- Boost - lists.boost.org

[repost][threads] final decision on making parts of Boost.Threads header-only
by joaquin＠tid.es 16 Jun '09

16 Jun '09

I'm reposting this in the hope of catching the attention of the lib's author, who is at the end of the day the person reponsible of making the decision: Hi, A few weeks ago we had a discussion about the convenience of making part of Boost.Threads, namely mutexes and locks, header-only. Some (including Anthony Williams) supported the idea, others disliked it: http://lists.boost.org/Archives/boost/2009/04/150524.php What the final decision on this issue is? If positive (i.e. favoring the change), can it be expected to be done for Boost 1.40? As a remainder, I've created an associated trac ticket: https://svn.boost.org/trac/boost/ticket/2964 Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

3 4

Proposal: null pointer class
by Martin Törnwall 15 Jun '09

15 Jun '09

Hello Boost developers, I'd like to propose the inclusion of a null_ptr class in Boost. While I am fully aware that this topic has been brought up several times in the past, it appears as if though all discussions slowly died off without any final decision as to whether it is a worthy candidate for inclusion in Boost or not. The null pointer class proposals I have seen before were all designed to function in a manner similar to that of the "nullptr" keyword of C++0x. That is, the classes typically provided a template conversion operator, which allows it to evaluate to any pointer type with the value zero. My approach is to make the class, rather than the operator, a template: template<typename T> class null_ptr { public: operator T*() const { return reinterpret_cast<T *>(0); } }; As we can see, passing a null_ptr requires us to explicitly specify the type of pointer. While slightly verbose, I believe that it can be highly beneficial, for the following reasons: 1. In cases where multiple overloaded functions that take pointers exist, there is no way that we could accidentally call the wrong overload (since we have to explicity specify the type). 2. It makes code easier to read and understand. When reading the code, another programmer will instantly be able to see what type argument(s) a particular function accepts, thus making it much clearer than using NULL or nullptr. As with all other null pointer classes that have been suggested to date, the primary benefit of using my null_ptr over the NULL constant is that it solves the following problem: void f(int) { ... } void f(char *) { ... } int main() { func(NULL); func(null_ptr<char>()); } When passing NULL, f(int) will be called -- probably not what the programmer intended. The second call utilizes null_ptr<char> to guarantee that the overload of f() taking char* will be called instead of the one taking int. As noted above, my null_ptr class also handles multiple overloads taking pointer types gracefully. Please let me know what you think! P.S: I have tested the code on Microsoft Visual C++ 2008 as well as gcc 4.3.3 (Linux). Regards, Martin Törnwall

6 12

A question about Units & Spirit
by John Phillips 15 Jun '09

15 Jun '09

In the middle of the directX thread, Daniel Duffy asked a question that seems more suited to its own thread. I'm starting that thread and giving others a chance to join into the discussion by giving you the content of the few messages that have passed by so far. John From Daniel --------------------------------------------------- Hello, This is my first mail, so I hope I have sent it to the right address :) I am using the Units library for Dimensional Analysis. So far it is very useful. Up till now I am using it in a hard-coded/compile time way as in a test.cpp. My inteterst is in using Units as part of a larger run-time system so I need some way of making my code more flexible. Lambda library is an option here but I would also like to create units and quanties from string input as well as being able to parse these strings to units and then store the units in a database. Can I use Spirit (and Serialization) for these features? Maybe the functionality already exist, and I have not found it. best regrads Daniel J. Duffy From Hartmut --------------------------------------------------- Sure Spirit is your friend for all parsing and output generation needs. What did you have in mind? Regards Hartmut From Daniel ---------------------------------------------------- Hartmut, I would like to generate well-formed units and create expressions of the followiong kinds: Examples of unit expressions: 1. 2m + 15m*32s/16s = 32m. 2. 3m*7s + 2m*10s/5m = 21m + 4s -> ERROR. Can't add meters to seconds. The second group looks more tricky. Examples of simplification: 1. (x-2)2 + 4x2 - 2^x + 1 = 5x2 - 8x + 10 2. x2 + 7x + 10 = (x + 2)(x + 5) 3. (a + b)(a - b) = a2 - b2 4. sin(x)2 + cos(x)2 = 1 5. 2sin(x)cos(x) = sin(2x) One tricky part is for example #5 where the code needs to 'know' trigonometry. best regards Daniel From John ---------------------------------------------------------- Daniel J. Duffy wrote: > Hartmut, > I would like to generate well-formed units and create expressions of the followiong kinds: > > Examples of unit expressions: 1. 2m + 15m*32s/16s = 32m. 2. 3m*7s + 2m*10s/5m = 21m + 4s -> ERROR. Can't add meters to seconds. So, you want this to happen at run time, or at compile time? At compile time, I think you are already set with the Units library. At run time, there are examples with the library that would help. If parsing the input is your concern, then Spirit is a good choice. (Spirit 2 is even better, by the way.) Using the two together, you could write a program where a user enters your example lines and gets correct responses. > > The second group looks more tricky. > > Examples of simplification: 1. (x-2)2 + 4x2 - 2^x + 1 = 5x2 - 8x + 10 2. x2 + 7x + 10 = (x + 2)(x + 5) 3. (a + b)(a - b) = a2 - b2 4. sin(x)2 + cos(x)2 = 1 5. 2sin(x)cos(x) = sin(2x) > One tricky part is for example #5 where the code needs to 'know' trigonometry. As far as I can tell here, you are looking for a program that does symbolic algebra. If that is true, then I don't think boost has a complete solution for you. Writing the parser for the input with Spirit wouldn't be that bad, but writing the processing engine that matches patterns and does simplifications is not covered by any current Boost libraries. I don't recall seeing anyone poll for interest on this, so I don't think there is even a partial library in the sandbox. Would you want this at run time, compile time, both? I haven't tried to write it, but I think a pure compile time implementation would be prohibitively slow. For a pure run time solution, you might consider linking to already existing solutions, such as Maple or Mathematica. > > best regards > > Daniel > John

2 2

[ptr_container] Can't store a pointer to const in multimap?
by Amit Kumar 15 Jun '09

15 Jun '09

Please consider the following code: #include "boost/ptr_container/ptr_map.hpp" int main() { typedef boost::ptr_multimap<char, const int> M; M m; char c = 'c'; int* j = new int(7); m.insert(c, j); return 0; } The gcc 4.3.3 compiler fails to compiler the above code. Am I doing something wrong by inserting a non-const? Can't I store a pointer to const in the multimap? try.cpp: In function ‘int main()’: try.cpp:8: error: expected primary-expression before ‘=’ token /usr/include/c++/4.3/bits/stl_pair.h: In constructor ‘std::pair<_T1, _T2>::pair(const std::pair<_U1, _U2>&) [with _U1 = char, _U2 = const int*, _T1 = const char, _T2 = void*]’: boost/boost_1_39_0/boost/ptr_container/ptr_map_adapter.hpp:765: instantiated from ‘typename boost::ptr_container_detail::ptr_map_adapter_base<T, VoidPtrMap, CloneAllocator, Ordered>::iterator boost::ptr_multimap_adapter<T, VoidPtrMultiMap, CloneAllocator, Ordered>::insert_impl(const typename boost::ptr_container_detail::ptr_map_adapter_base<T, VoidPtrMap, CloneAllocator, Ordered>::key_type&, typename boost::ptr_container_detail::ptr_map_adapter_base<T, VoidPtrMap, CloneAllocator, Ordered>::mapped_type) [with T = const int, VoidPtrMultiMap = std::multimap<char, void*, std::less<char>, std::allocator<std::pair<const char, void*> > >, CloneAllocator = boost::heap_clone_allocator, bool Ordered = true]’ boost/boost_1_39_0/boost/ptr_container/ptr_map_adapter.hpp:799: instantiated from ‘typename boost::ptr_container_detail::ptr_map_adapter_base<T, VoidPtrMap, CloneAllocator, Ordered>::iterator boost::ptr_multimap_adapter<T, VoidPtrMultiMap, CloneAllocator, Ordered>::insert(typename boost::ptr_container_detail::ptr_map_adapter_base<T, VoidPtrMap, CloneAllocator, Ordered>::key_type&, typename boost::ptr_container_detail::ptr_map_adapter_base<T, VoidPtrMap, CloneAllocator, Ordered>::mapped_type) [with T = const int, VoidPtrMultiMap = std::multimap<char, void*, std::less<char>, std::allocator<std::pair<const char, void*> > >, CloneAllocator = boost::heap_clone_allocator, bool Ordered = true]’ try.cpp:9: instantiated from here /usr/include/c++/4.3/bits/stl_pair.h:106: error: invalid conversion from ‘const void*’ to ‘void*’ I also tried but failed with const casting j before insertion.

3 2

[serialization][trunk] Latest patch for void_cast.cpp causes segfaults
by Hartmut Kaiser 15 Jun '09

15 Jun '09

Hi all, apparently the latest patch to boost/libs/serialization/src/void_cast.cpp (rev.53078, svn:date: 2009-05-18T06:25:56.023190Z, svn:author: ramey, svn:log: fix errors in extended_type_info management) causes an existing application to segfault during cleanup. The application was working properly before. The attached patch fixes the issue. Ok to commit? Regards Hartmut

1 0

Removing Bind placeholders from the global namespace
by Peter Dimov 15 Jun '09

15 Jun '09

This topic has again come up recently in boost-users. I have already assigned myself a ticket, https://svn.boost.org/trac/boost/ticket/2240 but have done nothing so far, in part because it would break other Boost libraries. (Patching placeholders.hpp is trivial in itself.) This is one case in which the trunk/release procedures don't work well as-is unless someone volunteers to produce a patch against all of Boost that at once moves the placeholders into namespace boost::placeholders _and_ fixes all errors that this would cause. It'd be easier to split the work: I put the breaking change in and then the libraries are fixed one by one. :-) Comments? -- Peter Dimov http://www.pdplayer.com

10 31

[candidate bloom_filter] nitpick
by JOAQUIN M. LOPEZ MUÑOZ 15 Jun '09

15 Jun '09

Hi Michael, I see that internals.hpp declares the following: namespace boost { namespace detail { template <class Input, class BitSet> class bloom_filter_internals {...}; } } boost::detail should not be used for internals of a specific library. The custom is to do it like this: namespace boost{ namespace bloom_filters{ namespace detail{ // you might want to rename to just "internals" // as the qualified name is sufficiently informative template <class Input, class BitSet> class bloom_filter_internals {...}; } } } Note the "s" in "bloom_filters", following the naming guidelines at: http://www.boost.org/development/requirements.html#Naming_consistency Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 2

Boost and XML-RPC
by Dmitriy Lyfar 15 Jun '09

15 Jun '09

Hello boosters, What do you think about boost.xmlrpc implementation? Some days ago I need simple xmlrpc library with https support, but I cannot find anything for windows. What I've found it's http://libiqxmlrpc.sourceforge.net/ library. But it has too many dependencies, I really don't understand why I should compile half of gnome to have an ability call xmlrpc methods. So after some libs I realized that there is no good xmlrpc lib for windows at least (if you know, please let me know). So what are you thoughs about implementation such lib as part of boost (asio based for example)? Is it needed for community? Thank you, Dmitriy.

2 3

Blechmann's lock-free
by Michael Caisse 14 Jun '09

14 Jun '09

Tim - I just downloaded your lock-free data structure code and was getting ready to take a look at it. Two questions immediately come to my mind: 1. What is your intent for this library? Do you plan on having it reviewed? 2. Do you have pre-made docs available somewhere? I seem to consistently have troubles getting the doc chain going. Best Regards - michael -- ---------------------------------- Michael Caisse Object Modeling Designs www.objectmodelingdesigns.com

2 1

[date_time] Change to the time_duration default format.
by Ilya Bobir 14 Jun '09

14 Jun '09

Hi all, I would like to propose a change to the default format for the time_duration objects. Current format is "%H:%M:%S%f". It does not allow one to input a time_duration that is longer than 99 hours. My proposal is to change the format to be "%O:%M:%S%f". time_duration objects are not restricted to 99 hours so the current behavior seems to be incorrect. I have submitted a patch but was told that the change is a braking change and need to be discussed on a mailing list: https://svn.boost.org/trac/boost/ticket/1861#comment:3 Please reply if you think there could be issues with the change to the default time_duration format. Here is a simple test that shows what I'm talking about: #include <iostream> #include <sstream> #include <boost/date_time.hpp> using namespace std; using namespace boost::posix_time; int main(void) { time_input_facet* facet = new time_input_facet(); locale loc(locale(locale(), facet)); time_duration td1(hours(100)); time_duration td2; stringstream ss; ss.imbue(loc); /* Input using default time_duration format. */ facet->time_duration_format("%H:%M:%S%f"); ss << td1; ss >> td2; cout << td2 << endl; /* Input using proposed time_duration format. */ ss.str(""); ss.clear(); facet->time_duration_format("%O:%M:%S%f"); ss << td1; ss >> td2; cout << td2 << endl; } For me it outputs: 00:00:00 100:00:00 It means that you can not serialize and deserialize arbitrary time_duration objects if you are using the default format. While the output facet do write extra digits the input facet will accept only two digits for the hours field. Ilya Bobir

2 1