- Boost - lists.boost.org

[doxygen] CSS/Temple files for Doxygen
by Artyom Beilis 10 Jul '11

10 Jul '11

Hello, Where can I find CSS files/Template files suitable for Doxygen documentation generation? So I can make Doxygen output to have more Boost-like style. Artyom Beilis -------------- CppCMS - C++ Web Framework: http://cppcms.sf.net/ CppDB - C++ SQL Connectivity: http://cppcms.sf.net/sql/cppdb/

6 13

[system] How are users expected to learn Boost.System?
by Lee Richmond 10 Jul '11

10 Jul '11

Hi, Since Boost.System doesn't appear to have any documentation other than its reference section, I'm wondering how users are expected to really learn this library? It would be nice if a tutorial section could be added to its documentation, including tutorials outlining how to extend the library with our (the library user) own error sources. That is: when to extend, what to extend and how to extend. The unclickable "Tutorial" link currently on the library's home page almost feels like a slap in the face ;) Thanks, Lee.

2 2

[1.47.0] Any remaining showstoppers?
by Beman Dawes 09 Jul '11

09 Jul '11

Any remaining 1.47.0 showstoppers? I'm still waiting for an OK to merge http://svn.boost.org/trac/boost/changeset/72855 Are there any other merges ready to go that still need permission to merge? --Beman

9 30

Bugs in regression/library_status.cpp on Windows
by Robert Kawulak 09 Jul '11

09 Jul '11

Hello, I've stumbled upon several bugs in tools/regression/src/library_status.cpp when compiling on Windows with gcc 4.5 and msvc 10.0. I'm attaching a patch file with all necessary fixes. The bugs are: - using path::native() instead of path::string() - the former is wstring on Windows, - calling path::string() twice to obtain begin() and end() iterators - on Windows path::string() returns a temporary, so the iterators point to different objects, - using whole path for each node in build_node_tree() instead of only the directory name, - using whole path instead of only the test name in do_table_body(), - the link to links file is relative to test directory, not to the results output directory, - generation of closing "</a>" without opening "<a>". Best regards, Robert

1 0

Re: [boost] Logging library
by Ivan Le Lann 09 Jul '11

09 Jul '11

----- "Tigran Hayrapetyan" <tigran.human(a)gmail.com> a écrit : > Hello dear Boost community > (...) > Existing logging libraries which I used, are writing log messages > linearly, > just one after another. > In contrast, as we know, almost any application's working structure is > not > linear, but hierarchical instead, i.e. functions which call other > functions, > and so on. (...) > So I decided to implement such logging library. I put it in namespace > "tlog" > by now, for "Tree Logging". In my implementation hierarchical messages > are > presented in XML format, as it is some standard for tree-like data. > There are two attached files. The library itself is in "tlog.hpp" > file, and > "main.cpp" contains a sample application which is using the library > and > writes log to "tlog.xml" file. They are written in MS Visual Studio > 2008. > Log messages are written with help of 4 macros: > > TLOG_INFO( Logger, Text ) : logs > message > with specified text to specified logger with priority "information" > TLOG_WARNING( Logger, Text ) : logs > message with > specified text to specified logger with priority "warning" > TLOG_ERROR( Logger, Text ) : logs > message > with specified text to specified logger with priority "error" > TLOG_SCOPED_NODE( Logger, NodeName ) : opens xml element in > the log > file. The element will be closed at the end of scope, so there is no > need to > close it manually. > > I think that the best way to implement entering/leaving subnodes of > the log > file, is to use scopes of the code. > > Please review run result and the code, and see if it can be useful > for boost. Have you tried Boost.Log and its attributes ? http://boost-log.sourceforge.net/libs/log/doc/html/log/detailed/attributes.… http://boost-log.sourceforge.net/libs/log/doc/html/log/detailed/utilities.h… Ivan

2 2

[assign v2] thoughts
by Gordon Woodhull 09 Jul '11

09 Jul '11

Hi Erwann, all, I had a thought about this library which I wasn't able to articulate during the review, and I don't think anyone else raised it. I think this library would be a useful utility if it had the right syntax, but I found issue especially with the use of operator| which is IMO not consistent with the way Boost.Range uses it. I would expect the pipe operator to just stream data and not affect the object on the LHS. I know everything is relative and in some EDSLs this might be appropriate, but it's a strain for me to think of pipe as anything other than unidirectional data flow. The syntax that you have now cal | do_csv_put<2>("jan", 31, "feb", 28, "mar", 31) looks kind of like Smalltalk or Objective C to me, like you are trying to send some arbitrary message to the container on the LHS. Did you consider keeping the overloaded operator+= from Assign v1, or another assignment operator that always modifies the LHS? cal += values<2>("jan", 31, "feb", 28, "mar", 31) is much clearer to me. And the assignment operators are expected to return a reference to the LHS. Yes, there is an ambiguity with other libraries which have meant += to mean "add to each element" but with the rare exception of containers containing elements of their own type, this should always be resolvable. If this is unacceptable, I think operator|=, meaning union-assign, also makes some sense. Or am I missing something? I always thought of Assign v1 as "the funny operator, library" and on at least one occasion didn't use it because I knew I'd have to explain what it's doing to my colleagues. So Assign v2 looks very powerful and useful to me, but in order to become a workhorse for everyone, it has to have the most blunt obvious syntax so that no one has to guess or explain anything. Cheers, Gordon

1 0

[graph] bug in ith_bandwidth?
by Joel Young 08 Jul '11

08 Jul '11

Is ith_bandwidth broken? It only measures the distance between indices when the source is greater than the sink. Should it look like this: template <typename Graph, typename VertexIndexMap> typename graph_traits<Graph>::vertices_size_type ith_bandwidth(typename graph_traits<Graph>::vertex_descriptor i, const Graph& g, VertexIndexMap index) { BOOST_USING_STD_MAX(); typedef typename graph_traits<Graph>::vertices_size_type size_type; size_type b = 0; typename graph_traits<Graph>::out_edge_iterator e, end; for (boost::tie(e, end) = out_edges(i, g); e != end; ++e) { int f_i = get(index, i); int f_j = get(index, target(*e, g)); using namespace std; // to call abs() unqualified if(f_i > f_j) b = max BOOST_PREVENT_MACRO_SUBSTITUTION (b, size_type(f_i - f_j)); else b = max BOOST_PREVENT_MACRO_SUBSTITUTION (b, size_type(f_j - f_i)); } return b; } Joel

2 1

Boost
by Joshua Boyce 08 Jul '11

08 Jul '11

Hey all, Ruben over at the MinGW-w64 project has started providing a Clang toolchain built upon MinGW-w64. http://lists.cs.uiuc.edu/pipermail/cfe-dev/2011-July/016005.html I just tried compiling Boost trunk r72754 with it, and figured I'd post the output here so work can commence bringing it in line with GCC via MinGW-w64. 90% of the errors seem to be a result of a 'call to deleted constructor' which seems to be caused by the shared_ptr library (or its usage)... Then the remaining 5% seem to be quite similar in nature, so it doesn't look like stuff will be too horrible to fix, but I don't know enough about the implementation of shared_ptr to fix it at the moment. One other thing to take into consideration will be the lack of a Clang Windows toolset. Currently when you set your toolset to Clang under MinGW, Boost treats it as Clang-Linux. I'm not sure what exactly that will affect... The script I used and the resulting output is available here: http://www.raptorfactor.com/code/Boost-Clang-MinGW-Output-20110708a.zip Thanks..

1 0

[fusion] random-access single_view
by Eric Niebler 07 Jul '11

07 Jul '11

Seems like a dumb thing, but it would actually help make my code simpler if single_view were random access. Patch attached. OK to commit (to trunk, of course)? The patch excludes the regenerated docs, which I would commit along with the rest. Thanks, -- Eric Niebler BoostPro Computing http://www.boostpro.com

2 1

More Boostcon videos available
by Marshall Clow 07 Jul '11

07 Jul '11

Monday: Micheal Wong - C++0x Progress and status http://blip.tv/boostcon/c-0x-progress-and-status-5349952 Ruediger Berlich - Distributed parametric optimization with the Geneva library http://blip.tv/boostcon/distributed-parametric-optimization-with-the-geneva… Tuesday: Joel Falcou - Automatic Hybrid MPI+OpenMP Code Generation http://blip.tv/boostcon/automatic-hybrid-mpi-openmp-code-generation-5349961 Sebastian Schaetz - Mastering the Cell Broadband Engine architecture through a Boost based parallel communication library http://blip.tv/boostcon/mastering-the-cell-broadband-engine-architecture-th… Beman Daves -- The proposed Boost B-tree library http://blip.tv/boostcon/the-proposed-boost-b-tree-library-5349968 Boris Kolpackov - Parsing C++ with GCC plugi http://blip.tv/boostcon/parsing-c-with-gcc-plugins-5349971 More next week.... -- Marshall Marshall Clow Idio Software <mailto:mclow.lists@gmail.com> A.D. 1517: Martin Luther nails his 95 Theses to the church door and is promptly moderated down to (-1, Flamebait). -- Yu Suzuki

1 0