- Boost - lists.boost.org

sun4v compile error
by Rohit Goel 04 May '10

04 May '10

Hi, I am using the interprocess package of boost for file locks. Though the code builds fine on Windows & Linux, I get the following errors on Solaris "/etc/sw//boost/interprocess/detail/move.hpp", line 649: Error: Templates can only declare classes or functions. "/etc/sw//boost/interprocess/detail/move.hpp", line 656: Error: A declaration was expected instead of "while". "/etc/sw//boost/interprocess/detail/move.hpp", line 656: Error: No direct declarator preceding "(". "/etc/sw//boost/interprocess/detail/move.hpp", line 658: Error: Multiple declaration for f. "/etc/sw//boost/interprocess/detail/move.hpp", line 658: Error: A declaration was expected instead of "++". "/etc/sw//boost/interprocess/detail/move.hpp", line 660: Error: A declaration was expected instead of "return". "/etc/sw//boost/interprocess/detail/move.hpp", line 668: Error: Templates can only declare classes or functions. "/etc/sw//boost/interprocess/detail/move.hpp", line 672: Error: A declaration was expected instead of "}". "/etc/sw//boost/interprocess/detail/move.hpp", line 686: Error: is_movable is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 686: Error: enable_if is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 686: Error: type is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 695: Error: is_movable is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 695: Error: disable_if is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 695: Error: type is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 696: Error: A previously specified default argument value cannot be changed. "/etc/sw//boost/interprocess/detail/move.hpp", line 696: Error: move_detail::uninitialized_move_move_iterator<move_detail::I, move_detail::F>(move_detail::I, move_detail::I, move_detail::F, int*) already had a body defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 706: Error: is_move_iterator is not a member of move_detail. "/etc/sw//boost/interprocess/detail/move.hpp", line 706: Error: Unexpected type name "I" encountered. "/etc/sw//boost/interprocess/detail/move.hpp", line 706: Error: Expression must have a constant value. "/etc/sw//boost/interprocess/detail/move.hpp", line 706: Error: enable_if is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 706: Error: type is not defined. "/etc/sw//boost/interprocess/detail/move.hpp", line 730: Error: is_move_iterator is not a member of move_detail. "/etc/sw//boost/interprocess/detail/move.hpp", line 730: Error: Unexpected type name "I" encountered. "/etc/sw//boost/interprocess/detail/move.hpp", line 730: Error: Expression must have a constant value. "/etc/sw//boost/interprocess/detail/move.hpp", line 730: Error: disable_if is not defined. Any ideas what could be wrong ? -Rohit

1 0

[xint] granularity
by vicente.botet 03 May '10

03 May '10

Hi, N1692 talks about granularity "For the operations on the integer data, its granularity must be defined, that is how wide are the data chunks that are operated on". Which is the granularity for Xint? _____________________ Vicente Juan Botet Escribá http://viboes.blogspot.com/

2 1

[date_time] Maintainer?
by conferno 03 May '10

03 May '10

Hi, boost::date_time has a lot of bugs: https://svn.boost.org/trac/boost/report/14?page=2 Their owner is az_sw_dude, who seems to be inactive since 2007: https://svn.boost.org/trac/boost/search?q=az_sw_dude http://groups.google.com/groups/search?hl=en&q=az_sw_dude Who is the maintainer of boost::date_time and is there any hope that the bugs will be fixed ?

3 2

DotNet FileSystemWatcher like library proposal.
by Fernando Pelliccioni 03 May '10

03 May '10

Hello all, I am working on a library like DotNet FileSystemWatcher ( http://msdn.microsoft.com/en-us/library/system.io.filesystemwatcher%28VS.80…) Basically the library monitors for file system changes and notifies throwing events. First, it will support the following platforms: - Windows NT 5.0 (Windows 2000) or higher using O/I Completion Port. - Linux 2.6.13 or higher using inotify. - FreeBSD 4.1 or higher using kqueue. - NetBSD 2.0 or higher using kqueue. - OpenBSD 3.6 higher using kqueue. (Not sure which version supports it) - MacOS/X 10.4 higher using kqueue. (Not sure which version supports it) Future supported platforms: - Linux from 2.4.0-test9 2.6.12 using dnotify. - Sun OpenSolaris using FEM API. Testing: I am testing the library on FreeBSD 7.2, Linux 2.6.31, Windows XP (Windows NT 5.1). I need help to test on older versions of these operating systems. Also in NetBSD, OpenBSD and mainly on MacOSX because I do not have the necessary hardware. I would like to include this library in Boost, I think would be useful. What do you think? If so, I would like to have a mentor (like GSoC) and experienced in Boost development. Thanks and regards, Fernando Pelliccioni --- References: - O/I Completion Port: http://msdn.microsoft.com/en-us/library/aa365198%28VS.85%29.aspx http://en.wikipedia.org/wiki/Input/output_completion_port - inotify: http://en.wikipedia.org/wiki/Inotify - kqueue: http://people.freebsd.org/~jlemon/papers/kqueue.pdf http://en.wikipedia.org/wiki/Kqueue - dnotify: http://en.wikipedia.org/wiki/Dnotify

6 8

[xint] what about a DSEL?
by vicente.botet 03 May '10

03 May '10

Hi Chad, I dont know the complexity of the algorithms behind the xint operations, but instead of making eager evaluation you could do lazy evaluation. You could define a DSEL which evaluates the expression x*y + z at once. it would be great if at the end w = x*y + z; resulted in only one allocation for the result of the expression, and a move to make the assignement. If given two integers i and j using n and m bits, you know the number of bits needed by i @ j, you can know the number of bits of x*y + z, isn't it? If in addition, the rhs expression resulted on a integer needing less bits than the lsh, you could do the operation directly on the lhs, avoiding any allocation! My 2cts, _____________________ Vicente Juan Botet Escribá http://viboes.blogspot.com/

1 0

[multi_array]regression test failures on trunk
by Thomas Klimpel 02 May '10

02 May '10

Hi, the Boost.MultiArray regression test for "resize" fails on trunk (<http://www.boost.org/development/tests/trunk/developer/multi_array.html>), but passes on the release branch (<http://www.boost.org/development/tests/release/developer/multi_array.html>). This recently created some confusion for me (<http://lists.boost.org/Archives/boost/2010/05/165798.php>), because I was looking at the regression test results for the release branch, but did my local testing on trunk. The failures on trunk are caused by the changes from the commit <https://svn.boost.org/trac/boost/changeset/53789>. These changes were never merged to the release branch. The attached patch seems to fix these failures, but I take no responsibility for its correctness (as I've only tried to find the changes that caused the failure, but have not tried to understand the code). Regards, Thomas

1 0

Re: [boost] [castor] Interest in Logic Paradigm for C++ ?
by Joris IJsselmuiden 02 May '10

02 May '10

My humble opinion to the matter: I think it is a great idea to try and make Castor part of Boost. I have been using Castor in my research for a while now, and I think that it is definitely Boost-material. A logic engine can be a powerful tool, and it is a shame that C++ doesn't have one readily available. I am confident that Castor is the best choice for adding logic support to Boost. As far as I know, Boost is the most popular code-base that is not part of the C++ standard library yet. So, making Castor part of Boost would be a great way of increasing Castor's visibility. It would be nice to keep Castor available as a stand-alone library, so that you don't need Boost if you only want to use Castor. But I am sure that this issue can be solved: Castor could be a Boost "companion" library, or one could have a stand-alone version and a Boost version side-by-side. Although this last solution doesn't seem that great to me. I hope I will find the time to do some testing for Castor 1.1. Keep in touch! Joris Ijsselmuiden _________________________________________________________________ Hotmail: Powerful Free email with security by Microsoft. https://signup.live.com/signup.aspx?id=60969

3 3

Parallel Boost Algorithm
by anil khadka 02 May '10

02 May '10

I am pretty new to both boost and parallel boost. I am using Brandes Centrality algorithm, but the performance really doesn't meet my expectation. If I used Java, which is only using 1 CPU core would complete it in around 30 hrs. The parallel version which use 8 CPU cores tooks 17 hours!! I don't know why there is just around 13 hours gain if i use 8 cores against 1, unless there are lots and lots of communication between each cores. I am using just 1 machine (not a cluster), right now. Graph size is Node 80K, edges 120K I am reading edges from files. I used distributed compressed_sparse_row to represent graph. Right now, it seems like all the cpu cores are reading files on its own, I am assuming each CPU core will takes its graph part automatically (distributed algorithm do it ?). Note: I downloaded the parallel version from: http://osl.iu.edu/research/pbgl/software/ Any help/suggestions would be really appreciated. Metanil.

2 1

[Wave][1.43] Asking for permission to merge low impact bug
by Hartmut Kaiser 01 May '10

01 May '10

Hey, I would like to ask for the permission to merge a low impact bug to the release branch (patch: https://svn.boost.org/trac/boost/changeset/61723) Regards Hartmut --------------- Meet me at BoostCon www.boostcon.com

2 1

[contract] next round of development
by Lorenzo Caminiti 01 May '10

01 May '10

Hello all, FYI, I am starting the next round of development of Boost.Contract. Most of the planned modifications derive from feedback I have received from this mailing list -- thanks! My goal is to request Boost.Contract formal review after this round of development and related documentation updates. You can check the planned development items at: http://sourceforge.net/tracker/?group_id=253844&atid=1330818 Comments are always welcome. Regards, -- Lorenzo

2 2