March 2018 - Boost - lists.preview.boost.org

[RFC] unique_val (handles, IDs, descriptors...)
by Miguel Ojeda 15 Oct '18

15 Oct '18

Hi all, While working on a medium-sized project, I had to wrap quite some handles, IDs, descriptors, etc. into RAII classes. Somehow, writing those and making them movable seemed more verbose than needed. I searched in the STL and Boost for something already implemented, but nothing seemed to come up. I decided to write my own prototype to see whether I was missing something important; but actually using it in my project felt quite natural and simple. At this point, I asked for a second opinion from a well-known expert (thanks Scott!) and he agreed that it seemed like a reasonable idea but, of course, he would also be surprised if others haven't come up with something similar. So I polished it up a bit and uploaded it to: https://github.com/ojeda/unique_val Below you have the introduction inlined [1]. If you have the time, please take a look (specially to the Rationale) and let me know your comments. The code itself is very short and straightforward. If you already have this class/template in Boost somewhere that I missed, please let me know! Otherwise, if you think this could be interesting to put into Boost in some existing library (not sure which) or as a new tiny library, I would be glad to make the effort to expand it, clean it up, etc. Thank you! Miguel [1] """ A single-header header-only library for representing unique values (handles, IDs, descriptors...) that default to a value on move for C++11, C++14 and C++17. It simplifies writing resource-owning classes (RAII) with move semantics support for resources with handles, IDs or descriptors that are supposed to be used as unique values, not as unique pointers. For instance, it can be used with non-pointer handlers (like the `int` file/socket descriptors in POSIX) and with opaque pointer handlers (like the `HWND` handles in the Win32 API). As such, it can be used as an alternative to `std::unique_ptr` with or without custom deleter, specially for resources that use a non-pointer type or resources that are not simply heap-allocated. It does not have any overhead compared to using the original type. It supports booleans, integers, pointers and enumerations. The default value can be different than the default-constructed value (i.e. different than 0, nullptr, etc.). """

7 13

[type_traits] revisiting is_virtual_base_of
by John Maddock 10 May '18

10 May '18

There is a problem with is_virtual_base_of in that it doesn't work especially well - most of the time it works OK, but can be made to fail catastrophically in a few situations, see for example: https://svn.boost.org/trac10/ticket/11323 Now... it turns out that with C++14 and later we now have is_detected which can test the validity of almost any expression... and I'm wondering if there's a cunning way to press this into action to create a more accurate is_virtual_base_of, but I'm having trouble coming up with anything. Any bright ideas? Note that an explicit static_cast can't be tested for, though we can use is_constructible if required. Thanks in advance, John. --- This email has been checked for viruses by AVG. http://www.avg.com

2 3

Reducing dependencies
by Antony Polukhin 08 Apr '18

08 Apr '18

Hello, I've noted that many libraries depend on container_hash while it depends only on a single header from Boost.Detail library https://pdimov.github.io/boostdep-report/master/module-levels.html Can we move "boost/detail/container_fwd.hpp" into Boost.Core or into Boost.ContainerHash? This will break the circular dependency, move the Boost.ContainerHash on level 4 and improve levels of Intrusive, Container, CircularBuffer, Unordered, Algorithm, Test, TypeIndex, Any, Function, Functional, ScopeExit, Signals, MultiArray, Variant. -- Best regards, Antony Polukhin

6 12

[1.67] Missing include in Boost.Graph
by Denis Davydov 03 Apr '18

03 Apr '18

Dear all, 1.67.0.beta1 has a regression of missing include in Boost.Graph which prevents deal.II https://www.dealii.org <https://www.dealii.org/> project from building. Even though the file was not modified since 2015, it’s a regression as we have not had this issue with previous versions of boost. The one-liner fix is here https://github.com/boostorg/graph/pull/104 <https://github.com/boostorg/graph/pull/104> I would appreciate if you could merge it prior to 1.67.0 release. ps. https://github.com/boostorg/graph/pull/101 <https://github.com/boostorg/graph/pull/101> is a better way to fix it but as suggested in https://svn.boost.org/trac10/ticket/13505#comment:4 <https://svn.boost.org/trac10/ticket/13505#comment:4> this might not get merged into 1.67.0, which would be pity. Regards, Denis

3 2

Build breaking changes
by Antony Polukhin 03 Apr '18

03 Apr '18

Hello, I've just noted that building Boost from master stopped working: [00:16:51] b2.exe -j4 toolset=gcc --layout=system variant=release architecture=x86 link=static --with-program_options --with-filesystem --with-system --with-test --with-atomic --with-thread --with-timer --with-date_time --with-chrono --with-regex --with-random --with-context stage [00:16:52] Performing configuration checks [00:16:52] [00:16:52] - default address-model : 32-bit [00:16:53] - default architecture : x86 [00:16:53] - symlinks supported : yes (cached) [00:16:54] - C++11 mutex : yes [00:16:55] - has_icu builds : no [00:16:55] - BOOST_COMP_GNUC >= 4.3.0 : yes [00:16:55] - C++11 mutex : no [00:16:56] - has_icu builds : no [00:16:56] - BOOST_COMP_GNUC >= 4.3.0 : no [00:16:56] error: Name clash for '<pstage\lib>libboost_atomic.a' [00:16:56] error: [00:16:56] error: Tried to build the target twice, with property sets having [00:16:56] error: these incompatible properties: [00:16:56] error: [00:16:56] error: - <address-model>64 [00:16:56] error: - <address-model>32 [00:16:56] error: [00:16:56] error: Please make sure to have consistent requirements for these [00:16:56] error: properties everywhere in your project, especially for install [00:16:56] error: targets. [00:16:56] [00:16:56] Command exited with code 1 Previously this build command worked fine. Full log available at https://ci.appveyor.com/project/apolukhin/boost-cookbook/branch/second_edit… Is this an intended breaking change?

12 21

Debate until Decision
by James E. King, III 03 Apr '18

03 Apr '18

I'm going to float some thoughts about how the group makes decisions below. I'm not suggesting a specific course of action, but I would like people to think about how these ideas can be used to improve Boost development. Recent discussions about default address-model, and past discussions about issues like cmake and CI build improvements (like whether or not to purchase more build slaves from travis or appveyor), and topics about legal use of code snippets leads me to the following conclusion: 1. We are geographically, culturally, and vertically distributed and have highly varying viewpoints. 2. We are all very good at expressing our viewpoint about various topics. 3. We collectively need to improve our skill at coming to a group decision and moving forward. Folks generally have individual control over their respective maintained repositories in the current structure. There are many topics that generally affect all of the repositories, such as the ones I mentioned previously, that affect all of us. At my most recent startup company we implemented an effective "Debate until Decision" mechanism by which: 1. The issue is discussed. (We do this well) 2. A decision is made and implemented (We do not seem to do this well): Not everyone will be happy with the outcome. Folks should not attempt to undermine the outcome after decision. The issue can be raised again if new information is brought to light. How is a decision made? Typically in a meeting where people vote face-to-face. For a group like this, voting would be done online with a time limit. Who can vote? That's a good question. In this environment it may fall to the group of folks who are maintainers of an affected repository. I don't necessarily have a good answer for that, as I am relatively new to the group. Some folks have been here for 10+ years and may feel differently that seniority should have a say. I don't know. What are the benefits of this? Simply: *Things get done.* Look at the cmake discussion, or the CI build job discussion, or the current address-model discussion. These need to be finished, agreed to, and implemented (or not), and everyone moves forward. The Apache Software Foundation maintains a fairly well documented structure for each project, with a project management committee consisting of a chair and members who have equal voting rights, however said group's primary function is to identify and grow the team of developers who have commit privileges. The committers as a whole vote on issues that are project-wide, like "should we split into multiple repositories", or "should we move to another source code control environment"? This mechanism could also be applied to how folks are brought into maintainer status for a repository, or how a library enters the CMT. If you look at the bug backlog in trac for some one the libraries, it's pretty clear that they are not properly maintained. The issue may be discussed, but then dropped and forgotten, and no action is taken to remedy it. When repositories have pull requests open for years, or a backlog of 100 defects going back 10 years, there's a very unhealthy problem for that project that needs to be addressed. It may be a similar structure that will work well here. Folks may disagree. This will be debated, certainly, as folks probably have varying opinions about how much structure is needed or desired. I simply hope this leads to an effective discussion about how this group identifies deficiencies, solves them, and moves forward. Any group large or small can benefit from examination and improvement of processes, including ours. Thanks, Jim

5 5

[ineterprocess] managed mapped file - recovery from process failure
by Assaf Ben-amitai 02 Apr '18

02 Apr '18

Hi, I have noticed that if a process is killed while it's allocating memory from a managed file, later loading of the managed file may segfault (even in read-only mode). A. What is the expected behaviour in this case? B. Should such a segfault be considered as a bug or is it the case that the user is not supposed to use the managed file after process failure? C. If unsupported, any pointers on efforts to improve this? Thx, Assaf

2 1

[cmake] Where to discuss CMake in Boost?
by Mateusz Loskot 02 Apr '18

02 Apr '18

Hi, TL;TR: Which is the right mailing list to discuss CMake in Boost? Numerous CMake discussions happened here lately which, I think, were mostly about the grand Boost vs CMake strategy, politics. The boost-cmake ml has been quiet for years now, according to its archives. https://lists.boost.org/mailman/listinfo.cgi/boost-cmake Whereas, there has been lots of activity in development of CMake for various aspects: to use CMake inside Boost. to use CMake to build Boost, etc. That is https://github.com/boost-cmake where, in particular, I'm interested in https://github.com/boost-cmake/bcm Where can I ask *technical* questions about those to avoid polluting wrong list? Best regards. -- Mateusz Loskot, http://mateusz.loskot.net

5 10

[gil] Locator bug in MSVC release builds
by Christian Henning 31 Mar '18

31 Mar '18

Hi all, there is this nagging bug (#49) with boost::gil which I have a hard time figuring out. Please consider the following code: #define _CRT_SECURE_NO_WARNINGS 1 #include <boost/gil/gil_all.hpp> #include <boost/gil/extension/io/bmp_all.hpp> using namespace boost::gil; int main() { using ref_t = const bit_aligned_pixel_reference<boost::uint8_t, boost::mpl::vector3_c<int, 1, 2, 1>, bgr_layout_t, true>; using ptr_t = bit_aligned_pixel_iterator<ref_t>; using pixel_t = std::iterator_traits<ptr_t>::value_type; using bgr121_image_t = image<ref_t, false>; ////////////////////// // works ////////////////////// { bgr121_image_t a(10, 10); fill_pixels(view(a), pixel_t(0, 0, 1)); auto loc = view(a).xy_at(0, 0); for (int y = 0; y < view(a).height(); ++y) { *loc = pixel_t(1, 0, 0); ++loc.x(); ++loc.y(); } auto v = color_converted_view<rgb8_pixel_t>(view(a)); write_view("a.bmp", v, bmp_tag()); } ////////////////////// // doesn't work ////////////////////// { bgr121_image_t b(10, 10); fill_pixels(view(b), pixel_t(0, 0, 1)); auto loc = view(b).xy_at(0, view(b).height() - 1); for (int y = 0; y < view(b).height(); ++y) { *loc = pixel_t(1, 0, 0); ++loc.x(); --loc.y(); } auto v = color_converted_view<rgb8_pixel_t>(view(b)); write_view("b.bmp", v, bmp_tag()); } } When running with VS 2017 all is working in Debug but with Release the second image is incorrect. The bug should be in the locator code. Anyone has an idea what is going wrong or what else I should try? Thanks, Christian

2 3

[boost.test] merging minor changes to master
by Raffi Enficiaud 31 Mar '18

31 Mar '18

Hi list, I merged a very minor commit that will make MinGW people happy. Rev of Boost.test on master is now 413b2f660118060db732dbf38a2bdffb527395e3 . Since Beta1: * doc updates * small bugfix PR 134 * include all lower case on windows for MinGW. PR 136 I would be grateful to the release managers if those changes go to the final 1.67 release. Thanks, Raffi

1 0