:] Hmm.... I believe the MPL book is an excellent body of work. The difficulty is because 1- I am either VERY smart about a subject or it is a black box. there seems to be little middle ground (this is learning style I think) 2- Metaprogramming is "training your brain" to think differently. Honestly with the subject matter being so DIFFERENT from what us normal blokes are used to, I am not sure if I could even suggest a better way to present it. Being able to write compile time code is a skill that belongs to people who have the commitment to struggle through and learn it. the TWISTED thing, is as I am getting more and more, it feels natural. Also I think I am wearing out my copy of the book- I continually go back and re-read sections and chapters. THANKS to both you and Aleksey for birthing that sucker. Lots of good information in there. TO ANYONE READING THIS BOOK: DON'T GIVE UP. KEEP AT IT. YOU WILL GET IT. You just have to read it enough that you begin to bleed through your eyes. When that happens, you know you are close......

2 1

[parameter] Patch for Tru64/CXX65
by Markus Schöpflin 06 Jun '05

06 Jun '05

Hello, attached patch fixes the current regression failure for the basics test. The code in questions contains a non-working workaround for this compiler, the patch removes this workaround as well. Ok to commit? Markus Index: basics.cpp =================================================================== RCS file: /cvsroot/boost/boost/libs/parameter/test/basics.cpp,v retrieving revision 1.8 diff -u -r1.8 basics.cpp --- basics.cpp 16 May 2005 12:58:48 -0000 1.8 +++ basics.cpp 3 Jun 2005 16:29:10 -0000 @@ -14,7 +14,6 @@ namespace test { - // A separate function for getting the "value" key, so we can deduce // F and use lazy_binding on it. template <class Params, class F> @@ -39,8 +38,13 @@ typename boost::parameter::binding< Params, index_, int - >::type i = p[index | 999]; - + >::type i = +#if BOOST_WORKAROUND(__DECCXX_VER, BOOST_TESTED_AT(60590042)) + p[test::index | 999]; +#else + p[index | 999]; +#endif + p[tester](n,v,i); return 1; Index: basics.hpp =================================================================== RCS file: /cvsroot/boost/boost/libs/parameter/test/basics.hpp,v retrieving revision 1.2 diff -u -r1.2 basics.hpp --- basics.hpp 15 May 2005 10:15:56 -0000 1.2 +++ basics.hpp 3 Jun 2005 16:29:10 -0000 @@ -88,13 +88,6 @@ { return values_t<Name,Value,Index>(n,v,i); } - -#if defined( __DECCXX_VER ) - && BOOST_WORKAROUND(__EDG_VERSION__, BOOST_TESTED_AT(245)) - // Some kind of nasty ambiguity arises here -- is there an index in - // the global namespace? - using test::index; -#endif } // namespace test

2 2

RE: [boost] Re: ICE ICE BABY! (my MPL code is ICEing)
by Brian Braatz 05 Jun '05

05 Jun '05

> >> > >>template <class type_t, class sequence_t> > >>struct copy_col_if_in_dest > >>: public if_< typename is_same< typename deref<typename > >>boost::mpl::find<sequence_t, type_t>::type >::type, type_t>, > > > > do_copy_col, > > > >>do_not_copy_col>::type > >>{ > >>}; > > This code has a problem when type_t is not in sequence_t: the > past-the-end iterator gets dereferenced. I wrote a short program to do > what you want (it's more explicit, it uses mpl::contains<>): > > #include <boost/mpl/vector.hpp> > #include <boost/mpl/if.hpp> > #include <boost/mpl/contains.hpp> > > namespace mpl = boost::mpl; > > struct do_copy_col {}; > struct dont_copy_col {}; > > template<typename T, typename Seq> > struct copy_col_if_in_dest > : mpl::if_< > mpl::contains<Seq, T>, > do_copy_col, > dont_copy_col > >::type > { > }; > > int main() > { > typedef mpl::vector<char, short, int, long> v; > do_copy_col do_copy = copy_col_if_in_dest<int, v>(); > dont_copy_col dont_copy = copy_col_if_in_dest<bool, v>(); > return 0; > } > [Brian Braatz Writes:] Wow. Many THANKS- the mpl::contains<> is REALLY what I was looking for. (Though the exercise of doing it with if_<> was a useful experience) THANK YOU ( I have been burning the morning and afternoon oil here the last few weeks trying to get a lib I am proposing ready for boost people to look at) -I am going to replace the code I had with the code you presented, and make a note of your contribution. THANKS AGAIN! Brian

1 0

problems buidling latest cvs version
by Christopher Lux 05 Jun '05

05 Jun '05

hi, bjam cancels buidling in v1 and v2 mode on the latest cvs checkout. i used bjam built from the source from the checkout. error message said: 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). C:\_development\boost_cvs\tools/build/v1\boost-base.jam:1966: in subvariant-target *** argument error * rule library-file ( target : sources + ) * called with: ( <@boost!libs!iostreams!build\libboost_iostreams.lib\vc-7_1\debug\runtime-link-static>libboost_iostreams-vc71-sgd-1_33.lib : ) * missing argument sources C:\_development\boost_cvs\tools/build/v1\boost-base.jam:119:see definition of rule 'library-file' being called C:\_development\boost_cvs\tools/build/v1\boost-base.jam:2403: in stage Jamfile:195: in load-jamfiles C:\_development\boost_cvs\tools\build\v1\bootstrap.jam:21: in boost-build C:\_development\boost_cvs\boost-build.jam:16: in module scope

1 0

[typeof] Small feature request
by Pavel Chikulaev 05 Jun '05

05 Jun '05

Hi all, I've got a small feature request to type_of library: 1) BOOST_TYPEOF_CONSTANT_COMPLEXITY Macro is defined if typeof can't evaluated with constant complexity 2) BOOST_TYPEOF_LINEAR_COMPLEXITY Macro is defined if typeof can't evaluated with linear complexity -- Pavel Chikulaev

2 2

[1.33] recent check-ins break date-time serialization tests
by Jeff Garland 05 Jun '05

05 Jun '05

I didn't look at the check-in logs, but in the last 24-48 hours a host of previously passing date-time serialization tests have started to fail. In particular code warrior 9-4, various gcc, and vc7 were previously passing. I haven't made any check-ins to date-time, so unless someone else has I expect the problem is elsewhere... summary: http://tinyurl.com/ayl5j detailed example: http://tinyurl.com/e2pa5 Jeff

4 4

boost::lock_error - documentation bug
by Paul Groke 05 Jun '05

05 Jun '05

Hi everybody, there seems to be a bug in the documentation (or implementation?) of boost::lock_error. The documentation (http://www.boost.org/doc/html/lock_error.html) describes lock_error as: Synopsis: class lock_error : public std::logical_error { ... Whereas them implementation (1.32.0, thread/exceptions.hpp) declares boost::lock_error as derived from boost::thread_exception which is derived from std::exception and nothing more. Even assuming std::logic_error was ment (there is no class std::logical_error ) I'd say that's a documentation bug. I haven't looked through all the other exception classes defined by boost::thread ... maybe someone should :-) bye, Paul

1 0

RE: ICE ICE BABY! (my MPL code is ICEing)
by Brian Braatz 05 Jun '05

05 Jun '05

HAHAHAHA Nevermind. copy_col_if_in_dest<boost::mpl::at<src::typelist, boost::mpl::int_<0> >::type,src::typelist>(); instead of being src::typelist (which is a ref to the param of type src_t) it needs to be src_t::typelist. it was indeed something stupid- I am undecided as to which is more stupid. The fact that I didn't catch that or the fact that MSVC ICE'd on it...... sorry for bugging you guys > -----Original Message----- > From: Brian Braatz > Sent: Saturday, June 04, 2005 2:38 PM > To: 'boost(a)lists.boost.org' > Subject: ICE ICE BABY! (my MPL code is ICEing) > > Sitting in the backyard. > > Beautiful day, got up at 3am to get something figured out (for proposed > boost profiler \ introspection library). > > Good news is I figured out the MPL code needed to accomplish my goal. > After typing in an incremental version of the code, I got the dreaded > internal compiler error (ICE). > > I am chewing on this- and thought I would post it so to provide an > opportunity for someone to show how incredibly stupid I am because I did > something that was obviously wrong. (you can even rub my nose in it if you > like :) ) > > here is the code: > > I am tearing it apart\deconstructing to try to figure out what is kicking > in the ICE. (I am also going to try it with an eval_if (which I should be > using anyway) just to see what happens) > > I am using boost 1.32 and vc7.1 > > struct do_copy_col > { > typedef do_copy_col type; > do_copy_col() > { > std::cout << "do_copy_col" << std::endl; > } > // todo- implement operator()(dest,src) > }; > struct do_not_copy_col > { > typedef do_not_copy_col type; > do_not_copy_col() > { > std::cout << "do_not_copy_col" << std::endl; > } > > // todo- implement operator()(dest,src) > }; > > template <class type_t, class sequence_t> > struct copy_col_if_in_dest > : public if_< typename is_same< typename deref<typename > boost::mpl::find<sequence_t, type_t>::type >::type, type_t>, do_copy_col, > do_not_copy_col>::type > { > }; > > template <class src_t, class dest_t> > void copy_union(src_t & src, dest_t & dest) > { > copy_col_if_in_dest<boost::mpl::at<src::typelist, boost::mpl::int_<0> > >::type,src::typelist>(); > }

2 1

ICE ICE BABY! (my MPL code is ICEing)
by Brian Braatz 04 Jun '05

04 Jun '05

Sitting in the backyard. Beautiful day, got up at 3am to get something figured out (for proposed boost profiler \ introspection library). Good news is I figured out the MPL code needed to accomplish my goal. After typing in an incremental version of the code, I got the dreaded internal compiler error (ICE). I am chewing on this- and thought I would post it so to provide an opportunity for someone to show how incredibly stupid I am because I did something that was obviously wrong. (you can even rub my nose in it if you like :) ) here is the code: I am tearing it apart\deconstructing to try to figure out what is kicking in the ICE. (I am also going to try it with an eval_if (which I should be using anyway) just to see what happens) I am using boost 1.32 and vc7.1 struct do_copy_col { typedef do_copy_col type; do_copy_col() { std::cout << "do_copy_col" << std::endl; } // todo- implement operator()(dest,src) }; struct do_not_copy_col { typedef do_not_copy_col type; do_not_copy_col() { std::cout << "do_not_copy_col" << std::endl; } // todo- implement operator()(dest,src) }; template <class type_t, class sequence_t> struct copy_col_if_in_dest : public if_< typename is_same< typename deref<typename boost::mpl::find<sequence_t, type_t>::type >::type, type_t>, do_copy_col, do_not_copy_col>::type { }; template <class src_t, class dest_t> void copy_union(src_t & src, dest_t & dest) { copy_col_if_in_dest<boost::mpl::at<src::typelist, boost::mpl::int_<0> >::type,src::typelist>(); }

1 0

Function type to function call, how to?
by Arkadiy Vertleyb 04 Jun '05

04 Jun '05

Hi all, I was trying to come up with a useful example of LVALUE_TYPEOF, so I decided to implement the "trace" function from the motivation section of decltype/auto proposal. However I ran into unexpected problem: Given a function type F, and parameter type T, how to create an expression for typeof to use in the return type? I use VC71, and the only thing that seems to work (but with a warning) is: F()(T()) The warning is something about using a null pointer to call a function... *(F*)0 doesn't work, and something like make<F>() doesn't work either -- 'expression can't be evaluated to a function call with one parameter', or something like this. What would be the correct way of doing this, and is there one? Thanks in advance. Regards, Arkadiy

3 5