May 2007 - Boost-users - lists.preview.boost.org

[1.33][tuple] Problem with tuples longer than 10 elements
by Seweryn Habdank-Wojewódzki 15 May '07

15 May '07

Dears The code: #include <boost/tuple/tuple.hpp> typedef boost::tuple<int,double,char,int,double,double,double,double,double,double> Foo; int main() {} Is working. But if I add one more field: #include <boost/tuple/tuple.hpp> typedef boost::tuple<int,double,char,int,double,double,double,double,double,double,double> Foo; int main() {} I have an error: tuple_type.cpp:3: error: wrong number of template arguments (11, should be 10) /usr/include/boost/tuple/detail/tuple_basic.hpp:75: error: provided for ?template<class T0, class T1, class T2, class T3, class T4, class T5, class T6, class T7, class T8, class T9> class boost::tuples::tuple? tuple_type.cpp:3: error: invalid type in declaration before ?;? token Why this error occurs? My kompiler is gcc 4.1 Regards. -- |\/\/| Seweryn Habdank-Wojewódzki \/\/

1 0

[bug] Suggestion for clarification to front page
by Paul Giaccone 15 May '07

15 May '07

In http://boost.org/libs/libraries.htm, I suggest changing the heading and link that reads "Broken compiler workarounds". 1. If this means "Workarounds for broken compilers", then change either to that phrase or to "Broken-compiler workarounds"; prefer the former for clarity. 2. If this means "Compiler workarounds that are broken", change to that phrase. As it stands, this is ambiguous: it is not clear which noun the adjective "broken" modifies. Does this section give workarounds (a good thing) or report workarounds that are broken (an unfortunate thing)? Following the links, I see it is the former (after all, what would be the point in giving workarounds that are broken?), but this is not obvious at first glance. Sorry if this is not the right place to post this.

1 0

Cross-compilation with mingw32 on Linux
by Maciej Blizinski 15 May '07

15 May '07

Hi Boost users, I have setup a MinGW cross-compilation suite on my Linux box. To create a Windows binary, the only thing I need to do is call specific C++ compiler name: $ i686-mingw32-g++ -o hello.exe hello.cpp Now I have a Windows executable on my disk, I can test it with wine: $ wine hello.exe Hello, world! I've written a program in C++; I would like to create binary for windows. I'm using Boost libraries, so I need to cross-compile Boost as well. I've tried to compile Boost like this: maciej@clover ~/src/boost_1_34_0 $ ./configure --with-toolset=mingw --prefix=$HOME/win32 Building Boost.Jam with toolset mingw... ./configure: line 199: ./bootstrap/jam0: No such file or directory tools/jam/src//bjam Detecting Python version... 2.4 Detecting Python root... /usr Unicode/ICU support for Boost.Regex?... /usr Backing up existing Boost.Build configuration in user-config.jam.6 Generating Boost.Build configuration in user-config.jam... Generating Makefile... Error messages... but Makefile has been created. Trying... maciej@clover ~/src/boost_1_34_0 $ make ./tools/jam/src//bjam -sICU_PATH=/usr --user-config=user-config.jam /bin/sh: ./tools/jam/src//bjam: No such file or directory Not all Boost libraries built properly. Can I somehow compile Boost using i686-mingw-c++ compiler? Any suggestions? -- Maciej Bliziński http://automatthias.wordpress.com

2 1

How to define _SCL_SECURE_NO_DEPRECATE when building signals? (or, win64 with intel compiler?)
by Nicholas Bastin 15 May '07

15 May '07

I caught the note that I needed to define _SCL_SECURE_NO_DEPRECATE on windows when building with vc8 (or, in my case the intel compiler, which uses the same SDK). Is there some mechanism for doing this in bjam? Or, asked another way, has anyone succesfully built boost.signals in 64-bit on windows using ICL 9.1? And if so, what magic bjam command line incantation did you use? :-) -- Nick

2 1

1.34.0 RC3
by Jeff 14 May '07

14 May '07

Hello, Might the fix for this problem be in there? http://article.gmane.org/gmane.comp.lib.boost.user/27122 Thanks, Jeff

7 11

a few boost.View questions
by abir basak 14 May '07

14 May '07

When I was searching for a view library to have some container adapters so that I can store a portion of a dynamic container in a class, I came across Boost View library and uBLAS range_vector and slice_vector. (Iterators are unsuitable for my purpose as I can't store a portion of a container through iterator when container are resizing (through push_back & pop_front) , and also iterators are dumb enough so they dont give the index for the container they are locating :( ) So I needed the alternative view classes or container adapters , which are store-able even when container is allocating/reallocating as long as the portion of the container which the view is pointing is available. At present I have my own view classes (only the classes I need ) , but planning to switch to boost View classes. A few questions regarding boost View library (from sandbox) 1) is it an accepted library ? if not when it will be accepted ? 2) Apart from range_view all other views are based on container (i.e they are container adapter ). So all other views are store-able while range view is not ( when the iterator becomes invalid , the view is also) any reason for this decision ? 3) for step_iterator, why no step_view is given ? 3) How can I make a composite view mixing a set of views like 2 range view, or one range_view and one permutation_view. For eg. I want to make a container adapter for a container taking a range 2-10 (range_view) and 15,18,17 (permutation_view ). AFAIK some of this mixed views are supported by blitz library. Thanks abir -- Abir Basak, Member IEEE Software Engineer, Read Ink Technologies B. Tech, IIT Kharagpur email: abir(a)abirbasak.com homepage: www.abirbasak.com

2 3

[BoostCon 07] Some First Pictures from Aspen
by Jeff Garland 14 May '07

14 May '07

I know some of you were looking for 'blogs' from BoostCon. I won't promise much on the text side, but I borrowed my wife's new camera and will be taking some photos during the conference and posting them. Thru a variety of circumstances I arrived a day early in Aspen, so I've had a chance to look around a bit. These are photos of the outside of the Aspen Institute and Aspen Meadows hotel. I think the word 'idyllic' pretty much sums up the surroundings! http://www.crystalclearsoftware.com/boostcon_07/FirstPictures/index.html Jeff

3 3

[serialization] macros and template classes
by alex solovyov 14 May '07

14 May '07

Hi all, It is not possible to use the macro BOOST_CLASS_EXPORT_GUID or BOOST_IS_ABSTRACT etc. with a template class. You can certainly redefine the macro to use the template classes as follows: #define BOOST_TEMPLATE_CLASS_IS_ABSTRACT(CLASS_NAME, TYPENAME_ARG_1 …, TYPENAME_ARG_n) \ namespace boost \ { \ namespace serialization \ { \ template<> \ struct is_abstract<CLASS_NAME<TYPENAME_ARG_1, …, TYPENAME_ARG_n> > \ { \ typedef mpl::bool_<true> type; \ BOOST_STATIC_CONSTANT(bool, value = true); \ }; \ } \ } But here we have the following problem: 1. How many definitions are written? 2. What happens with the following template class: Class1<int, Class2<int, char> >; it defines another new macro (: The answer for this problem I found in MPL library. We need a wrapper (traits): struct abstract_class { typedef Class1<int, Class2<int, char> > type; }; and the macro: #define BOOST_TEMPLATE_CLASS_IS_ABSTRACT(CLASS_WRAPPER) \ namespace boost \ { \ namespace serialization \ { \ template<> \ struct is_abstract< CLASS_WRAPPER::type > \ { \ typedef mpl::bool_<true> type; \ BOOST_STATIC_CONSTANT(bool, value = true); \ }; \ } \ } It is not an optimal solution, but reduces the source code. Best regards Alex Solovyov

1 0

[Boost.Graph] unused variable warning
by Christoph 13 May '07

13 May '07

please fix the unsused variable warning in graph/topological_sort.hpp:40 Thank you, Best regards Christoph

1 0

Installing boost with bloodshed dev c++ and windows xp
by Paul Epstein 12 May '07

12 May '07

I am trying to install boost -- I have a bloodshed dev c++ compiler 4.9.9.2 and windows xp. I am trying to install boost and then quantlib but am stuck at the boost stage. I find that the instructions at boost.org appear to assume more knowledge of dos commands and path environment variables than I possess. I sometimes get the message that bjam is not recognized, and I sometimes get the message that the toolsets are not configured. I do know how to add paths to the path environment variable but not much else. I tried to follow Terrence August's online instructions (which are excellent) http://www.stanford.edu/~taugust/quantlib.html but this interfered with my dev cpp compiler. Thank you very much for your help. Paul Epstein ____________________________________________________________________________________Be a better Globetrotter. Get better travel answers from someone who knows. Yahoo! Answers - Check it out. http://answers.yahoo.com/dir/?link=list&sid=396545469

1 0