August 2014 - Boost - lists.preview.boost.org

[interprocess] atomic_write32 on GCC/x86/x68_64 lacks memory barrier
by Lars Hagström 12 Aug '14

12 Aug '14

Hi! In a piece of legacy code I've had a few uses of the atomic operations inside Boost.Interprocess (yes, I know they're internal, but I had no other choice at the time, so sue me :-). I've been debugging a thread synchronization issue which ended up in me actually looking for a problem inside interprocess, and I believe that I found the error. The atomic_write32 function, when running gcc on intel platforms does not have a memory barrier, which causes my threads to not see changes that they should see. A little bit of googling gave me the following link http://boost.2283326.n4.nabble.com/interprocess-atomic-write32-td4173289.ht…, which discusses exactly this issue. I've now fixed my code to use Boost.Atomic instead of relying on boost interprocess internals, so I'm good, but I worry that the issue might affect boost.interprocess as well. So I propose that atomic_write32 should be implemented using atomic_cas32. Or indeed that Boost.Atomic should be used instead, which is more work, obviously. Thoughts? /Lars

5 20

[1.56] Sorry but no PDF doc build (build script has been broken)
by John Maddock 11 Aug '14

11 Aug '14

Sorry folks, but as things stand there will be no PDF build of the docs for 1.56 as the build script appears to be hopelessly broken :-( Each time I fix one Boost.Build error it just leads to another, and frankly at present I have neither time nor inclination to keep tracking these down. If anyone wants to experience some pain then trying to invoke bjam in /doc/pdf/ will give you an idea of some of the issues. Regards, John.

6 7

Re: [boost] [multi_index]: Migration of code to VS12 does not compile anymore
by Brian Wood 11 Aug '14

11 Aug '14

Johannes Brunen writes: >Hello, > > > I'm porting some code fromVS10 to the VS12 compiler on windows. I have a >problem compiling code like the following: Clang didn't like the code either: clang++ -c -I ../boost-1_56_0_b1/ mi.cc mi.cc:21:35: error: use 'template' keyword to treat 'nth_index' as a dependent template name typedef typename container_t::template nth_index<0>::type sequence_index_t; ^ 1 error generated. -- Brian Ebenezer Enterprises - So far G-d has helped us. http://webEbenezer.net

1 0

[multi_index]: Migration of code to VS12 does not compile anymore
by Johannes Brunen 11 Aug '14

11 Aug '14

Hello, I'm porting some code fromVS10 to the VS12 compiler on windows. I have a problem compiling code like the following: //Test.cpp #include <boost/multi_index_container.hpp> #include <boost/multi_index/random_access_index.hpp> #include <boost/multi_index/hashed_index.hpp> #include <boost/multi_index/identity.hpp> template<typename T> class test { public: typedef typename boost::multi_index_container < T*, boost::multi_index::indexed_by < boost::multi_index::random_access<>, boost::multi_index::hashed_non_unique < boost::multi_index::identity<T*> > > > container_t; typedef typename container_t::nth_index<0>::type sequence_index_t; }; This does not compile and gives the following error message: 1> Test.cpp 1>D:\test\Test.cpp(18): error C2059: syntax error : '<' 1> D:\test\Test.cpp(19) : see reference to class template instantiation 'test<T>' being compiled 1>D:\test\Test.cpp(18): error C2238: unexpected token(s) preceding ';' I'm unable to find the reason for this error. Is it a compiler error, or is something wrong with my declaration? Regards, Johannes

2 1

Missing <boost/detail/endian.hpp>
by Edward Diener 11 Aug '14

11 Aug '14

I am not seeing a <boost/detail/endian.hpp> after pulling the latest 'master' and pulling all the submodules. What have I done wrong ?

2 1

Showstopper for Boost 1.56?
by Niall Douglas 10 Aug '14

10 Aug '14

Is anyone else seeing this warnings spew on Boost 1.56 beta when compiled with clang 3.4? https://ci.nedprod.com/view/Boost.AFIO/job/Boost.AFIO%20Build%20POSIX_ Linux64_clang%203.4/11/consoleFull GCC 4.8 produces no such spew. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

10 39

Boost 1.56.0 has been released
by Marshall Clow 09 Aug '14

09 Aug '14

Release 1.56.0 of the Boost C++ Libraries is now available. These open-source libraries work well with the C++ Standard Library, and are usable across a broad spectrum of applications. The Boost license encourages both commercial and non-commercial use. This release contains one new library and numerous enhancements and bug fixes for existing libraries. === Modularization === Boost version control has migrated to a system using git submodules. This shouldn't make too much of a difference to users, although the directory structure is now a bit different. Parts of some libraries have been moved into different modules, and several new modules have been extracted from existing code. All header paths should remain the same. The new modules are: • Assert: Customizable assert macros. Maintained by Peter Dimov. • Core: Core utilities used by other libraries, with minimal dependencies. Maintained by Peter Dimov, Glen Fernandes and Andrey Semashev. • Lexical_Cast: General literal text conversions, such as an int represented a string, or vice-versa, from Kevlin Henney. • Throw_Exception: A common infrastructure for throwing exceptions from Boost libraries, from Emil Dotchevski. • Winapi: Windows API declarations without <windows.h>, for internal Boost use. === New Libraries === • Align: Memory alignment functions, allocators, and adaptors, from Glen Fernandes. • Type_Index: Runtime/Compile time copyable type info, from Antony Polukhin. For details, including download links, see http://www.boost.org/users/news/version_1.56.0 You can also download directly from SourceForge: http://sourceforge.net/projects/boost/files/boost/1.56.0/ To install this release on your system, see http://www.boost.org/doc/libs/release/more/getting_started/index.html Thanks, --The Boost release team Vladimir Prus Rene Rivera Marshall Clow Eric Niebler Daniel James Beman Dawes

8 7

[config] Macro for initializer-list literals?
by Peter Dimov 09 Aug '14

09 Aug '14

We have BOOST_NO_CXX11_HDR_INITIALIZER_LIST, which tells us whether we can #include <initializer_list> and use std::initializer_list; we also have BOOST_NO_CXX11_UNIFIED_INITIALIZATION_SYNTAX, which tells us whether we can use X{ 1, 2 }, or X x{ 1, 2}. But we don't have a macro that tells us whether this: struct X { int a, b; }; void f( X x ) { } int main() { f( { 1, 2 } ); } is supported. This is a problem for http://www.boost.org/development/tests/develop/developer/output/Sandia-inte… because the test is guarded by BOOST_NO_CXX11_UNIFIED_INITIALIZATION_SYNTAX: https://github.com/boostorg/smart_ptr/blob/develop/test/allocate_shared_arr… and this macro is not set for Intel 13.1 in C++0x mode. Presumably the compiler allows braced initialization, but not braced literals.

1 0

[1.56.0] Release candidates available
by Marshall Clow 08 Aug '14

08 Aug '14

Release candidate files for 1.65.0 are available at http://boost.cowic.de/rc/ As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. This helps ensure the candidates build OK before we push them out to SourceForge. The files (and associated md5s) are: MD5 (boost_1_56_0_rc1.7z) = 0ac22b92b24209286d48e4c41925cd9f MD5 (boost_1_56_0_rc1.tar.bz2) = 5b5bfa99244b1ac5737323decceb19c2 MD5 (boost_1_56_0_rc1.tar.gz) = 2047cdd4e09cda34e7f10fec0f821175 MD5 (boost_1_56_0_rc1.zip) = 0c544f6acc85550b15cdc9221f2ac0c9 Thanks! -- The release managers

13 23

A quieter clang build
by Marshall Clow 08 Aug '14

08 Aug '14

I added the following lines to my user-config.jam: using clang-darwin : noC99W : "/usr/bin/clang++" : <cxxflags>"-Wno-c99-extensions -Wno-variadic-macros" ; and then built the 1.56.0 release with "toolset=clang-darwin-NoC99W", and lo! there were a *sane* amount of warnings. Just a quick tip for people drowning under the warnings. — Marshall P.S. I’m leaving for vacation this afternoon, but would like to pick this back up when I return.

1 0

Regression testing unobservable behavior?
by Beman Dawes 08 Aug '14

08 Aug '14

Certain library function errors are difficult to regression test because it is not obvious how to observe their behavior in a test framework. Examples include: 1) Selection of correct overload or template specialization when the only difference in effect is a performance enhancement. 2) Verification that BOOST_ASSERT fires when a precondition is violated. 3) Verification that I/O and other operating system API errors are handled correctly. 4) More generally, something occurs in a function under test that is hard to observe but needs to be verified. How do other boost developers address these regression testing needs? It seems to me all of the examples can be solved by creating some form of back channel (or channels) to pass information between caller and callee functions. All require getting additional information from a function after it is called. (3) also requires passing additional information to a function when it is called. An ideal solution requires: * Works in the existing Boost regression test framework. So much the better if it works in other test frameworks. * Does not violate the ODR or depend on undefined behavior. * Does not alter the interface of the function under test. * Does not generate any code except when the function is actually being tested. * Does not require that each test case be run as a separate program. * Works even if the function under test is noexcept. * Is lightweight, utterly reliable, and has few if any dependencies beyond those already present. --Beman

3 5

Error Handling issue in NonCopyable Interface
by John Pallino 08 Aug '14

08 Aug '14

Hello, I am having error handling issues that is in regards to the NonCopyable Interface library. Here is the source code: #include <boost/noncopyable.hpp> #include <iostream> class A : public boost::noncopyable { public: A() { std::cout << "In A \n" << std::endl; } class AnError {}; void Func() { if () //error condition throw AnError(); } }; int main() { try { A object1; A object2(object1); object1 = object2; } catch (A::AnError) { //tell user about error. } return 0; } } So far, I have tried on the forums in CPlusPlus.com and MSDN, but no responses. Since that Boost.org has created and authored the NonCopyable.hpp library, I would like to get some assistance. Can anybody give me some solutions on how to suppress any compilation errors? Thanks in advance. Email me back at the Cc: if you have any answers or comments with detailed information and examples. John P.

2 1

The release branch is now available for merging
by Marshall Clow 07 Aug '14

07 Aug '14

Now that 1.56.0 is out the door …. — Marshall

2 2

[1.56.0] Release Candidate 3 available
by Marshall Clow 07 Aug '14

07 Aug '14

We've created a third release candidate to fix a one more issue with Boost.Range. The files for 1.56.0 RC3 are available at: http://boost.cowic.de/rc/ As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. The files (and associated md5s) are: MD5 (./boost_1_56_0_rc3.7z) = b7a72a6a6f487fa445333b4e115c6171 MD5 (./boost_1_56_0_rc3.tar.bz2) = a744cf167b05d72335f27c88115f211d MD5 (./boost_1_56_0_rc3.tar.gz) = 8c54705c424513fa2be0042696a3a162 MD5 (./boost_1_56_0_rc3.zip) = fc16da21d641e03715a4144cb093964e Thanks! -- The release managers

7 10

Clang, -pedantic, and variadic macro warnings
by Edward Diener 06 Aug '14

06 Aug '14

My test show that if you compile with clang, pass the '-pedantic' option, and also pass 'std=c++0x' or 'std=c++11', that the many warnings about variadic macro use do not show in the output. I know there was a discussion and an assertion that the warnings do show in the output oonce '-pedantic' is passed as a compiler option even though a 'std=xxx' is passed that covers variadic macros, but I am not seeing that in my tests. If anybody else can tell me of a Boost library build where clang is the compiler, 'pedantic' is passed, and the appropriate 'std=xxx' to cover variadic macro use is passed, and the clang compiler warnings about using variadic macros show up, it would be appreciated.

1 0

[python] Building and Testing needs to be updated for modular-boost
by Edward Diener 05 Aug '14

05 Aug '14

Attempting to follow the building and testing explanation for Boost.Python in modular-boost, it appears it no longer works with the new structure of modular-boost and Boost Build. Staring with the 3.1 Basic Procedure section I invoke in the libs/python/example/quickstart/ directory the command: > b2 toolset=gcc-4.8.1 --verbose-test test only to receive: > Unable to load Boost.Build: could not find build system. > --------------------------------------------------------- > C:\Programming\VersionControl\modular-boost\libs\python\example\quickstart\boost-build.jam attempted to load the build system by invoking > > 'boost-build ../../../../tools/build/v2 ;' > > but we were unable to find "bootstrap.jam" in the specified directory > or in BOOST_BUILD_PATH (searching C:\Programming\VersionControl\modular-boost\libs\python\example\quickstart\../../../../tools/build/v2). > > Please consult the documentation at 'http://www.boost.org'. I do not know who is taking care of Boost Python now but it appears that it needs to be updated for modular-boost.

2 8

Files in Github
by Axel Ismirlian 05 Aug '14

05 Aug '14

Which subrepository do the files under /home/boost/boost belong to? More specifically, I made a change to /home/boost/boost/date_time/time_resolution_traits.hpp and I am looking to push it upstream, but I cannot find who this file belongs to. Sincerely, -Axel

4 3

[filesystem] develop broken after current merge
by Jürgen Hunold 05 Aug '14

05 Aug '14

Hi Beman, the last merge broke Boost.FileSystem on Linux using gcc(-4.8) and clang(-3.6) in c++11 mode. Diagnosic: include/boost/filesystem/path_traits.hpp contains #ifndef BOOST_NO_CXX11_CHAR16_T inline void convert(const char16* from, const char16* from_end, This should most probably be "char16_t" for C++11. The same for "char32" which should be "char32_t". This affects several other files. This one I could patch myself. But then I encountered: BOOST_ROOT/boost/filesystem/path_traits.hpp:131:13: error: ‘interop’ does not name a type typedef interop::conversion_iterator<interop::narrow, interop::utf16, const char16*> I can't find a definition of "interop" anywhere in the current repository... If you need more diagnostics, please ping me. Yours, Jürgen -- * Dipl.-Math. Jürgen Hunold ! * voice: ++49 4257 300 ! Fährstraße 1 * fax : ++49 4257 300 ! 31609 Balge/Sebbenhausen * jhunold(a)gmx.eu ! Germany

2 1

Is there any interest in a dependency injection library?
by Krzysztof Jusiak 05 Aug '14

05 Aug '14

Hi, Is there any interest in a C++03/C++11/C++14 header only, type safe, library providing compile time, macro free *constructor dependency injection <http://en.wikipedia.org/wiki/Dependency_injection>*? *Key features:* * *Compile time *- no exceptions - if application compiles all dependencies will be be created accurately * *Macro free* - by default no need to specify constructor traits or register anything (less intrusive) * *Scopes deduction* - scopes are deduced based on type semantic * *Automatic conversion* between std/boost smart pointers * *Compile time policies* - example to detect circular dependencies or limit supported types only to specified *Hello World* ----------------------------------------------------------------------------------------------------------> #include <memory> #include <boost/shared_ptr.hpp> *#include <boost/di.hpp>* *namespace di = boost::di;* struct interface { virtual ~interface() { } }; struct implementation : interface { }; class hello_world { public: hello_world(std::unique_ptr<interface> up , const std::shared_ptr<interface>& sp1 , boost::shared_ptr<interface> sp2 , int i) : i(i) { assert(dynamic_cast<implementation*>(up.get())); assert(dynamic_cast<implementation*>(sp1.get())); assert(dynamic_cast<implementation*>(sp2.get())); assert(sp1.get() == sp2.get()); assert(i == 0); } int run() { return i; } private: int i = 0; }; int main() { // create injector and configure bindings auto injector = *di::make_injector*( *di::bind*<interface, implementation>() // scope will be deduced ); // create and run hello_world return injector.*create*<hello_world>().run(); } <---------------------------------------------------------------------------------------------------------- *Overview*: * boost library incubator: http://rrsd.com/blincubator.com/bi_library/di-dependency-injection-2 * repository: https://github.com/krzysztof-jusiak/di * documentation: http://krzysztof-jusiak.github.io/di/boost/libs/di/doc/html * issues: https://github.com/krzysztof-jusiak/di/issues * download: https://github.com/krzysztof-jusiak/di/archive/master.zip *Supported Compilers* - 100% test coverage: * Linux (x86/x86-64) * Clang 3.2/3.3/3.4/3.4.2+ (clean with Clang Static Analyzer and Valgrind) * GCC 4.7.3/4.8.2/4.9.0+ (clean with Valgrind) * Intel C++ 14.0.3+ (clean with Valgrind) * Windows (x86/x86-64) * MinGW 4.7.3/4.8.2+ * Visual Studio 2013+ (clean with DrMemory) * Darwin/Mac OS (x86-64) * Clang 503.0.40+ (Apple LLVM version 5.1) *Continuous integration*: * Linux, Darwin/Mac OS: https://travis-ci.org/krzysztof-jusiak/di * Windows: https://ci.appveyor.com/project/krzysztof-jusiak/di * Coverage: https://coveralls.io/r/krzysztof-jusiak/di *Similar libraries*: * https://bitbucket.org/cheez/dicpp * https://code.google.com/p/infectorpp * https://github.com/phs/sauce * https://code.google.com/p/wallaroo * https://code.google.com/p/hypodermic * https://github.com/admiyo/CppInject * http://qtioccontainer.sourceforge.net * https://code.google.com/p/autumnframework * http://sourceforge.net/projects/cpp-resolver (proposal for boost library in 2009: http://lists.boost.org/Archives/boost/2009/10/157477.php) * https://code.google.com/p/pococapsule * https://code.google.com/p/spring-cpp * http://www.cs.utah.edu/~arichard/inject.html * http://bobah.net/d4d/source-code/misc/cpp-dependency-injection * https://code.google.com/p/ffead-cpp * http://codebros.github.io/DiLite * https://code.google.com/p/kindi Java: https://github.com/google/guice C#: http://www.ninject.org/ Source code was internally reviewed. DI was successfully used in a commercial project as well. Looking forward to hear from you. Any feedback is more than welcome. Best regards, Kris

3 13

[1.56.0] Release Candidate 2 available
by Eric Niebler 05 Aug '14

05 Aug '14

We've created a second release candidate to fix a few issues. The three fixes since RC1 are listed at the end of this mail. The files for 1.56.0 RC2 are available at: http://boost.cowic.de/rc/ As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. The files (and associated md5s) are: MD5 (boost_1_56_0_rc2.7z) = 2e1688886aede1c3348e131a0e73a031 MD5 (boost_1_56_0_rc2.tar.bz2) = 2079fabbfb9fb046e6e12e381eff20fa MD5 (boost_1_56_0_rc2.tar.gz) = 96fe9c5f70759399217dc7ee113ecdd3 MD5 (boost_1_56_0_rc2.zip) = df42b60e120841fc3893de7c2529fc4a Thanks! -- The release managers === Changes Since RC1 === Boost.Thread bug is causing coroutine build to fail on gcc 4.6 in C++11 mode. https://github.com/boostorg/thread/pull/21 Boost.Graph broken on MSVC-12 and gcc-4.4. https://github.com/boostorg/graph/pull/14 Boost.System broken on cygwin, local_free_on_destruction not defined. https://github.com/boostorg/system/pull/4

9 12

Re: [boost] [1.56.0] Release Candidate 2 available
by Glen Fernandes 04 Aug '14

04 Aug '14

> We've created a second release candidate to fix a few issues. The three > fixes since RC1 are listed at the end of this mail. > > The files for 1.56.0 RC2 are available at: > http://boost.cowic.de/rc/ Builds successfully on: 1. OSX 10.9.4, x64, clang 3.4 (C++03, C++11) 2. Ubuntu 14.04, x64, gcc 4.8.2, (C++03, C++11) Glen

1 0

Re: [boost] How to establish boost configurations for cross-compilers not currently covered?
by Dave Gomboc 04 Aug '14

04 Aug '14

> > The boost config manual > > does not explicitly address the case where one is attempting to > > determine the relative lack of C++98 compliance of a cross-compiler > > (e.g., for an embedded system) that is not already within Boost's > > supported set of compilers. > > [Christopher Kormanyos] > Dave, may I ask what cross compilers you are considering for > your investigations? I am not sure that I can say precisely which, but I feel safe enough saying that none of them are intended to run under Unix or Linux. Unfortunately, this means I can't directly use "configure" as per John Maddock's suggestion. Though, it is true that some of the compilers run under Windows, so perhaps I could install Cygwin to obtain ./configure capability... or, perhaps I could try running under Wine. For the others, who knows... > Although I can not answer your question directly, I am also > very interested in finding ways to adapt or extend Boost support > to embedded systems compilers --- even for small micros. I am > particularly interested in extending the range of Boost and C++ to > compilers from suppliers such as IAR systems, Green Hills, > and also selected GCC crosses. > > I am also sensing increased interest in the Boost and C++ > communities regarding compilers not running *nix, Win32. > There is, in my opinion, lots of work to do here --- especially > when plain "int" is 16-bits in width or if an STL is lacking, or etc. Int is 32 bits on most, if not all, of the platforms that I was considering when I wrote my email. That said, there appears to be a surprising amount of (hobbyist) retrocomputing activitly over the past few years. So, I don't doubt at all that there might be interest even where plain "int" is even just 8 bits! Dave P.S. I ordered your book recently, Chris, and am looking forward to reading it when it arrives.

2 1

[random] Contributing Well equidistributed long-period linear (WELL) implementation
by Sergiu Dotenco 04 Aug '14

04 Aug '14

Some time ago, I implemented the WELL family of PRNGs introduced in: Francois Panneton, Pierre L'Ecuyer and Makoto Matsumoto. "Improved Long-Period Generators Based on Linear Recurrences Modulo 2". ACM Transactions on Mathematical Software, 32 (1, March) 2006, pp. 1-16. WELL is supposed to provide better equidistribution and bit-mixing properties for equivalent period lengths compared to Mersenne twister. The WELL family of generators provide period lengths ranging from 2^512-1 to 2^44497-1. The implementation can be found at https://bitbucket.org/sergiu/random/src/tip/well.hpp?at=default. The code has to be boostified, which I'm willing to do if there's any interest.

2 4

Hana Typeclass
by Robert Ramey 04 Aug '14

04 Aug '14

I can't review the whole library. So I'm confining my observations to just a couple of aspects. My method is to review some section of the documentation. it looks like a critique of the documentation, but it's really observations on the underlying design of the library. In this post I'm addressing Typeclass. Typeclass "Conceptually, type classes are an artifice allowing humans to manipulate objects of heterogeneous types with well-defined semantics. OK They serve a purpose very similar to C++ concepts (which are not in yet) and to Haskell type classes, except they do not have language support" OK - But I would argue that C++ concepts do have library support. It may not be perfect, but it's good enough for government work. template <class T> struct Printable { private: T t; std::ostream os; public: BOOST_CONCEPT_USAGE(Printable) { os << t; } }; "Type classes in Boost.Hana are a library-level implementation of such type constraints allowing us to organize our generic programming." When I read this it seems to be identical to the definition of what Boost Concept Checking is. "Getting concrete" Here is where we start to have problems. Rather than the simple BCC example above, we've got something a lot more complex and hard to understand. the example struct Printable { BOOST_HANA_TYPECLASS(Printable); // This is valid, but it's probably not a good idea unless those are // strongly related to the type class. int foo; void bar() const { }; struct baz { }; }; Describes what not to do - but doesn't describe what one is supposed to do. "As you will see later with minimal complete definitions, it is often useful to put other related members in a type class; don't hesitate to do it when it makes sense." Upon reading this - we don't know enough to know that "related members" means and what it means to "make sense". "When I introduced type classes, I said they allowed us to bundle together related operations called methods. This is because type classes can be seen as defining some kind of public interface consisting of the operations that are valid with any object satisfying some constraints" which is the C++ concept - or my own "type constraint" "I also said that they made it possible to explicitly state that a type satisfies those constraints, and how it does so" I would phrase this differently: explicitly state what operations must be valid when applied to a given type in order to deem it as fulfilling the "type constraints" "From now on, I will refer to the set of all types satisfying the constraints of a type class T as the instances of T" A big problem as far as I'm concerned. In C++ we use the phrase "instance of type T" for an entirely different thing. I would use the phrase: a "Printable" type - in the case of the example above. When referring to a set of type constraints I would pick something like "modeling the type constraints T or modeling the Printable type". This is familiar language to us. "To associate methods to a type class, we create a layer of indirection through the instance member of the type class. For example, let's say we want to have a method named print in the Printable type class: …" The whole rest of this section goes into how to create a set of types modeling "Printable". OK - But why is this necessary at all? If one tries something like: template<class T> void f(T & x) { BOOST_CONCEPT_ASSERT((Printable<T>)); std::cout << x; } class x { }; f(x); // compile time error here the program will trap at compile time with an error inside of Printable indicating that f doesn't model Printable because the operation ostream << x isn't supported by x; That's what I need. With type class it looks like that is expected that there be a default implementation for f if x isn't printable. It's not at all clear what this default definition is supposed to do. Print default error message at runtime? or what? The documentation notes that the usage of a default would be rare. But if it were to be used - it conflicts with the original purpose if type constraints (or concepts or type classes, etc). It's still not clear to me what happens if I use your type class system and invoke f(x). Do I get a compile time error? - what kind? where does it point to? How does it help the user find his error? To me the "type class" of the Hana library is a muddled re-implemenation of the boost concept check library. Its much more complicated and subtle and not as well documented and doesn't add anything to BCC. Robert Ramey -- View this message in context: http://boost.2283326.n4.nabble.com/Hana-Typeclass-tp4665978.html Sent from the Boost - Dev mailing list archive at Nabble.com.

5 7

[msm] State Machine and Dependency Injection
by Kris 04 Aug '14

04 Aug '14

Hi, Recently I have implemented Dependency Injection library, which I proposed to boost last week (http://boost.2283326.n4.nabble.com/is-there-any-interest-in-a-dependency-in…) Right now I wanted to shared how dependency injection might be integrated with Boost.MSM (eUML) and would really glad for any feedback. Boost.MSM is a great library, unfortunately actions/guards have to have empty constructor and whole logic has to be handled via Finite State Machine. Such approach is hard to maintain and test, so would be great to have possibility to implement action/guards without FSM dependency and with proper dependency injection instead. More information: http://krzysztof-jusiak.github.io/di/boost/libs/di/doc/html/di/examples/int… Hello World: class guard : public euml_action<guard> { public: guard(shared_ptr<data> data = nullptr, shared_ptr<other_data> = nullptr); // default constructor is required by eUML template<typename TEvent> bool operator()(const TEvent&) const; // no FSM dependency }; class action : public euml_action<action> { public: action(shared_ptr<data> data = nullptr, shared_ptr<other_data> = nullptr); // default constructor is required by eUML template<typename TEvent> void operator()(const TEvent&); // no FSM dependency }; class state_machine_ : public state_machine_def<state_machine_> { public: using initial_state = state1; BOOST_MSM_EUML_DECLARE_TRANSITION_TABLE(( state1() + event() [guard()] / action() == state2() ), transition_table) }; using state_machine = back::state_machine< state_machine_ , back::use_dependency_injection // Boost.DI dependency injection policy >; auto sm = di::make_injector().create<state_machine>(); // create state machine / actions, guards and their dependencies will be injected via DI framework sm.process_event(event()); Hot it works? * we do implement actions/guards with any dependencies we need * we add use_dependency_injection policy in state machine definition * we create state machine using dependency injection framework library What MSM with DI actually give us? * actions/guards are really easy to test / code doesn't require FSM dependency - loosely coupled * signature change of actions/guards may not require binding configuration changes (if it's not an interface or it's an interface already bound) * FSM is just a declarative abstraction - we don't need to test it * we can use all DI goods for controlling how dependencies will be injected - see https://github.com/krzysztof-jusiak/di Project (with workings tests): https://github.com/krzysztof-jusiak/msm Comparison between usage of Boost.MSM with DI and without it: http://krzysztof-jusiak.github.io/di/boost/libs/di/doc/html/di/examples/int… Looking forward to hear from you. Any feedback is more than welcome. Cheers, Kris -- View this message in context: http://boost.2283326.n4.nabble.com/msm-State-Machine-and-Dependency-Injecti… Sent from the Boost - Dev mailing list archive at Nabble.com.

1 0