- Boost - lists.boost.org

Re: Singleton
by CHAOS 05 Jan '05

05 Jan '05

Dave Handley wrote: > Ordering of Singletons is, in my opinion, one of the most important issues with Singletons. It is almost certainly insufficient in most applications to create in a specific order, and delete in reverse order. One reason for this is start-up performance - you don't want every Singleton to have to create at startup (necessarily). Another reason can be memory efficiency. Loki again provides a number of policies including a Phoenix Singleton and a Longevity managed singleton. I personally tend to use Longevity managed singletons linked to the at_exit function. Nevertheless, there are some additional things that I would like to see - for example, a phoenix singleton that could be created and destroyed more than once during a program run could be useful - either explicitly or even if not used for a while. Also, allowing thread-safety as an option is also crucial. The way that my design works is to use reference counting in the form of dependency objects, which create the singleton when the first dependency is constructed and destroy the singleton when the last dependency is destroyed. If a singleton A owns a dependency on singleton B, that dependency lifetime will be greater than the lifetime of A itself, meaning that B will exist for the duration of A's lifetime. The only property of my singleton that forces it to be constructed before main and destructed after is the existence of a static dependency in the singleton class to itself. If I removed this static dependency, it would allow for singletons that could be created and destroyed any number of times, and lifetime would be completely controlled by the scope at which the dependencies lived. Implementing this change would make it possible for GetInst to fail, when it would be more desirable to have GetInst somehow utilize a dependency itself. Perhaps I could enable this by having GetInst return a type of smart pointer, which would own a singleton dependency as well as pointing to the singleton object. Thoughts / suggestions? I have no experience in dealing with thread safety. Would ensuring thread safety require that the singleton have access to a threading library, or would it be possible to just write the code more carefully to ensure safety, similar to writing exception safe code? In addition, I have no familiarity with the other design patterns you mentioned, nor with Loki. I learned about the singleton concept through a friend and thought that it made for an interesting challenge. I just ordered Modern C++ Design, so hopefully I can familiarize myself with these topics in the near future. For reference, I will leave a link to the code below: http://www.ezequal.com/chaos/libs/singleton.h

1 0

Yet another variant-style type, and boost inter-library dependencies
by christopher diggins 05 Jan '05

05 Jan '05

I have written a union-style type similar to boost::variant which allows single values, and multiple types. This type is called a union-list ( ul ) and is described at http://www.codeproject.com/useritems/union_list.asp I will be honest, there are several disadvantages of the union-list when compared to boost::variant 1. Declaration style is uglier : typdef ul<int, ul<char, ul<float, ul<string, ul_end> > > > my_union_list; 2. the data size is huge = summation of the size of the individual types plus sizeof(bool) * #types. 3. union-list has very few features The advantages (and my motivation for creating it) of the union-list is that it requires no other boost libraries and has a very tiny code base. Another difference ( but not neccessarily an advantage ) of the union-list is that it uses and provides an index to access the type of its contents. I don't know if this is sufficient to warrant eventual submission as a boost mini-library. Any comments? In general I am curious about how other boosters feel about boost library sizes and inter-library couplings. For me this is a big obstacle to using boost in many circumstances. It doesn't seem to be a concern at all for the boost community, but I wonder if other programmers have similar sentiments and find themselves often seeking out simpler alternatives. Christopher Diggins http://www.cdiggins.com http://www.heron-language.com

4 9

Plans to submit the Boost Software License for OSI approval?
by Christopher D. Russell 05 Jan '05

05 Jan '05

Are there plans to submit the Boost Software License to OSI to get it added to the "Approved" list (http://opensource.org/licenses/)? OSI certification would be useful: - allow SourceForge project admins to publish under "Boost Software License" as opposed to "Other / Proprietary" - less explaining to business people / lawyers - Regards Chris

2 2

RE: [boost] Re: Portable alternative for VS.NET __if_exists
by Brian Braatz 05 Jan '05

05 Jan '05

> -----Original Message----- > From: boost-bounces(a)lists.boost.org [mailto:boost-bounces@lists.boost.org] > On Behalf Of Roman Yakovenko > > > -----Original Message----- > > From: David Abrahams [mailto:dave@boost-consulting.com] > > Mårten Rånge wrote: > > > I'm looking for a portable version of __if_exists (VS.NET > > keyword) and > > > unfortunately google turned up nothing good. > > > > > > I use boost fairly regular and checked boost::type_traits > > but found nothing > > > that seemed to similar. > > > > > > Has anyone heard of something that can accomplish this? > > > > It might help if you described what it does. > > full explanation: > http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vclang/htm… > [Brian Braatz] I Find it a TAD bothersome that the link above does NOT state it is a MS specific implementation. OR - Is this something recently added to the standard I never noticed???? Anyone know?

2 1

Anyone received Dave A's MPL book yet?
by Brian Braatz 05 Jan '05

05 Jan '05

http://www.amazon.com/exec/obidos/tg/detail/-/0321227255/102-2639086-432 8160?%5Fencoding=UTF8&v=glance I ordered it on amazon, and it now says it will arrive Feb 16th. If anyone knows of a faster way to get it, Ill cancel my amazon order. (Maybe Dave should sell them directly that we he could keep all the profit)

7 6

RE: [boost] Re: Anyone received Dave A's MPL book yet?
by Brian Braatz 05 Jan '05

05 Jan '05

> -----Original Message----- > On Behalf Of David Abrahams > Subject: [boost] Re: Anyone received Dave A's MPL book yet? > > Brian Braatz wrote: > > http://www.amazon.com/exec/obidos/tg/detail/-/0321227255/102-2639086-432 > > 8160?%5Fencoding=UTF8&v=glance > > > > I ordered it on amazon, and it now says it will arrive Feb 16th. > > > If anyone knows of a faster way to get it, Ill cancel my amazon order. > > > > (Maybe Dave should sell them directly that we he could keep all the > > profit) > Dave Abrahams Wrote: > Correction: Dave A's and Aleksey G's MPL book. I'd consider selling [Brian Braatz] Aleksey Gurtovoy: MY APOLOGIES TO YOU. Writing a book is not a easy undertaking and helping the rest of us get better at metaprogramming is extremely cool all by itself. I am sorry I referred to the book as "DaveA's book".

1 0

[Random] uniform_int vs. uniform_real vs. uniform_01
by Doug Gregor 05 Jan '05

05 Jan '05

I'm having a few problems dealing with uniform_real and uniform_01. The basic issue is that uniform_real does not always produce values between its min and its max. The problem consistently occurs when I'm using minstd_rand as my generator (one of the linear_congruential variants). For instance, if I write: template<typename RandomNumberGenerator> void foo(RandomNumberGenerator& gen) { uniform_int<int> rand_int(some_min_int, some_max_int); // Okay, some_min_int <= x <= some_max_int int x = rand_int(gen); uniform_real<double> rand_double(some_min_double, some_max_double); // Uh-oh, some_min_double <= y <= some_max_double doesn't always hold double y = rand_double(gen); } Looking at the source to uniform_real, the problem is obvious: result_type operator()(Engine& eng) { return eng() * (_max - _min) + _min; } uniform_real assumes that the engine returns a value in [0, 1], but doesn't check that. If this is correct (I hope it isn't!), the "obvious" answer is to use uniform_01<RandomNumberGenerator...> as input to uniform_real. However, uniform_01 is also broken in several ways: 1) It isn't a random distribution because its operator() does not accept an engine. Instead, the engine is embedded in the uniform_01 object. 2) uniform_01 stores the incoming random number generator by _value_, so the original generator "gen" does not advance. You can't even fake it with uniform_01<RandomNumberGenerator&>, because that creates reference-to-reference problems. Below is a small patch to make uniform_real<...> behave nicely. As for uniform_01: we should deprecate it in 1.33.0 and remove it afterwords, because it is hopelessly broken and not even included in the TR. Doug Index: uniform_real.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/random/uniform_real.hpp,v retrieving revision 1.14 diff -u -r1.14 uniform_real.hpp --- uniform_real.hpp 27 Jul 2004 03:43:32 -0000 1.14 +++ uniform_real.hpp 30 Dec 2004 22:11:33 -0000 @@ -50,7 +50,11 @@ void reset() { } template<class Engine> - result_type operator()(Engine& eng) { return eng() * (_max - _min) + _min; } + result_type operator()(Engine& eng) { + return static_cast<result_type>(eng() - eng.min()) + / (eng.max() - eng.min()) + * (_max - _min) + _min; + } #if !defined(BOOST_NO_OPERATORS_IN_NAMESPACE) && !defined(BOOST_NO_MEMBER_TEMPLATE_FRIENDS) template<class CharT, class Traits>

3 3

Re: Question and idea for smart pointers
by Ion Gaztañaga 05 Jan '05

05 Jan '05

> typedef pointer_to_other<NewType>(my_smart_ptr)::type new_pointer_t; Sorry, I mean obviously, typedef pointer_to_other<MyPtr, NewPointeeType>::type new_pointer_t; Regards, /Ion

1 0

Boost Shmem new version available in Boost Files
by Ion Gaztañaga 05 Jan '05

05 Jan '05

Hello to all, I'm announcing a new version of Boost.Shmem library. The file is in Boost Files section (Boost.Shmem.2005.01.04.zip). After some months working with other people from boost, the library has suffered some serious rework. First of all, all documentation has been ported to BoostBook format and can be built with bjam --v2 from doc directory. It has been a hard work to port from word format to boostbook, but now, the boostbook format is much nicer. Please, this is the first rough version of the documentation, so I would be glad to receive feedback about this. Regarding to library, new features have been added: -> Shared memory mapping to fixed addresses -> Templatized mutex types, so that different locking mechanisms can be used indepently from the framework -> Templatized shared memory allocation algorithm, so that the framework can be expanded with new allocation algorithms. -> Templatized pointer types in shared memory STL allocator, so that the pointer type is independent from the framework. This allows using all STL implementations with the framework if the memory is mapped in the same address in different processes. -> Use of boost preprocessor to implement a new, wider, more clear calling syntax. See documentation to see more details. -> New relative pointer types, to balance pointer size, pointed object range, etc... There are many new features in the library but I wanted to mantain this announcement short. The library has been tested in Visual C++ .NET 2003 and gcc 3.4.3 in Suse 9.2 Linux. Please, feel free to report any comment, bug or idea. Regards, Ion

1 0

Re: [ublas] gmane news archive not updated
by Toon Knapen 05 Jan '05

05 Jan '05

Paul Baxter wrote: > I note that the gmane news interface to this list (and the other boost > lists) got stopped in mid December. The others have been restarted, but > ublas is still not showing newer posts (mailman archive is fine). I'm > afraid I don't know what's involved in restarting them.... me neither ;-( Anyone ? > Perhaps it might make sense to have an additional 'common administrator' > for the various boost mailing lists for carrying out activities common > to all lists (like restarting the feeds). I'm not suggesting the > administrator also does ublas-specific work as the current ublas team > are doing a great job. (BTW thanks for setting up the new list!) > > I also note that http://www.boost.org/more/mailing_lists.htm still > refers to the yahoo ublas list. It might be useful to update that link > as well to refer to both the mailman page and the gmane ublas news feed > in a similar way to the spirit and language bindings projects now do. Anybody can explain me the process for doing this ?

2 2