January 2017 - Boost - lists.preview.boost.org

[regression] Are tests not running?
by Andrey Semashev 31 Jan '17

31 Jan '17

Hi, Looking at the regression tests table, it looks like the latest regression tests were run last Saturday, November 26th. It seems inlikely that all testers decided to stop running 4 days ago or are running tests for that long. Is there some problem with the regression infrastructure?

6 16

Re: [boost] [sort] pdqsort
by Orson Peters 30 Jan '17

30 Jan '17

Hi, Orson here, I'm the original author of pdqsort. As discussed in this long github issue ( https://github.com/boostorg/sort/issues/11) I have full copyright on the pdqsort implementation (which is freely available under the zlib license on https://github.com/orlp/pdqsort) and am willing to provide it to Boost free of charge under the Boost license. @Marc Glisse, since your 2015 experiment the code has changed, and pdqsort now uses a technique called block quicksort. It's substantially faster for branchless comparisons (50-80%), but it seems that on some systems it also give significant (5-15%) slowdowns for branchful code. I will revert pdqsort's default implementation to the old version, except when the passed in type is built-in, and the comparison operator is std::less / std::greater, then I will use the branchless code. I will provide the branchless code explicitly as well, under a different function name, for experienced users that know their comparison function is branchless. The old implementation was never slower than std::sort in my experiments, for the record. If any of you have further questions, please do let me know.

2 1

[rational] Compilation issues on VS2013+
by Raffi Enficiaud 30 Jan '17

30 Jan '17

Hi Boost, I am having the following issues when compiling boost.test on develop: http://www.boost.org/development/tests/develop/developer/output/teeks99-09e… It seems that the problem is related to rational (worked for the 1.63 release), but I am not sure. Thanks! Raffi

1 0

[Beast] WebSocket is now feature-complete!
by Vinnie Falco 29 Jan '17

29 Jan '17

I'm pleased to announce that Beast 1.0.0-b24 has been pushed to the master branch, making Beast's WebSocket implementation feature complete with respect to RFC 6455 ("The WebSocket Protocol") and draft-ietf-hybi-permessage-compression-28 ("Compression Extensions for WebSocket). The latest version adds support for the "permessage-deflate" extension, which allows WebSocket messages to be compressed using the deflate algorithm. Beast comes with a header-only port of ZLib to modern C++11, no additional library is necessary to use the extension. The library now passes every test in the Autobahn TestSuite including compression tests. Official Repository https://github.com/vinniefalco/Beast/ Documentation http://vinniefalco.github.io/beast/index.html Autobahn TestSuite results http://vinniefalco.github.io/autobahn/index.html Thanks! -- Follow me on Github: https://github.com/vinniefalco

3 2

[Boost.Lockfree] Interest in a ringbuffer class
by Raphaël Londeix 27 Jan '17

27 Jan '17

Hi everyone, Is there any interest in having a lower level ringbuffer type in Boost.Lockfree ? The containers provided today in the library can only store homogeneous types of fixed size. At work, we have refined many implementations of ringbuffer classes that allow one to push/pop raw buffers. Before investing some time into a pull request, I wanted to know if there's some interest in such a type. The interface looks like this: template<typename Self, typename... Policies> class basic_ringbuffer { public: typedef ... buffer; public: buffer* start_write(size_t); void commit_write(size_t); void commit_write(buffer const*); buffer const* start_read(); void commit_read(size_t); void commit_read(buffer const*); public: void reader_reset(); void writer_reset(); private: buffer* _ring() { return static_cast<Self>(*this).ring(); } }; Of course, these are only the lowest level methods, one could implement push()/pop() or other higher level methods. Beyond the fact that this API provides by default zero-copy semantics, this interface is also agnostic to the actual owner of the memory, which allow users to implement custom ringbuffers easily (shared memory, mmap trick to map the beginning at the end, etc.), for example: class custom_ringbuffer : basic_ringbuffer<custom_ringbuffer, SomePolicies...> { buffer* ring() { return some_memory; } }; The usage would look like: custom_ringbuffer rb; // the write side if (auto buf = rb.start_write(100)) { memcpy(buf.data, "some data", 9); rb.commit_write(9); } // the read side if (auto buf = rb.start_read()) { std::cout << std::string_view(buf.data, buf.size) << std::endl; rb.commit_read(buf); } What do you think ? Did I miss something in boost itself ? Should I give a more complete overview of the API ? Cheers, -- Raphaël Londeix http://hotgloupi.fr

6 13

[outcome] Requesting second pre-review of Boost.Outcome tutorial
by Niall Douglas 27 Jan '17

27 Jan '17

Dear Boost, Last November you gave me some valuable feedback on rewriting the tutorial for Boost.Outcome, a C++ 14 library providing a factory and family of policy driven lightweight monadic value-or-error transports. I recently came out of employment contract and so have had the time to rewrite Outcome's Tutorial (anyone interested in employing a remote working C++ consultant please do get in contact). I'd appreciate feedback on https://ned14.github.io/boost.outcome/: 1. Is this rewritten tutorial too simple, too detailed or about right? 2. Do you feel you could use Outcome in your own code without difficulty after reading it? If not, what information is missing? 3. Are you persuaded of the merits of using Outcome after reading it? If not, what else do I need to say? The first FOUR sections at https://ned14.github.io/boost.outcome/ have been rewritten by me to hopefully reflect the feedback you gave me last November, so these have been refactored so far: * Description, Prerequisites and Support * Installing, building and testing Outcome * Quick start for the expert * Tutorial The package repo install methods haven't been implemented yet, these are marked in the docs as todo. These sections have yet to be refactored: * Frequently Asked Questions * Advanced usage: Outcome as a Monad * Outcome and the upcoming std::expected<T, E> Many thanks in advance for any feedback received. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

7 20

Why is this message in Spam?
by Olaf van der Spek 26 Jan '17

26 Jan '17

Hi, > Why is this message in Spam? It has a from address in microsoft.com but has failed microsoft.com's required tests for authentication. Learn more I think this list isn't configured properly.. and this causes some messages to end up in Spam consistently.. Who's responsible for the mailing list infrastructure? Gr, -- Olaf

7 12

[sort] pdqsort
by Steven Ross 26 Jan '17

26 Jan '17

I've been reviewing a proposed addition to Boost.Sort called pdqsort. When using a general-case partitioning algorithm, it performs comparably to std::sort in my testing (Windows and Linux). When used as a replacement for the std::sort fallback in spreadsort, using a branch-reducing optimization it is ~20% faster on ints and floats, and ~40% slower on strings. The concerning part is that any comparison operator that involves a branch may have a comparable slowdown to the string case. pdqsort is also significantly faster than std::sort on mostly-sorted data and some other (relatively common) special cases. This makes little difference when used as a fallback by spreadsort, but does make a difference when used on its own. I see these possible things to do with pdqsort: 1) Reject it completely as too similar to std::sort, which is already highly optimized. 2) Add it as another optional library in the Boost.Sort library. 3) Add it to Boost.Sort, and only use it as a fallback for ints and floats. 4) Add it to Boost.Sort, use it as a fallback for all of spreadsort, and have pdqsort itself only use its branch-reduction optimization on ints and floats. 5) #4, except eliminate the branch-reduction optimization completely from pdqsort for simplicity. Does anyone have a strong preference, advice, or comments on how they use Boost.Sort? Should I run a mini-review for pdqsort?

5 9

Re: [boost] bad include guards
by Larry Evans 26 Jan '17

26 Jan '17

On 08/21/2007 03:11 PM, Steven Watanabe wrote: > AMDG > > The following pairs of files have identical include guards > > ./boost/concept_check/general.hpp ./boost/concept/detail/msvc.hpp > ./boost/concept_check/msvc.hpp ./boost/concept/detail/msvc.hpp > ./boost/filesystem/cerrno.hpp ./boost/cerrno.hpp > ./boost/interprocess/containers/tree.hpp > ./boost/interprocess/containers/detail/tree.hpp . . . Steven, do you have a program to detect files with identical include guards. I could you one at the moment :) -regards, Larry

4 6

Boost XML Serialization Stability across library versions
by Adlai Shawareb 26 Jan '17

26 Jan '17

Hi, We want to generate an Serialized XML archive now, and be able to deserialize it and get the same data it in the future. Is Boost XML Serialization guaranteed to be compatible across versions of the library? The XML serialized archive would be stored in a memory chip and, in the future, read back and deserialized. I don't believe that would impact the answer, but want to offer that. Adlai

6 12