January 2020 - Boost - lists.boost.org

[Boost.Config] Should BOOST_FORCEINLINE depend on NDEBUG?
by dariomt＠gmail.com 29 Jan '20

29 Jan '20

Right now BOOST_FORCEINLINE does not depend on NDEBUG. This means that on Debug builds the code asks to be forced-inlined with optimizations disabled, which does not make much sense to me. This is reflected by a myriad warnings in msvc under level 4 warnings. In fact Multiprecision library works around this issue by defining their own macro BOOST_MP_FORCEINLINE which decays to just inline in Debug builds (boost/multiprecision/detail/number_base.hpp lines 25-29). I think this behavior should be adopted by BOOST_FORCEINLINE. I'm happy to do a PR if there is agreement from maintainers.

3 2

Re: [boost] Yet Another Container: DynamicArray
by Glen Fernandes 29 Jan '20

29 Jan '20

I wrote: > Yes quite. i.e. What you believe is not true about boost::allocate_unique. > > i.e. For auto p = boost::allocate_unique<T[]>(a, n); > assert(p.get().size() == n); > > This is because the unique_ptr stores a pointer adaptor around the > Allocator::pointer which also stores the size. > > So the pointer and size are stored together in the unique_ptr result. Note, that the size is only stored for allocate_unique<T[]>(a,n) but not for allocate_unique<T[N]>(a) since there the N is known at compile time. For auto p = allocate_unique<T[N]>(a); your p.get().size() will return N, but there's no need to store a size_t for N, so we don't. > To get the plain Allocator::pointer out of it you can use p.get().ptr() > > Or, for example, p.release().ptr() for someone who wants to manage > it themselves. Glen

1 0

New allocator that uses global memory
by Frederick Gotham 26 Jan '20

26 Jan '20

I propose a new allocator that uses global (i.e. static duration) memory. Here is my current experimental code: #include <cstddef> /* size_t */ #include <new> /* Only for bad_alloc */ static int this_translation_unit; /* This serves the same purpose as __FILE__ */ template<typename T, std::size_t t_capacity, std::size_t t_counter, int *t_file = &this_translation_unit> class StaticAllocator { protected: static T buf[t_capacity]; public: typedef T value_type; template<typename U> struct rebind { typedef StaticAllocator<U, t_capacity, t_counter, t_file> other; }; T *allocate(std::size_t const n) { if (n > t_capacity) throw std::bad_alloc(); return buf; } void deallocate(T *, std::size_t) { /* Do Nothing */ } }; template<typename T, std::size_t t_capacity, std::size_t t_counter, int *t_file> T StaticAllocator<T, t_capacity, t_counter, t_file>::buf[t_capacity]; using std::size_t; #include <vector> using std::vector; #include <iostream> using std::cout; using std::endl; auto main(void) -> int { vector< char, StaticAllocator< char, 4, __COUNTER__ > > v; v.push_back('a'); v.push_back('b'); v.push_back('c'); v.push_back('d'); for (auto const &elem : v) cout << elem << endl; vector< char, StaticAllocator< char, 4, __COUNTER__ > > v2; v2.push_back('x'); v2.push_back('y'); v2.push_back('z'); for (auto const &elem : v2) cout << elem << endl; // Now try the first vector again for (auto const &elem : v) cout << elem << endl; } The trick I'm using with "this_translation_unit" compiles just fine on GNU and Clang, however it fails on Microsoft because the variable has internal linkage. I'm sure this can be done another way using Boost preprocessor stuff. The preprocessor macro "__COUNTER__" is available on every C++ compiler I know of. Frederick P.S. This is my second idea to the Boost development list. My first one is "istream_iterator_limited".

2 1

boost/asio/awaitable.hpp including <experimental/coroutine>
by Paps Olo 25 Jan '20

25 Jan '20

Hi, i've been trying to get the coroutine chat example to work, but it seems that in 1.72 and 1.71 boost/asio/awaitable.hpp does not compile when BOOST_ASIO_HAS_CO_AWAIT has been defined. Is there something i overlooked (i've been away from Linux development for some years...) or is boost really broken with this define? Which version should i try, or is there any sense in trying to fix this myself? Best regards papsolo

4 4

Re: [boost] dynamic_*<T>
by Glen Fernandes 24 Jan '20

24 Jan '20

Alexander wrote: > So if T[] and T[N] are valid, then T[N][M] should also be. This increases the scope greatly but has an incredible difficult questions to answer: What would operator[] return? I don't have a good answer... Simple questions that already have an answer: The same way shared_ptr<T[N][M]> works today. :) Its operator[] returns element_type& So: auto p = allocate_shared<int[5][1024]>(alloc); p[3][675] = 8; auto q = allocate_unique<double[][2][3][4]>(alloc, size); q[index][1][1][1] = 0.95; Then: dynamic<int[5][1024], Allocator> a(alloc); a[3][675] = 8; i.e. Making it work for T[N] also automatically makes it work for T[N][M]. Glen

2 1

[license] Boost license question
by Edward Diener 24 Jan '20

24 Jan '20

If some compiler vendor takes a version of Boost and makes changes to that version of Boost to suit the particular compiler, while distributing that version with full source code as an optional part of some release of that compiler, does this mean that the code is now proprietary for that compiler vendor ? Or does the code still remain open source so that a programmer can use the changes made to update his own version of a Boost release or provide PRs to some later release of Boost based on the code in the compiler vendor's release ?

4 6

[Boost][GSoC]New algorithm for multiplication
by Debajyoti Dasgupta 23 Jan '20

23 Jan '20

Sir/Ma'am, I am highly interested in working on the multiprecision big number types . I wanted to propose that since In the proposed project there is an increase in precision from 1000 to 10000 decimal digit , we can implement toom-cook algorithm instead of the standard karatsuba multiplication.Toom cook algorithm has a complexity of O(n^1.465) which is quite faster than karatsuba multiplication O(n^1.58). Please share your views on this. Thank you

2 2

Willing to take part in GSOC 2020 with Boost.
by Ram Nad 23 Jan '20

23 Jan '20

Dear Mentors, I am a second year undergraduate student from India. I would like to work with Boost.Multiprecision library. I have completed Programming competency test and would like to familiarise myself with the code base. It would be really helpful if you could guide me on where to start. Thanks You Ram Nad

2 1

Anyone got AT commands work using asio serial port?
by JH 22 Jan '20

22 Jan '20

Hi, There are many boost serial port sample code out there, but none of them I can run AT command. I have a device connected to my ubuntu 18, I can run AT command on minicom to get correct response, but I could not get any AT response from boost serial port. Has anyone made it work? Thank you. Kind regards, - jh

5 11

New Boost Review Wizard: Mateusz Loskot
by Glen Fernandes 20 Jan '20

20 Jan '20

Our thanks to Mateusz for agreeing to help Boost with the formal review process. https://www.boost.org/community/reviews.html has been updated to say: > The role of Boost Review Wizard is currently played by: > * Mateusz Loskot (mateusz at loskot dot net) > * John Phillips (johnphillipsithaca at gmail dot com) > > Past Review Wizards (we thank you for your service): > * Ronald Garcia Glen

2 1