- Boost - lists.preview.boost.org

GSOC 2013
by Niranjan Singh 14 Apr '13

14 Apr '13

@antoshk Hello sir, As I go through the graph library of boost, everything seems to be done like all major algorithms for graphs. That's pretty good however we can use it to implement a structure of methods which deals with problems like Chinese postman problem, Decision making poblem , Project management.,Graph partitioning and clustering , Distribution of Eigen values and compression e.t.c With graph we can solve friend-of-friend problems and also implement a simple command line social network based on c++ with the help of graphs. This would be really interesting. However shortest path algorithms can be implemented from one cutting plane algorithm , and Ant algorithm. It would be better if we can change the input and output display by using graphics so that user gets an interactive output of graph. With these graph techniques and algorithms can be use for calculating polynomials time verification via hamiltonian cycles and NP problems. May be useful in Boost uBLabs.

1 0

GSOC 2013
by Niranjan Singh 13 Apr '13

13 Apr '13

Hello, My name is Niranjan , It was really a pleasure seeing a poject like this, and the ideas were also very good. I like to contribute to the vector ,making new matrix class. I would like to propose you my idea as below As *Graph theory* as a subject also and is becoming increasingly significant as it is applied to other areas of mathematics, science and technology. It is being actively used in fields as varied as *biochemistry (genomics), electrical engineering (communication networks and coding theory), computer science (algorithms and computation) and operations research (scheduling).* The powerful *combinatorial methods* found in graph theory have also been used to prove fundamental results in other areas of pure mathematics. And so is becoming an important subject in mathematics and algorithms also. So I was thinking to make a library to implement all the graph functions and its various properties and applications in it so that it can be used by student to understand the concept and it can also be used to solve problems like *Travelling Salesman problem, finding chromatic number, edge colouring. * Finding *shortest path , face colouring* e.t.c. With the help of small functions user can create a program which tackles the big problem. Lots of data-structures can also be implemented. Thanks. P.S. Sir, As my end semester are exams are on .. so its possible that i couldn't reply you fast but I do want to do this project. Exams will over on 30 april.. Thanks for you concern.

4 3

Niranjan Singh has invited you to open a Google mail account
by Niranjan Singh 13 Apr '13

13 Apr '13

I've been using Gmail and thought you might like to try it out. Here's an invitation to create an account. You're Invited to Google Mail! Niranjan Singh has invited you to open a Google Mail account. Google Mail is Google's free email service, built on the idea that email can be intuitive, efficient and fun. Google Mail has: *Less spam* Keep unwanted messages out of your inbox with Google's innovative technology. *Lots of space* Enough storage so that you'll never have to delete another message. *Built-in chat* Text or video chat with Niranjan Singh and other friends in real time. *Mobile access* Get your email anywhere with Google Mail on your mobile phone. You can even import your contacts and email from Yahoo!, Hotmail, AOL or any other web mail or POP accounts. Once you create your account, Niranjan Singh will be notified of your new Google Mail address so you can stay in touch. Learn more<http://mail.google.com/mail/help/intl/en/about.html>or get started<http://mail.google.com/mail/a-b2df9a7699-186fc6feff-CUGUbjU3pJUkhZ0HbL1w6xE…> ! Sign up<http://mail.google.com/mail/a-b2df9a7699-186fc6feff-CUGUbjU3pJUkhZ0HbL1w6xE…> Google Inc. | 1600 Amphitheatre Parkway | Mountain View, California 94043

1 0

JSON Parser GSoC 2013
by Stephan Bourgeois 13 Apr '13

13 Apr '13

Hi everybody, I am studying for an MSc in Computer Science at Oxford Brookes University. I have taken a compiler construction course this year and I am therefore interested in the JSON Parser idea for the Boost library. Open Source JSON parsers have already been implemented in C++ and in Java. Examples of Java libraries are: Gson, quick-json. Even if other libraries already exist, developers who are using Boost for their project will appreciate having a JSON parser within Boost. >From a compiler construction background, writing a JSON parser is not difficult. The JSON grammar is simple and the specification is easy to find. Potential difficulties can exist with robustness of error handling and recovery. The question is what data structure we should use to represent JSON objects, and how the user can access key/value pairs in those objects. (examples: Boost.PropertyTree, pre-existing C++ object, ...) Ideally we should offer validating and non-validating implementations. We should also offer JSON generation as well as parsing. Let me know what you think. Kind regards, Stephan.

10 14

Regarding proposal for GSOC
by Niranjan Singh 12 Apr '13

12 Apr '13

If I have to send my proposal for the poject in GSOC then whom should I mail. This link only or is there some other email. Thanks

2 1

JSON Parser GSoC 2013 (part 2)
by Stephan Bourgeois 12 Apr '13

12 Apr '13

Hi everybody, thank you for all your replies. I will try to summarize my impressions so far: 1. Although there is a solution in Boost.PropertyTree, a majority of users consider it to be incomplete, and quite a few developers are using external libraries at present. 2. As Bjorn Reese pointed out, a solution should offer several different interfaces. Some being already implemented in the protoc library. Michael Marcin suggested that something similar to JsonFX in C# be added to Bjorn's list. 3. There is already a Lexer implementation as part of the tokenizer API in protoc. This is how I imagine I could contribute to the project: 1. Collect requirements for functionality and interfaces. 2. review existing Java, C++, C#, Android library solutions. 3. Write a Parser with error recovery using the Lexer from the protoc library. Or adapt an existing parser if there is a good reason for doing so. 4. review JSON specific issues like type casting and how to deserialize arrays with multiple types. (see posts by Arindam Mukherjee and Michael Marcin on deserialization) 5. With the help of my mentor, implement the different interfaces, preferably in a modular way. (I know an example of this in compiler design using the visitor pattern for code generation) Please don't be offended as I only had the time to have a quick look at a single library. I understand that I need to look at a wider code base from other libraries and familiarize myself with general issues related to JSON and serialization. Bjorn Reese offered to mentor this project, please let us know if someone more "officially involved" with Boost needs to approve this. Please let me know what you think. Stephan

1 0

[gsoc-2013] Physics Library Abstraction Layer
by Preston Hamlin 12 Apr '13

12 Apr '13

Hello, my name is Preston Hamlin and I am a current CS undergraduate at the Florida State University. I was reading over the list of ideas provided and then referenced the source code libraries to get a feel for what was being asked for. I found the prospect of writing an abstraction layer for PhysX or related libraries to be interesting. I am somewhat familiar with THREE.js which is an abstraction of WebGL. In THREE.js there is a heavy amount of abstraction to the degree of providing geometric primitives, particles effect presets, simple materials and basic data structures (matrices, vectors, etc...). Given the platform that THREE.js operates on, such a degree is understandable. How much of the physics library is to be abstracted away in generic structures and functions? I would think that having a few simple primitives (cube, low-poly sphere, etc...) would be required since they are commonly used. However more advanced meshes like a torus might be better produced through using modifiers on said primitives rather than a built-in preset. I think the degree of abstraction is most important for modifiers, and I desire to get a feel for what is being looked for. I assume the end goal is to be able to produce Blender/Unity/3dsMax usable code as well as standalone utilities. Many thanks, -Preston

4 7

Re: [boost] JSON Parser GSoC 2013
by John Venarchick 12 Apr '13

12 Apr '13

> 2013/4/12 Michael Marcin <mike.marcin(a)gmail.com> > >> Arindam Mukherjee wrote: >> >>> In JSON we typically deal with maps and arrays. The arrays themselves >>> could >>> have arbitrary types (string, object, array, numeric, boolean, null) as >>> elements. The key types in the maps are always strings and the value types >>> in the maps could be anything that can appear in an array, including >>> another map or array. >>> >>> Due to this, I'd imagine being able to use Boost.Variant or Boost.Any in a >>> list and as a value_type in a map would help. >>> >>> I have been doing some work with JSON and boost by modifying the current 1.53 ptree. I created a type for use as the data_type of the ptree that represents the various options for JSON data (no value, null value, bool, integer, float, string). It uses boost::variant and visitors. I have it working for my purposes, but would want to make it a little nicer before submitting a patch. I also implemented property_tree::translator_ between to make it work with the spirit classic parser in json_parser_read and json_parser_write. I can post more if anyone is interested. I'm also working a generic way to load arbitrary structures from the created ptree. -John V

1 0

Re: [boost] [gsoc-2013] Physics Library Abstraction Layer
by Preston Hamlin 12 Apr '13

12 Apr '13

Considering that I have received positive feedback on this project, I shall likely make a proposal for it when the time comes. Before then, I would like to pin down the scope and breadth of this library. I am in possession of the PhysX SDK and have been reading it over to come up with some questions to ask so that (should my proposal be accepted) I would not have to be overly concerned later. On the first page of the SDK documentation, it is denoted that several distributable (PhysX) DLLs are required for applications on the Windows platform. Since Boost is a library project, I am unsure what to do with DLL files other than unceremoniously stick them in a Resources folder until the time comes, or have a script built into one of the library files. Perhaps contact Nvidia? I have worked a little with WinAPI, but not in the case of using an SDK. Since this would be a beginning point for the library, might I assume that complex mesh deformations and simulations (e.g. fluid dynamics) are beyond the scope of what should be done over the summer? Certainly one could implement certain models with effort, but I mean that the library will lack explicit methods for something of the following degree: http://www.youtube.com/watch?v=7Who8EpbvCY The same question goes for Multithreading. I will continue to tinker with and read over the SDK(s) and brush up on my matrix arithmetic. For the proposal, should I have a couple concrete classes? Or would pseudocode/feature descriptors suffice? More questions on the way, -Preston

1 0

Fwd: How to start contributing for Boost
by Hasit Bhatt 12 Apr '13

12 Apr '13

I'm 3rd year B.Tech student in Computer Science with quite good proficiency in c,c++ and python. I'm applying for Google Summer Of Code. How should I start?

2 1

Can C++11 INVOKE be defined with C++98 compilers?
by Vicente J. Botet Escriba 12 Apr '13

12 Apr '13

Hi, I would like to implement C++11 INVOKE on C++98 compilers. I have a C++11 implementation on boost/thread/detail/invoke.hpp based on the one of libc++. Could this be done or there is something that can not be done inherently? Best, Vicente

2 4

GSoC 2013: ODE or B-trees?
by Артём Серов 11 Apr '13

11 Apr '13

Hello! My name is Artem Serov. I would like to participate GSoC in Boost and I have a dilemma. I don't have enough information to choose one of two projects in boost: B-trees and Ordinary Differential Equations and want to find out which project matches me better. Before composing a formal application (with all formal requirements) for a project and deeper investigating of the environment i hope to get a piece of advice about my doubts. A few facts: - the 3rd year student of Moscow Institute of Physics and Technology (State University), Russia. - ~1 year exprerience of C/C++ industrial programming in Intel as intern. - no experience in open souce project contribution. Both projects are very challenging and interesting for me. ODE: I had cources of Theory of ODE, Computional Mathematics and Parallel programming (OpenMPI) at the university, so i am familiar with the topic. B-trees: The topic "B-trees" is located in the additional list of ideas for GSoC, so there is no concrete information except the title. Could someone (maybe who have put this in the ideas list) will tell anything: maybe requirements, any thoughts, why is it mentioned in GSoC ideas list, why isn't this idea in a main list or any information like in projects in the main ideas list, please? What would you advise me: to apply for the concrete project from the main list or to develop the project from additional list? What is preferable for the community? Could you give a kind of very brief estimation and comparisson of difficulty of both projects? -- Thanks, Artem Serov

3 3

[Graph] Efficiently finding cycles in a (multi)graph
by Louis Dionne 11 Apr '13

11 Apr '13

Dear Boost, I have implemented the `hawick_circuits` algorithm described in [1] for finding all the elementary circuits of a directed (multi)graph. I implemented it in a generic fashion, planning to propose it for inclusion in Boost. Boost.Graph currently contains the `tiernan_all_cycles` algorithm for finding all the cycles in a graph. However, it is undocumented and it has a time bound (for which no precise analysis is provided) that makes it unsuitable even for moderately sized graphs. In fact, I implemented `hawick_circuits` because `tiernan_all_cycles` did not cut it for my application. My implementation is available at [2]. It is licensed under the Boost license and was tested using the test suite at [3]. I am posting this message for two reasons: 1. To know whether there is interest for such an algorithm in Boost.Graph. 2. To spread the word that this implementation exists for C++, to save people the hassle of reinventing the wheel before Boost.Graph officially provides it. I will create a Trac ticket if there is interest, but note that I won't submit a patch immediately because I am in deep juice with school right now. Regards, Louis Dionne P.S.: More information on the test suite can be found at [4]. [1] http://bit.ly/12LMwIr [2] http://bit.ly/10MSAhW [3] http://bit.ly/YdLHVn [4] http://bit.ly/16SvlW6

1 1

GSoC 2013
by İslam Yaşar 11 Apr '13

11 Apr '13

Hi, I'm a graduate student and work since 2 years as a C++ developer on Unix-like systems (FreeBSD and Linux), i want to implement a library for Boost. It is a thread-safe queue that is implemented on disk. You can use it as a IPC. Is anyone interested to mentor this project ? Thanks Islam Yasar

3 2

Boost.uBLAS for GSoC 2013
by Sathyam Vellal 11 Apr '13

11 Apr '13

Hello. I'm Sathyam Vellal, a 3rd year undergraduate student pursuing Computer Science. I'm interested in contributing to Boost.uBLAS in unifying the vector and matrix representations. I've looked at the docs ( http://www.boost.org/doc/libs/1_53_0/libs/numeric/ublas/doc/index.htm) to get an idea of the library. Does anyone have any suggestions on how I could proceed more and connected to this project? Thank you -- Sathyam M Vellal Computer Science and Engineering, P.E.S. Institute of Technology, Bangalore, India.

2 1

GSoC 2013 : Boost.uBLAS
by Jignesh Jain 11 Apr '13

11 Apr '13

I am interested in this project for Gsoc, which involves unification of vector and matrix libraries. I have been browsing boost ublas code, but have no idea about how should I begin as I am totally new to Boost. Can Someone please guide me some approach for getting started in this project. Thank you.

1 0

Boost.Multiprecision GSoC 2013
by Nikhar Agrawal 10 Apr '13

10 Apr '13

Hi, My name is Nikhar Agrawal. I'm a 2nd year Computer Science student. I'm interested to work for Boost in GSoC and I found the project Boost.Multiprecision really really interesting. Any pointers on how should I go about it? Meanwhile, I'll go through this book:- http://www.loria.fr/~zimmerma/mca/mca-cup-0.5.1.pdf Thanks

3 3

[thread] using Boost.Thread with move-only arguments
by Fernando Pelliccioni 10 Apr '13

10 Apr '13

Hi Vicente, I have an issue with Boost.Thread contructor using move-only type arguments. The "Thread Constructor with arguments" section of the documentation says ... template <class F,class A1,class A2,...> thread(F f,A1 a1,A2 a2,...); Preconditions: F and each An must by copyable or movable. http://www.boost.org/doc/libs/1_53_0/doc/html/thread/thread_management.html… "An" could be copyable or movable, but the following example fails to compile... error: use of deleted function 'A::A(const A&)' //Begin code ------------------------------------ #include <boost/thread.hpp> struct A { A() = default; A( const A& other) = delete; A& operator=( const A& other) = delete; A( A&& other) = default; A& operator=( A&& other) = default; void run() {} }; void f( A a ) { a.run(); } void exec_boost() { A a; boost::thread t( f, std::move(a) ); t.join(); } //End code -------------------------------------- I think we should modify ... template <class F,class A1> thread(F f,A1 a1,typename disable_if<boost::thread_detail::is_convertible<F&,thread_attributes >, dummy* >::type=0): thread_info(make_thread_info(boost::bind(boost::type<void>(),f,a1))) { start_thread(); } ... by ... template <class F,class A1> thread(F&& f,A1&& a1,typename disable_if<boost::thread_detail::is_convertible<F&,thread_attributes >, dummy* >::type=0): thread_info(make_thread_info(boost::bind(boost::type<void>(),f,boost::forward<A1>(a1)))) { start_thread(); } ( In the all family of related ctors in "boost/thread/detail/thread.hpp" ) But, I don't know if boost::bind is forwarding its arguments I think we should modify Boost.Bind to be similar to std::bind. Is there any reason why Bind has not changed so far? Have you found any limitations? If Bind can not be changed, I think Boost.Thread should use some other tool similar to Bind to make forwarding parameters. GCC Libstdc++ use a Bind-lite tool for std::thread. I can work on this if you want. What do you think? Environment: Windows 7 64bits MinGW - GCC 4.7.2 MSVC 2010 MSVC 2012 Boost 1.53.0 Non-Variadic template code branch Regards, Fernando Pelliccioni

2 10

Re: [boost] GSoC 2013 : Implementing Dynamic Programming Algorithms
by Aman Gupta 09 Apr '13

09 Apr '13

Hi, Thanks for the reply. The first thing I want to implement is a Integer-Knapsack problem which takes in a precision error(i.e. an epsilon which is a indicator of the allowed error in the input) The input parameters are Weights, Values and the error. The first thing to do is massage the values according to the error (e.g. if the value is ~10^4 divide it down to ~10^3 according to the error) then run the a dynamic programming algorithm, which has a running time of O(n^3/e) (n : number of items, e :error) Hence, an arbitrarily close approximation is possible with a running time trade-off. Also, a vanilla variety of the knapsack problems(with small instances) which gives the optimal solution can also be implemented. I also want to implement the assignment problem, which also arises in a lot of situations in real life ( http://en.wikipedia.org/wiki/Assignment_problem) using the Hungarian Algorithm. I realize this requires effort but I'm really interested in adding a feature to the boost library. Regards, Aman Gupta

1 0

GSoC 2013 - Boost registration?
by Paul A. Bristow 09 Apr '13

09 Apr '13

http://www.google-melange.com/gsoc/events/google/gsoc2013 March 18: 19:00 UTC Mentoring organizations can begin submitting applications to Google. March 29: 19:00 UTC Mentoring organization application deadline. Just checking that Boost has/is registering? Paul --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow(a)hetp.u-net.com

3 3

GSoC 2013 : Implementing Dynamic Programming Algorithms
by Aman Gupta 09 Apr '13

09 Apr '13

Hi, I am Aman Gupta from BITS-Pilani India, I am a 4th year Undergraduate Student majoring in Computer Science. I've noticed that boost does not contain dynamic programming algorithms like Integer-Knapsack, Longest Common Sub-sequence. I was thinking that implementing this would be a good project. I realize most of these problems are non-polynomial in complexity but I was thinking implementing them to a certain size will not be problem. Also, since some of the dynamic programming algorithms have alternate greedy algorithms which compromise on correctness but are faster, implementing those can be useful. The knapsack problem(which comes in a lot of situations) for which there exists an algorithm which gives an arbitrarily close approximation to the optimal value can be implemented. ( e.g. f(weights,values,error) could be the function, the user has a trade-off between error and computation time). I think it could be useful in situations where the user just decides the error he wants or perhaps how fast he wants his computation. Please let me know if this seems like a valid project idea to you! Thanks, Aman Gupta

2 1

[move] move's is_convertible does not compile with aligned data
by Tobias Loew 09 Apr '13

09 Apr '13

The implementation of move's is_convertible-type has the same problems on MSVC with aligned types as reported for type_trait's is_convertibale in Ticket #5271 https://svn.boost.org/trac/boost/ticket/5271 As type_trait's implementation has been fixed, a simple solution would be letting the user (and not the code in move.hpp:41) [not] define BOOST_MOVE_AVOID_BOOST_DEPENDENCIES regards Tobias -- View this message in context: http://boost.2283326.n4.nabble.com/move-move-s-is-convertible-does-not-comp… Sent from the Boost - Dev mailing list archive at Nabble.com.

2 2

boost/thread.hpp UBICOM processor
by Aswin Paranji 09 Apr '13

09 Apr '13

Hi, I'm trying to compile boost library 1.53 using the ubicom toolchain provided by our embedded partner. During compilation of the boost library there was a linker error (_sync_fetch_add). This was overcome by adding define statement BOOST_SP_USE_PTHREADS, along with FENCED in user-config.hpp. The application was linked properly. Running the application within the embedded box exits with a segmentation fault happening before printing the first line of the main function. This was narrowed down to boost/thread.hpp header. Removing the header allows the application to print the first line and return. The problem I have is, I do not have GDB/GDBServer to assist me. Also, i could not use back_trace as the toolchain does not have the necessary header files. Can anyone tell me what might be causing the problem. I'm sure there is some sort of conflict between the processor architecture with thread or the toolchain which was used to compile boost. The GCC version is 4.4.1 Appreciate if someone can help me on this. Thanks, Aswin Paranji

2 2

[phoenix] doesn't compile with gcc 4.9 snapshot
by Nathan Ridge 08 Apr '13

08 Apr '13

Hi, I'm trying to compile some code that uses Boost with a GCC 4.9 snapshot, but something seems to be going badly wrong. Something as simple as a file containing the include: #include <boost/phoenix/core.hpp> fails to compile, with the following errors: In file included from boost/proto/matches.hpp:42:0, from boost/phoenix/core/domain.hpp:12, from boost/phoenix/core/actor.hpp:17, from boost/phoenix/core.hpp:15, from test.cpp:1: boost/proto/traits.hpp: In instantiation of 'struct boost::proto::tag_of<boost::proto::_>': boost/phoenix/core/domain.hpp:25:18: required from here boost/proto/traits.hpp:241:42: error: no type named 'proto_tag' in 'struct boost::proto::_' typedef typename Expr::proto_tag type; ^ In file included from boost/phoenix/core/argument.hpp:15:0, from boost/phoenix/core.hpp:17, from test.cpp:1: boost/phoenix/core/terminal.hpp: In instantiation of 'struct boost::phoenix::detail::placeholder_idx<boost::proto::make<boost::is_placeholder<boost::proto::_value>()> >': boost/phoenix/core/terminal.hpp:122:18: required from here boost/phoenix/core/terminal.hpp:115:16: error: 'value' is not a member of 'boost::proto::make<boost::is_placeholder<boost::proto::_value>()>' struct placeholder_idx ^ The exact same code compiles fine with GCC 4.8 and earlier. Any ideas as to what might be going wrong? Thanks, Nate

1 1

Boost.Graph for GSoC
by Daniel Mitchell 08 Apr '13

08 Apr '13

Hi all, I'm a researcher and graduate student at the University of Texas at Austin and I'm interested in doing a Boost.Graph project for GSoC. I was thinking of doing one or more planar graph algorithms, in particular the multiple source shortest paths algorithm for planar graphs described at http://courses.csail.mit.edu/6.889/fall11/lectures/L11.html The algorithm assumes that the sources are located on a single face. It proceeds by calculating the shortest paths tree from one source and then minimally modifying the tree to efficiently get the shortest paths from the other sources. The algorithm uses a dynamic tree data structure, so developing that data structure would be part of the project too. I've made prior contributions to BGL. I worked with Doug Gregor to implement color maps for the breadth first and Dijkstra's algorithms. One of my current areas of research is movement ecology, where planar graphs can be used to study movement. Let me know what you think. Daniel Mitchell

3 4