- Boost - lists.boost.org

[type_traits][core] modularisation and moving stuff about (again)
by John Maddock 23 Aug '18

23 Aug '18

So... I have a couple of PR's to move some of type_traits into config so it can be used by core. It's a modest proposal so I can go with that I guess... but this moving stuff about all over the place just to satisfy dependencies does wind me up a bit - I know that I sometime struggle to figure out where stuff is on Github - half of Integer is in Config for example. How on earth would an end user figure that out in order to file a PR or a bug report? I know... "calm down dear" and all that ;) So radical suggestion - does type_traits just belong in core and lets be done with it? What would happen to the file history if we moved it (can I still "git blame")? DARFC, John. --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

5 16

Re: [boost] [type_traits][core] modularisation and moving stuff about (again)
by Glen Fernandes 23 Aug '18

23 Aug '18

Andrey Semashev wrote: > I would rather encourage people use type traits, the official > interface. Boost.Config or Boost.TypeTraits could provide macros that > allow users to detect whether is_final is emulated or is the real > deal. One thing about this suggestion: If Boost.Config was to provide macros that tell users whether Boost.TypeTraits' boost::is_final<T>::value is emulated or the real deal, then the compiler detection logic currently in Boost.TypeTraits for detecting __is_final would already have to be duplicated in Boost.Config. That does not sound like a great idea. Right now TypeTraits provides both, which is better. i.e. It provides boost::is_final<T>::value and BOOST_IS_FINAL(T). The former uses the latter if it is defined, or is always false if not defined. If BOOST_IS_FINAL(T) and all its compiler detection logic lives in TypeTraits, then that all that compiler detection logic should not be duplicated in Config to provide a BOOST_HAS_IS_FINAL_INTRINSIC or similar feature detection macro. Glen

5 13

[test] I want my core dumped
by Alain Miniussi 22 Aug '18

22 Aug '18

Hi, I have assert in my library code, but when the assert fire in a test, the main function in minimal.hpp intercept the signal and explains that, yes, you asked for an abort, but no: [...] ../../../boost/test/minimal.hpp(136): exception "signal: SIGABRT (application abort requested)" caught in function: 'int main(int, char **)' Is there a way to avoid that behavior ? I need that core file. Thanks Alain

3 2

[multiprecision] Design question on variable precision types.
by John Maddock 22 Aug '18

22 Aug '18

Folks, I have a tricky design question that Boost.Multiprecision has been dodging relating to variable precision types - these have just expanded somewhat now that we support complex numbers via mpc as well as mpfr's reals etc. The basic use case is this: We have a set of numbers at very high precision, some calculations are performed on those directly, others we want to temporarily drop precision, get an approximate answer and perhaps later refine it. Generally speaking, performance trumps everything, on a really long running calculation 10-20% speedup can actually make a big difference (in $$'s in some cases if the CPU time is from the cloud). Up till now Boost.Multiprecision has been dodging this issue - it's actually pretty hard to program in a sane and consistent way, so the message has been "don't mix precisions" when using a type like boost::multiprecision::mpfr_float - just set what default you want and get on with it. But we really should do better than that. So... with the help of Danielle Brake I've been trying to nail this down in develop, what we have at present is: * Copy construction or move assignment copies the precision of the source. * Regular assignment retains the precision of the target variable. This is efficient in that assignment doesn't reallocate to a new precision, and provides an easy way to up/down sample precision when required. But it breaks equality after assignment. It does match what MPC/MPFR do in mpfr_set etc though. The fact that move assignment behaves differently from regular assignment bother me too - it's too easy to write code that triggers move semantics without noticing. The trouble is, being able to assign while retaining precision of source is sometimes useful too, so having move-assign behave differently provides a useful back door, but maybe one that's just too cute for it's own good? And then.... we get on to arithmetic. * Currently temporaries are created at current default precision. So it's easy to shoot yourself in the foot by not noticing that temporaries are being created in an expression and not having the default precision set correctly. I actually think this one might be solvable, but it's a huge job involving banning default constructors in all the backends to our number types, and then catching all the cases where they might have been used. * What should happen in "a = expression", where the variable used in "expression" are at different precision to a? The uncontroversial answer might be to adjust the result to the precision of a. But hold on... we have expression templates, and these can use variable a as working space while evaluating the expression... so the change in precision might occur *during* the computation and not after it. * Should we try to ban mixed precision arithmetic altogether? It's a tempting idea, but I can't see how to enforce it in code without also banning mixed precision assignment. Thanks for your consideration, John. --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus

2 1

[MSM] Still maintained?
by Deniz Bahadir 22 Aug '18

22 Aug '18

Hi Boost-list, because I realized that my simple PR for Boost.MSM [1] (which makes it compilable with C++17) and an even older one did not receive any attention, I wonder if Boost.MSM is still actively maintained. The latest changes seem to be from around December 2017. Does anyone know, if Christophe is still active? Best regards, Deniz [1] https://github.com/boostorg/msm/pull/15 -- BENOCS GmbH Dipl.-Inform. Deniz Bahadir Reuchlinstr. 10 D 10553 Berlin Germany Phone: +49 - 30 / 577 0004-22 Email: deniz.bahadir(a)benocs.com www.benocs.com Advisory Board (Chairman): Oliver Fietz Board of Management: Stephan Schröder, Dr.-Ing. Oliver Holschke, Dr.-Ing. Ingmar Poese Commercial Register: Amtsgericht Bonn HRB 19378

1 0

Increasing BOOST_PP_VARIADICS_SIZE
by Damien Buhl 20 Aug '18

20 Aug '18

Dear Edward, In comparison to BOOST_PP_SEQ, BOOST_PP_VARIADICS is limited to 64 elements while BOOST_PP_SEQ is limited 255. Is there a reason for the choice of 64 or could I patch BOOST_PP_VARIADICS to go as BOOST_PP_SEQ to a limit of 255 ? Or asked differently do you think there are in newer compilers/standards a way to have an unbounded BOOST_PP_VARIADICS implementation nowadays ? Warm Regards, -- Damien Buhl Software Developer +33 6 77 43 10 05

2 2

ABI issues with -std
by Peter Dimov 20 Aug '18

20 Aug '18

At some point, we'll need to acknowledge the problem which results from building Boost with one -std and user code with another, and deal with it somehow, if just by documenting. Most of our libraries don't suffer from it at the moment but Boost.System does, because of my changes; its ABI changes between 03/11/14. So I'm going to take the initial hit. :-) POSIX platforms have long been able to ignore ABI issues so we don't mangle the library names there at all, but now whether one builds Boost for C++03, C++11 or C++14 matters, and there's nothing to prevent linking C++xx to C++yy. Even with --layout=versioned, the cxxstd level is not encoded. The fact that the default C++ standard differs between g++ and clang++ doesn't help; when following the instructions of doing "b2 install", with g++6 and above you get C++14, but with g++5 or clang++ you get C++03. The latter is very rarely what one wants today.

7 26

Formal review request: Synapse
by Emil Dotchevski 16 Aug '18

16 Aug '18

Hello, I'm requesting a formal review for Synapse, a non-intrusive signal programming library. Synapse is a non-intrusive C++ signal programming library. It is able to register connections and emit signals from any C or C++ object of any type whatsoever. If two contexts share access to a C or C++ object, they can use that object as a randezvouz point of sorts, and communicate with each other through Synapse signals. A system of meta signals provides interoperability with other signal programming libraries and callback APIs. In particular Synapse can be used to safely register C++ function objects, including lambda functions, with any C-style callback API that is designed to work only with C function pointers. In a multi-thread environment signals can be emitted asynchronously and scheduled for synchronous execution when requested in other threads. There have been numerous improvements since the first review, including the the ability to create persistent connections which don't require the user to explicitly keep them alive. See https://zajo.github.io/boost-synapse/. Thanks, Emil

2 1

Strange libstdc++ versions in travis-ci testers
by Andrzej Krzemienski 15 Aug '18

15 Aug '18

Hi Everyone, Maybe someone could help me understand the configuration on the travis-ci testers. Whenever I push a new commit I get a number of failures from travis-ci about missing std::is_trivially_default_constructible from the standard library; like here: https://travis-ci.org/boostorg/optional/builds/397519342?utm_source=email&u… Some tests with Boost.Predef indicate that they are clang compiler 3.X, and clang 5 with libstdc++ of strange version nubers: macro BOOST_LIB_STD_GNU gives values 450400026 and 460700026, which would correspond to releases 2015-04-26 and 2016-07-26, but when I look at the release notes for GCC there are no releases mentioned of these dates: https://gcc.gnu.org/news.html Maybe someone knows what versions of libstdc++ these are? Or maybe I am missing something obvious? Regards, &rzej;

5 12

numeric directory/namespace
by Robert Ramey 14 Aug '18

14 Aug '18

I'm preparing to make adjustments to safe_numerics directory and namespace in preparation to incorporating safe_numerics into the official boost distributions. I have a couple of decisions to make that someone/everyone will want to weigh in on. There is already a directory boost_root/libs/numeric which seems to be an "umbrella" for libraries dealing with numerics. It includes conversion, interval, odeint and ublas. There is an index.html in there also. But instead of referring to the above libraries, it points to the ublas documentation. So it's sort of tempting the place safe numerics in libs as a sub directory. But I really have strong reservations about the whole concept of have this "umbrella" directory in the first place. Personally I'd just prefer to use boost_root/libs/safe_numerics. I don't think the extra structure adds anything and makes it harder to implement tools which scan the directories. It would also mean that I should ? use #include <boost/numeric/safe_numeric/ ....> and I'm not sure want to do that. Also I expect I'll want to sync the namespace boost::numeric::safe_numeric:: ... with the file system. It just seems more complex and flexible than it has to be. I see the future of boost as diminishing coupling between libraries and I see the current setup as not helping with this. Please don't wait to answer until I've made the changes. Robert Ramey

3 2