- Boost-users - lists.preview.boost.org

[uBLAS] Identity Matrix Constructor vs Resize
by Ryan 14 Mar '11

14 Mar '11

I have an identity matrix that when initialized with the constructor I don't get an error. If I do a default construction and then resize the matrix I get a runtime error. I'm not sure the difference. Could someone enlighten me? Ryan using namespace boost::numeric::ublas; template<typename matrixT> matrix_range<matrixT> topleft (matrixT & M) { return subrange(M, 0, 3, 0, 3); } int main(int, char const* []) { vector<double> initial(3); initial[0] = 10; initial[0] = 100; initial[0] = 2; //I realize that "matrix<double> P(zero_matrix<double>(9,9));" would do the same thing as below. //This is extracted from a class and the constructor needs to to this. matrix<double> P; //This is a class variable P.resize(9,9); //This happens in the constructor P.clear(); //This happens in the constructor //The following runs without errors //identity_matrix<double> identity(3); //This causes a run time error. identity_matrix<double> identity; identity.resize(3); topleft(P) = initial[0] * identity; return 0; }

1 0

backporting filesystem v3
by Arash Abghari 14 Mar '11

14 Mar '11

Hi, I wonder if it is possible to backport filesystem v3 to 1.40 by replacing source folder from 1.46. Thanks Arash

2 2

boost::mutex over a socket
by Marco Piacentini 14 Mar '11

14 Mar '11

Hi guys...I've 2 thread and I would like that one of these performs a synchronous read, menawhile the other performs a synchrounous write...adn obviously they share the same socket.....the signature of the functions are the following: length=boost::asio::read(socket_,boost::asio::buffer(msg,sizeof(msg)), ec); boost::asio::write(socket_,boost::asio::buffer(msgToSend,sizeof(msgToSend))); socket_ is a private field of the classe Server. How can I use the mutex to perform concurrently and safetly these operations? Thanks..

2 4

[GIL] Compilation times
by Germán Diago 14 Mar '11

14 Mar '11

Hello. I'm using g++-4.4 with boost libraries and they're great. However, I have to use some image manipulation algorithms for two types of images (rgb8 and gray). I'm using it successfully, however, compilation times have increased a lot: I use apply_operation and any_image and any_image_view with two types each (rgb8 and gray8) Is there a way to decrease compile times? Thanks in advance.

2 2

Small logic::tribool riddle
by Thomas Jarosch 14 Mar '11

14 Mar '11

Hello, here's a small boost::logic::tribool riddle: -<snip>------------------------------------- #include <iostream> #include <boost/logic/tribool.hpp> int main(void) { boost::logic::tribool foo(boost::logic::indeterminate); if (foo != true) { std::cout << "reached1\n"; } if (foo == boost::logic::indeterminate) { std::cout << "reached2\n"; } return 0; } -<snap>------------------------------------- Try to determine the printed result without looking at the boost::logic::tribool implementation / documentation. Intuitive? A coworker just shot himself in the foot... Best regards, Thomas Jarosch

4 6

Re: [Boost-users] Boost-users Digest, Vol 2660, Issue 1
by Leon Sit 14 Mar '11

14 Mar '11

> > ---------- Forwarded message ---------- > From: Joel Falcou <joel.falcou(a)lri.fr> > To: boost-users(a)lists.boost.org > Date: Sat, 12 Mar 2011 08:24:53 +0100 > Subject: Re: [Boost-users] [proto] Object got sliced when passed by > reference > I just compiled your smaple using g++-4.5 and i got: > > logical_and( > logical_or( > terminal(4baseI2F1E) > , terminal(4baseI2F2E) > ) > , terminal(4baseI2F3E) > ) > hodqhoqw > 0 > > Why is I think the correct answer. So this is tied to some MSVC specific > behavior, i guess mayeb on how the terminal< base<_> const &> is matched. > matching terminal< base<_> > should be enough I think no ? Could you try ? > > HTH > > > > Thats what I did at the very beginning but fail. So I moved on to try const& but still fail. Before I started using proto, I hand-craft some overloaded operator to return the expression object too but failed with the same error. This makes me think that it is a compiler bug too. Should I try something else too?

1 0

[hash] Performance & distribution, variable bit width
by Sebastian Karlsson 14 Mar '11

14 Mar '11

As I use boost::hash quite a bit I have a few questions which the documentation seems to leave out. First of all it would be terrific if there was a performance & distribution benchmark as a part of the documentation to make it easier to evaluate the merits of using it. Secondly I'm a bit stumped on why the return value is fixed to size_t for a couple of reasons. Bit width of size_t isn't guaranteed to be constant, wouldn't this severely affect the distribution properties of the hashing algorithm used? Furthermore it would be great if the user could choose the bit width of the hash. Often the birthday problem is relevant, and being able to swap to using a 64 bit hash easily is something that doesn't seem possible but often desirable. Kind regards, Sebastian Karlsson

1 0

punits][accumulators] buggy behavior for min accumulator of quantities
by alfC 14 Mar '11

14 Mar '11

Hi, Sometimes it is difficult to make boost::units quantities with other libraries. For example to have boost::accumulators mean work with units, one has to specialize division by integer for quantities. In this example the accumulators and units compiles out of the box but gives an incorrect answer. #include<boost/accumulators/accumulators.hpp> #include <boost/accumulators/statistics/min.hpp> // or other stat features #include <boost/accumulators/statistics/max.hpp> // or other stat features #include<boost/units/systems/si.hpp> #include<boost/units/io.hpp> using namespace boost::accumulators; using namespace boost::units; int main(){ accumulator_set<quantity<si::time>, features<tag::min, tag::max> > a; a(2.1*si::second); a(2.0*si::second); a(4.1*si::second); std::cout << extract::min(a)<< " " << extract::max(a)<< std::endl ; return 0; } It prints: 0 s 4.1 s instead of: 2.1 s 4.1 s So, it seems that the min extractor fails for quantities. I would like to know if there is a way to correctly specialize "min" for quantities. (By the way, std::min for quantities works fine). However the behavior is unfortunate in the first place. Regardless of the solution I would label this as a bug of one of the two libraries. Thank you, Alfredo

2 3

Preventing iostreams from accessing freed memory
by Adam Nielsen 14 Mar '11

14 Mar '11

Hi all, I'm unable to prevent Boost iostreams from accessing memory once it has been freed, and I'm hoping someone here familiar with iostreams has some advice. As far as I can tell, when I push a standard iostream onto a filter, the filter holds on to a reference of it. If this happens inside a function, then when the function returns the iostream is freed, followed by the filter. In its destructor the filter tries to perform a flush/close and promptly segfaults, as it is trying to access the iostream which has just been freed. There is a set_auto_close() function I can call on the filter which sounds like it would do the trick, but it doesn't and according to the docs it's because it "does not prevent the Filters and Devices in the chain from being closed [flushed] automatically if the chain is complete at destruction." So is there any way I can prevent a filtering_stream from performing I/O in its destructor? Thanks, Adam.

1 1

[property tree][json]how to create empty array?
by Kulti 12 Mar '11

12 Mar '11

Hi, all! It's possible to get following output from boost::property_tree::json_parser::write_json ? { "id": "id", "method": "test_method", "params": [] } I know how to create array with some items, but when I try to create empty array, I've get "params": ""

3 3