January 2019 - Boost - lists.preview.boost.org

C++Now 2019 Registration is Open
by Jon Kalb 24 Jan '19

24 Jan '19

Boosters, Registration for C++Nows 2019 (aka BoostCon) is now open: http://cppnow.org/announcements/2019/01/2019-registration-is-open/ Please help us to spread the word about this announcement. Thanks. Note that the Aspen Meadows Resort online registration link is also up on the lodging page. Because there will be another group in the hotel, we’ve been allocated fewer rooms this year. Please either reserve a room very soon or plan to stay at the Lamplighter or another hotel in town. See you in Aspen in May. Jon

1 0

[gsoc19][GIL] Project proposal
by Miral Shah 23 Jan '19

23 Jan '19

Hello everyone, I am Miral Shah a computer engineering student from India. I have been talking with Boost.GIL community on GIL's Gitter channel since last few weeks to come up with the new idea for GSoC19. After many suggestions and a long discussion on various ideas with the contributors, we have come up with the idea to expand the GIL by introducing new image processing algorithms into it. I would also like to thank Stefan Seefeld and Mateusz Loskot for their guidance and support. Proposal idea can be found here: https://github.com/boostorg/gil/wiki/GSoC-2019 Thank you, Miral Shah

1 0

Boost log issues and workarounds
by hh h 21 Jan '19

21 Jan '19

Hi, I like the boost log, but there are many small issues, appreciate your kind help: I followed the following link to use a good macro LOG: https://stackoverflow.com/questions/35895199/boost-log-and-severity-local-a… But calling LOG(error) caused much trouble for the name space collision, alternative is to use full namespace boost::log::trivial::error, but that is too cumbersome,I tried to use an uppercase ERROR to replace it in an enum ERROR = boost::log::trivial::error, obviously that could not be complied, although it is defined as enum severity_level in trivial.hpp, it does not look like the boost::log::trivial::error is an enum value. Any workarounds? Another issue from that link is it has to call LOGGABLE prior to call LOG(error) in each function, that makes more clumsy, software developers in my team are not happy about it. Any workaround to have a fixed argument in BOOST_LOG_STREAM_WITH_PARAMS, without need to claim boost::log::sources::severity_logger<boost::log::trivial::severity_level> slg; in each function? You guys are genius that must be a way to do it, appreciate your advice. Thank you

4 22

[lockfree] wait-free multi-queue?
by Shawn Yarbrough 18 Jan '19

18 Jan '19

Is it practical to have a wait-free, multi-producer/multi-consumer queue? I ask because Boost's Lockfree multi-queue is only advertised as being "lock-free". (Boost does offer a wait-free single-queue.) https://www.boost.org/doc/libs/1_69_0/doc/html/lockfree.html Shawn Y.

2 1

Re: [boost] [parameter] Helping user libraries adjust to recent changes in GitHub develop branch
by Cromwell Enage 17 Jan '19

17 Jan '19

Happy New Year, everyone! > I've been contributing upgrades to Boost.Parameter functionality > and documentation in the GitHub develop branch for the past > couple of months. These upgrades include the addition of code > generation macros for Boost.Parameter-enabled function call > operators and support for perfect forwarding, and more upgrades > are yet to come. It's been an enjoyable experience so far, > but understandably there are some growing pains. I've taken > pains to ensure that as long as user libraries access only the > public interface of Boost.Parameter, they won't be affected by > the changes. However, there are several notable libraries that > currently use non-public portions of Boost.Parameter, so > whenever possible I've submitted PRs to those libraries that > either change the usage to access only public portions > (preferred) or make quick fixes (temporary). Update: As of this writing, I have active PRs for the following libraries: Boost.Accumulators Boost.Convert Boost.Graph Boost.Log AFAIK, these are the libraries most directly affected by the changes. To reiterate: I'm willing to help the library maintainers resolve these issues as quickly as possible. I should be able to answer any questions you have about the changes to Boost.Parameter. Cromwell D. Enage

3 2

[histogram] discussion of accessor interface design
by Hans Dembinski 15 Jan '19

15 Jan '19

Dear all, work on boost.histogram is progressing fast. I am still implementing feedback from the review, simplifying the interface, adding requested features, such as axes that can grow. STL compatibility was further improved, you can now also write to histograms via iterators. The normal iterators make sense when you iterate over a 1D histogram. When you iterate over a multi-dimensional histogram, you also want to know the current multi-dimensional index. After considering many options, and I really thought about this a lot, I went for the following design, which is a bit unusual. Therefore I would appreciate feedback. I think it is great, once you overcome an initial feeling of awkwardness. The accessor class is "polymorphic", it behaves like a pointer to the histogram value, and like an array for the multi-dimensional index. In code, this is how you use it: auto h = make_histogram(…) // make 2D histogram for (auto && x : indexed(h)) { // indexed produces a range of accessors // x is a special accessor type, combining two non-overlapping concepts // - it acts like a pointer to the histogram value // - it acts like an array to the current index std::cout << "current value " << *x << std::endl; // "dereference" to get value std::cout << "current index " << x[0] << " " << x[1] << std::endl; // use subscript operator to get index } This syntax is beautifully terse, e.g. see https://github.com/HDembinski/histogram/blob/develop/examples/guide_access_… for a full example, especially line 66. Pros: - really terse - you can access methods on the pointee with x->method() as well! (useful when histogram counters are not PODs) - you can iterate over x to get the indices, for (auto i : x) { … } works - since x acts like an array to the indices, you can pass it to functions which accept ranges or iterators (it has .begin() and .end()) - x can be (and is) enhanced with other useful methods * x.bin(N) returns the current bin interval for the N-th axis, allowing you to access the central value, width, edges * x.density() returns the current density (bin value divided by product of current bin widths) Cons: - *x and x[0] do completely different things: *x gives you the bin value, x[0] gives you the first index The last point is, of course, where people have a problem. But if you take C++ concepts seriously then the accessor is a perfect model of a pointer and a perfect model of an array. These two roles are non-overlapping and they have non-overlapping sets of interfaces, which I exploit here. If you have the expectation that *x and x[0] should do the same thing, it is so because of C. C has no extensive type system like C++ and does not distinguish between arrays and pointers, although these are very different concepts. A pointer points to a value, and an array is a collection of values. "dereferencing" a collection of values makes no sense, we are just used to it because of our C heritage. C++ has a better type system, and better classes for pointers and arrays than raw pointers. The stdlib authors recognize that *x has no meaning when x is a sequence of values. *x fails when x is a std::vector, std::deque, std::list, or any kind of collection in the stdlib. Even for std::unique_ptr, they made sure that the interfaces for the pointer-to-object and pointer-to-array specializations behave differently: ``` #include <memory> int main() { std::unique_ptr<int> p(new int); // pointer version // p[0]; // fails *p; // OK std::unique_ptr<int[]> a(new int[3]); // array version // *a; // fails a[0]; // OK } ``` Once you accept that the two concepts of pointers and arrays have non-overlapping interfaces in C++, it becomes possible to make a combined object which has both interfaces, and uses these two sets to return different information. If you have read this far, I hope your initial reaction of "woah, this looks really inconsistent and confusing!" turned to "hmm, maybe it is not inconsistent after all". I am looking forward to hear your thoughts. Best regards, Hans PS: The alternative would be to return a std::pair<index_type, value_reference_type>, but this has disadvantages. Unpacking the pair is going be nice in C++17 with structured bindings, but not so much in C++14. Also, it prevents me from adding convenience methods, like the above-mentioned bin(N) and density() methods.

1 0

boost log to concatenate just one log file
by hh h 15 Jan '19

15 Jan '19

Hi, I added boost::log::keywords::open_mode = std::ios_base::app to the log configuration, but it did not stop log file proliferation, every time I started the program, it created a new log file mylog.log0000x, how can I stop creating new logs? Thank you.

2 2

Is it possiable to use a simple LOG(info) in boost log?
by hh h 15 Jan '19

15 Jan '19

Hi, After trying many days on boost log, still have many issues. Google log has a nice LOG(INFO) << "debug";, looks like that kind thing cannot do in boost log, the close online code in https://stackoverflow.com/questions/35895199/boost-log-and-severity-local-a… has to call LOGGABLE first then to use LOG(info) << "my log", it is really cumbersome, is it possible to call LOG(info) like glog? Thank you.

1 1

Boost.Build requirements to exclude GCC 5
by Niall Douglas 14 Jan '19

14 Jan '19

Outcome is now into Boost, and is being tested nightly: Docs in BoostDoc theme: https://www.boost.org/doc/libs/develop/libs/outcome/doc/html/index.html Regression matrix: https://www.boost.org/development/tests/develop/developer/outcome.html One problem is how to exclude from testing GCC 5 which Outcome cannot support due to GCC 5's incorrect parsing of variable templates in nested template classes [1], so I need some sort of Boost.Build project requirement which explicitly excludes GCC 5. Unfortunately, GCC 5 appears to Boost.Config as fully implementing C++ 14, so this is not straightforward. What might work is to specify a requirement for one of the library items at https://github.com/boostorg/config/blob/develop/checks/Jamfile.v2 which we know for a fact libstdc++ 5 does not implement, but libstdc++ 6 does. However, clang with libstdc++ 5 would then be excluded incorrectly, as clang does not have this parsing bug. I guess ideally speaking Boost.Config gains a new test testing for this specific compiler parsing bug. But it seems a bit much to ask for this just for Outcome. Guidance is sought! Niall [1]: https://www.boost.org/development/tests/develop/developer/output/teeks99-02…

4 5

C++Now 2019 Submission Deadline
by Jon Kalb 14 Jan '19

14 Jan '19

The deadline for submitting to C++Now 2019 is just a few days away. If you’ve been thinking of submitting, please do not delay: http://cppnow.org/announcements/2018/12/2019-call-for-submissions/ <http://cppnow.org/announcements/2018/12/2019-call-for-submissions/> Thanks. Jon

1 0

High number of integer_sequence/index_sequence implementations
by Nikita Kniazev 14 Jan '19

14 Jan '19

Bunch of Boost libraries have their own integer_sequence/index_sequence implementation. Wouldn’t it be better if there was boost::integer_sequence/boost::index_sequence? boost\beast\core\detail\integer_sequence.hpp boost\callable_traits\detail\polyfills\make_index_sequence.hpp boost\context\detail\index_sequence.hpp boost\fusion\support\detail\index_sequence.hpp boost\mp11\integer_sequence.hpp boost\phoenix\core\detail\index_sequence.hpp (alias of fusion one) boost\poly_collection\detail\functional.hpp (uses mp11 one) boost\test\data\index_sequence.hpp boost\variant\detail\multivisitors_cpp11_based.hpp

4 4

(no subject)
by Siddhartha Sen 14 Jan '19

14 Jan '19

Hey, This is Siddhartha. I have been interested in contributing to boost and hopefully work with boost in GSOC 2019. I apologise for the delay as my exams were going on and I hope it is not too late to start. I have looked in to the ideas list as well as last year's projects. Currently, I am proficient in c and c++,especially algorithms,data structures and maths. Any help regarding how to get started would be appreciated. I myself have some ideas about implementation of a few libraries which I would like to discuss as well. Thank You

1 0

Regression matrix stopped updating
by Nikita Kniazev 13 Jan '19

13 Jan '19

For past month or two regression matrix update frequency was fluctuating in a big range, sometimes not updating for two-three days. As for now master is not updated for 7 days and develop for 5 days. https://www.boost.org/development/tests/master/developer/summary.html Summary: Master Report Time: Tue, 11 Dec 2018 01:44:54 +0000 https://www.boost.org/development/tests/develop/developer/summary.html Summary: Develop Report Time: Thu, 13 Dec 2018 06:05:04 +0000 Is it down for maintenance or what is going on?

6 8

Safe numerics at Meeting C++
by Robert Ramey 12 Jan '19

12 Jan '19

News Flash for both of you interested in the Boost Safe Numerics Library. Checkout https://www.youtube.com/watch?time_continue=942&v=tcyb1lpEHm0 at 1 hour in. The whole rest of the talk is about a safe integer implementation written D. He uses it to illustrate his concept of "Introspective Programming". ALL the features and techniques he mentions (and more) are implemented in the safe numerics library. The Safe Numerics library is the Next Big thing! Robert Ramey

1 0

question/guidence regarding merge to master
by Robert Ramey 12 Jan '19

12 Jan '19

I'm looking to merge my develop branch into the master so that they will be in sync. I'm thinking that this the changed in develop should be squashed so that the master branch doesn't included all the the intermediate changes made during the develop phase. Is there any boost rule/practice/guidence regarding this?. Ideally, I'd like to see the master have only one set of consolidated changes for each release. But I'm not the person who decides these things or does the work. Any insight by other parties would be appreciated. Robert Ramey

10 29

Re: [boost] A class for embedded friendly callbacks.
by Bjorn Reese 12 Jan '19

12 Jan '19

On 1/4/19 1:30 PM, Mikael Rosbacke via Boost wrote: > I'm Mikael Rosbacke and new to this list. The last couple of months I've > been writing on a class for storing callbacks to different forms of > callables (member functions, functors, free functions). The main focus A while back, a somewhat similar library underwent a formal review. You may want to read the review report to see if any of the comments applies to your library: https://lists.boost.org/boost-announce/2016/12/0484.php

1 0

Re: [boost] Feedback desired for reflective enum library
by Nir Friedman 12 Jan '19

12 Jan '19

I had to fill in the subject for this manually, because of settings for the mailing list. I really hope it ends up on the right thread. Apologies if it doesn't. Hi Geert, There is an ADAPT macro, which is similar to your concept of CONVERT_TO_WISE_ENUM. However, it does require listing the enumerators; after all there's no way to find them out after the fact or else this library wouldn't exist in the first place. That said, I'm not quite sure what you mean by "changing your whole codebase". To take advantage of wise enums, all you'd need to do is change the original enum declaration to the wise enum macro. Everything else in your codebase will continue to work exactly as before. No typedef shenangians are necessary. Changing the original enum declaration is exactly the same amount of work as writing an adapt macro, but you don't have to maintain the list of enumerators twice. So I can't imagine a reason to write things the way you have (unless you want a conditional build dependency on WISE_ENUM but that seems unnecessary). The ADAPT macro is designed to be use wise_enum functionality generically with enumerators whose source you don't control, and therefore cannot change the declaration. Hope that clarifies things. Cheers, Nir

1 0

boost log to display source code name and lines
by hh h 11 Jan '19

11 Jan '19

Hi, I am looking for good log library to replace google log, my plan is to use boost log, but I am not sure if it is mature or it has been used seriously. It lacks some simple basic features such as display source code name and lines in log file, I've searched on line discussion, there are many suggestion to fix it in piece of code, but I could not get that work, difficult to use it may be overstated. Are there any good examples to display source program name and lines similar to asio socket examples on line? What is your opinion about boost log? Should I still pursue it? Thank you. - jhh

3 10

GSoC19
by Pranam Lashkari 10 Jan '19

10 Jan '19

Hello, Organization application date is going to start soon. I was trying to find our idea page for this year but couldn't find it. can anyone help!? Thank you, Pranam Lashkari

1 0

Requesting formal review for (Boost) LEAF
by Emil Dotchevski 10 Jan '19

10 Jan '19

LEAF is a small C++11 error handling library (the acronym stands for Low-latency Error Augmentation Framework). Features: - Header-only, NO dependencies. - No dynamic memory allocations. - Associate objects of arbitrary types with any failure — when it is initially reported or at a later time. - Compatible with std::error_code, errno and any other error handling API. - Use with or without exception handling. - Support for multi-thread programming. Official documentation: https://zajo.github.io/leaf. LEAF is designed for use in low-latency environments without exception handling, but it also works with exception handling; in that case it is a better, more optimal replacement for Boost Exception. This link can be helpful for refactoring existing Boost Exception code: https://zajo.github.io/leaf/#boost_exception.

5 9

async_read SEGFAULT
by hh h 09 Jan '19

09 Jan '19

I use one io_service in one process to run two servers A and B each has its own port, own context and SSL configuration to accept inbound connections. Although both A and B have boost::asio::deadline_timer, but I think there is only one thread is running: (gdb) info threads Id Target Id Frame * 1 Thread 0x7ffff7fe0bc0 (LWP 31204) "ssl_connection" 0x00007ffff4a690a3 in __epoll_wait_nocancel () at ../sysdeps/unix/syscall-template.S:84 The problem I got: - If I started only one server A, it worked just fine for both async_read and async_write. - If I started both servers A and B, even B was in idle, server A could sometime be crashed after receiving and responding just one message, then crashed at next async_read in the first second, it was more stable if it did not call async_write. SEGFAULT: Fault address=0x55b3700d00cf, #4 0x000055555557f98a in boost::asio::detail::task_io_service_operation::complete (this=0x55555584adc0, owner=..., ec=..., bytes_transferred=0) at /usr/include/boost/asio/detail/task_io_service_operation.hpp:38 #5 0x0000555555581f82 in boost::asio::detail::task_io_service::do_run_one ( this=0x555555814710, lock=..., this_thread=..., ec=...) at /usr/include/boost/asio/detail/impl/task_io_service.ipp:372 #6 0x0000555555581ab4 in boost::asio::detail::task_io_service::run ( this=0x555555814710, ec=...) at /usr/include/boost/asio/detail/impl/task_io_service.ipp:149 #7 0x000055555558221f in boost::asio::io_service::run (this=0x7fffffffe150) at /usr/include/boost/asio/impl/io_service.ipp:59 It seems that async_write and async_read messed up, but that problem was intermittent, especially when the server A passed first second, it could be running for hours for doing consequent many async_read and async_write simultaneously. It was only one thread, what could cause that problem? Thank you.

2 11

b2 --exec-prefix=...
by Peter Dimov 09 Jan '19

09 Jan '19

Our "b2 --help" documentation says --exec-prefix=<EPREFIX> Install architecture dependent files here. Default: <PREFIX> --libdir=<LIBDIR> Install library files here. Default: <EPREFIX>/lib and bootstrap.sh also supports --exec-prefix and when given, records it in project-config.jam. But it doesn't work and has never worked. Only --prefix is honored, --exec-prefix is simply ignored. I think that we should bring the documentation and the actual behavior in sync, either by making --exec-prefix work as documented, or removing it from the documentation. Which one should we do? Comments?

5 4

Boost wiki - revised Boost guidance for authors on Warnings - RFC
by Paul A. Bristow 08 Jan '19

08 Jan '19

As promised at the end of a long (hijacked!) thread on 'Warnings' http://boost.2283326.n4.nabble.com/Current-Guidance-on-Compiler-Warnings-td… starting from a query by Brian Kuhl of Wind River on Boost's policy on compiler warnings, I have spent some time revising the previous guidance document on Trac that I edited many years ago. This was a long and tedious task as Markdown, as most conveniently used on Github, is rather different from the mark-up language on Trac, leaving me slightly in fumble mode :-( I have made some updates en passant, but there is still much to do updating for the much improved granularity of warning suppression and perhaps more about other code improvement tools like Lint, UBSan. The overall structure leaves a lot to be desired, but I don't want to put any more work into this without some encouragement that it will prove useful. I have posted the result of my efforts on https://github.com/boostorg/boost/wiki/Guidelines:-Warnings and I am as promised I am starting this new thread on warning for your comments, corrections, condemnations etc. HTH Paul --- Paul A. Bristow Prizet Farmhouse Kendal UK LA8 8AB +44 (0) 1539 561830

3 2

[cmake] Pull request announcement - update
by mike.dev＠gmx.de 07 Jan '19

07 Jan '19

I think it is time for an update on this issue. #################################################### # TLDR: There are now ~25 boost libraries that can be directly used in a cmake project via the add_subdirectory & target_link_library - workflow. And I want to encourage library maintainers to comment/merge outstanding PRs or make similar additions to their libraries themselves. ################################################### # A summary what that thread was about: In September last year I was informing the mailing list that I was about to make a bunch of PRs adding a set of minimal CMake files to various boost libraries. Their sole purpose was to enable the following workflow from within a cmake project: add_subdirectory( <path-to-boost-libfoo-directory> ) target_link_library( my_app Boost::libfoo ) Without needing b2 or the boost super project for that matter (although support in the super project would be very welcomed) For that, the cmake file in libfoo has to - provide the target Boost::libfoo - attach the dependency information for Boost::libfoo - attach the proper include directory for Boost::libfoo - if necessary, build libfoo with whatever toolchain and compile flags the parent cmake project defines - if necessary, attach any necessary compile definitions to Boost::libfoo Although things like running the unit-tests or provide and install target can be added by any library maintainer (and some have done just that) it is not part of my PRs. ################################################### # What is the current state. For a quick overview, I uploaded a picture of the boost dependency graph (code based on boostdep) that shows which libraries Have a cmake file in their root and which don't [1]. So far, about 25 libraries have gained cmake files as part of this effort (not all from me). And of course there are a couple of libraries that Already used cmake before I started my efforts, but AFAIK, most of them can not be used in the above mentioned work-flow. Of course, there remains a lot to do, but at least progress is happening and we now have fitted almost all "foundational" libraries , so from here on out, the pace should increase. Currently, I have 7 outstanding PRs and I would ask the Respective maintainers to comment on them as soon as they can find some spare time, as further PRs are blocked on them. Even if the PRs get completely rejected, I at least know for which part of boost I should no longer pursue this quest and/or find a workaround. Finally a big thanks to all maintainers that added such cmake files on their own. Best Mike [1] https://gist.github.com/Mike-Devel/46d04409d6ae6053861706e3ad91e4dc Explanation for the linked picture: It shows the boost dependency graph (generating code based on boostdep). - Green libs already have a cmake file in their root directory - Yellow libraries don't have such a cmake file, but all their dependencies have one, so one could be added and most likely there is already a PR for that - Red libraries have at least one direct or indirect Dependency that lacks a cmake file, so I consider Them "blocked" for now.

4 3

A class for embedded friendly callbacks.
by Mikael Rosbacke 07 Jan '19

07 Jan '19

Greetings. I'm Mikael Rosbacke and new to this list. The last couple of months I've been writing on a class for storing callbacks to different forms of callables (member functions, functors, free functions). The main focus has been to make it suitable for embedded work with small footprint and dependable exception and heap behavior. The code can be found on github: https://github.com/rosbacke/delegate It has reached a stage where I'm getting sort of 'done'. I feel it could be useful to a wider audience so I would like to get some feedback about if this is suitable for that and if it is worth starting a new branch for possibly adapting it to boost for inclusion in some future version of boost. Any feedback is appreciated. Cheers, --- Mikael Rosbacke

2 2