January 2017 - Boost - lists.boost.org

[SmartPtr] alternative to enable_shared_from_this
by Roberto Hinz 16 Jan '17

16 Jan '17

Dear Boost Community, I've been following this list for a time with the desire of contributing to Boost. It seems to me a good idea to start with small and easy to review proposals, before attempting anything big. So my first shot is shared_handle class, an alternative to enable_shared_from_this that doesn't suffer from multiple inheritance issues. It also provides the capability of instantiating a shared_ptr to a member object. doc: https://robhz786.github.io/smart_ptr/shared_handle.html source: https://github.com/boostorg/smart_ptr/compare/develop... robhz786:develop Unfortunately, the implementation produces at least one regression: clang 3.1 crashes when it compiles cpp11_pointer_cast_test.cpp with -std=c++11. There may be others since I couldn't test in all compilers used in the regression tests. Nevertheless, is there any interest? Kind regards, Roberto Hinz

9 11

Logging Documentation issue
by jack nade 13 Jan '17

13 Jan '17

There is no visible Design Figure in the Design Overview at http://www.boost.org/doc/libs/1_63_0/libs/log/doc/html/log/design.html using Chrome or IE. The Design Overview refers to it 4 times.

3 3

shutdown ticket-system on svn.boost.org
by Oliver Kowalke 13 Jan '17

13 Jan '17

Hi, I encountered that the ticket system at svn.boost.org is not up-to-date: components coroutine2 and fiber are missing. At least the reports created for coroutine2 and fiber do not know the components (not in drop-down list). I would vote for shutdown of the ticket-system in favour of github - for me it makes no sense to manage two bug reporting systems. best regards, Oliver

9 19

Stringify
by Olaf van der Spek 13 Jan '17

13 Jan '17

On Wed, Jan 11, 2017 at 8:03 PM, Roberto Hinz <robhz786(a)gmail.com> wrote: > On Wed, Jan 11, 2017 at 3:40 PM, Roberto Hinz <robhz786(a)gmail.com> wrote: > > > > > Hi, > > > > I've been working in a format library that contains a function template > > called appendf that can do this: > > > > std::string str("blabla"); > > boost::stringify::appendf(str) () (" AAA ", 25, " BBB ", {255, "x"}); > > assert(str == "blabla AAA 25 BBB ff"); // ( 255 formated in > > hexadecimal ) > > > > This library - I call it Boost.Stringify - is in very early stage of > > development and I though it would be to premature to mention it now. But > > given the repercussion in this thread so far, I changed my mind. So I > will > > soon publish it in github and I will start to write a brief documentation > > so that I can soon present here to gauge interest. > Hi, I think your library deserves it's own thread. ;) How does it compare to for example fmtlib? Gr, Olaf

2 1

gregorian::date limited to years 1400 - 9999
by Viacheslav Usov 12 Jan '17

12 Jan '17

Conceptual question: why is the Gregorian calendar's time range limited to years 1400 - 9999? If this or any of the below has already discussed in some public forum (which I have failed to found, gmane's rebuild being possibly the culprit), kindly point me it to it. I am specifically interested in 1400. It seems completely arbitrary to me. It predates the time the calendar was introduced by almost two centuries, so that cannot be the reason why it cannot support earlier times. ISO 8601 allows the entire 0 - 9999 range, even though dates earlier than 1582-10-15 are subject to an agreement among the parties involved. Extending the proleptic Gregorian calendar all the way back to year 0 seems like one of the very few things that make sense. So my question is, is there any technical problem in supporting the entire 0 - 9999 range? If not, then my next question would be, what else would speak against supporting the 0 - 9999 range? If you expect some compatibility could be broken by that, then perhaps the low end could become a template parameter with the default at 1400? Cheers, V.

5 7

Tagged pointer library?
by Hans Dembinski 11 Jan '17

11 Jan '17

Hi all, I recently learned about tagged pointers, https://en.m.wikipedia.org/wiki/Tagged_pointer. Over the weekend, I was working on a tiny_string implementation which has a width of only one pointer. Essentially it's a null-terminated char* array managed by std::unique_ptr. On my system, std::string is four times larger (probably 8 byte for the pointer, 8 byte for the length, 8 byte for the allocator and 8 byte for … good measure?). With a tagged pointer, I could even implement a small-string optimisation in that tiny_string, by storing small strings with up to 7 bytes in the memory that otherwise holds the address to the char* array. The tag then allows you to discriminate between the two uses of the memory. You may rightfully question whether it is really practical to do this, but it is certainly one possible application of a tagged pointer. Looking into boost, I found that the lockfree library has a tagged pointer implementation in its "detail" namespace. The implementation is nice, because it falls back to a safe version with a separate tag field on systems which do not follow the required alignment conventions. Wouldn't it make sense to make the tagged pointer accessible by users, i.e. place it in a separate library (or in the existing smart_ptr library)? Best regards, Hans

3 3

tensor library
by Paulo Pereira 09 Jan '17

09 Jan '17

Hello everyone, I am new in the community. I would like to ask, how interesting would be to have in boost a new container with multiple indices (tensor) using variadic templates. I know that Boost.MultiArray library already exists in boost. However I think it does not fulfill the readability and friendliness of something like: newlib::tensor<3,double> t(2,2,3); t(0,0,0)=1; also with extra possibilities similar to Matlab/Octave t(newlib::all,1)=0; I would also like to add mathematical operations t1=t2+t3; using lazy evaluation like in armadillo package or in the David Abrahams and Aleksey Gurtovoy book (C++ Template Metaprogramming: Concepts, Tools, and Techniques from Boost and Beyond). I fell that something similar to this is very important for numerical calculations and somehow it is not a standard in C++ (maybe I am missing something). Best regards, Paulo

4 3

Interested in contributing
by vivekroyandroid＠gmail.com 09 Jan '17

09 Jan '17

Sir, I would like to contribute in whichever way possible. I have a decent amount of knowledge in C/C++, but I am new to contributing to open-source. If someone might guide me and tell me how to proceed, I would appreciate it. Thank You. Vivek Roy --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

2 1

Crosscompiling context and fiber without multithreading
by Vitaly Osipov 09 Jan '17

09 Jan '17

Dear all, I am developing for a PowerPC target under real-time os und studying a possibility to employ fibers. They seem to be a good fit, since we have a single thread but several entities, which could in fact be considered parallel. Data exchange through channels also is relevant for the project. The problem is, that though we are provided with a quite recent compiler gcc 4.7.4 with c++11 support, the multithreading part of std is missing, since the OS has its own implementation of multithreading functionality different from pthreads and std::thread. Is there any chance to compile context and fiber without multithreading support? Or it is used under the hood and can not be avoided? Best, Vitaly

2 3

Boost 1.63 vs GCC 5.4.0
by Andrzej Krzemienski 06 Jan '17

06 Jan '17

Hi Everyone, I am trying to determine if Boost 1.63 (or 1.62) will work on GCC 5.4.0 on Linux. I can see in release notes that Boost 1.62 was tested on Linux with "GCC 5.4.1" (I am not aware that v 5.4.1 even exists). However in release notes for 1.63 there is no GCC 5.4 listed for Linux (only for Android). Does anyone know if there is any reason to suspect that Boost 1.63 may have problems on Linux with GCC 5.4.0 ? Regards, &rzej;

4 3