August 2017 - Boost-users - lists.preview.boost.org

Re: [Boost-users] Boost 1.65.0 Release Candidate 4
by Marshall Clow 17 Aug '17

17 Aug '17

On Wed, Aug 16, 2017 at 8:33 PM, Marshall Clow <mclow.lists(a)gmail.com> wrote: > The fourth (and hopefully final) release candidates for the 1.65.0 release > are now available at: > > https://dl.bintray.com/boostorg/release/1.65.0/source/ > [snippage] As always, the release managers would appreciate it if you download > the candidate of your choice and give building it a try. Please report > both success and failure, and anything else that is noteworthy. * Builds successfully with Apple Clang [Apple LLVM version 8.0.0 (clang-800.0.42.1)] with C++03/11/14/1z * Builds successfully with Clang Trunk [clang version 6.0.0 (trunk 310587)] with C++03/11/14 * Built successfully with Clang Trunk [clang version 6.0.0 (trunk 310587)] with C++1z except for Boost.Locale. -- Marshall

2 1

Boost 1.65.0 Release Candidate 4
by Marshall Clow 17 Aug '17

17 Aug '17

The fourth (and hopefully final) release candidates for the 1.65.0 release are now available at: https://dl.bintray.com/boostorg/release/1.65.0/source/ [ Note: The RC1s/RC2s/RC3s are in that directory, too - make sure you get the right files! ] Changes since RC3: * Fix a bug in Boost.Thread where timed waits were waiting for a very long time. Changes since RC2: * Fix a crashing bug in Boost.ASIO * Fix a compilation error when including both <boost/math_fwd.hpp> and <boost/math/common_factor_ct.hpp> * An update to Boost.Spirit to fix compilation errors on Mac OS with -std=c++17 Changes since RC1: * An update to Boost.Config so that the flags BOOST_NO_CXX98_RANDOM_SHUFFLE and BOOST_NO_CXX98_BINDERS are now set appropriately for libc++. (This fixes the compilation errors for Boost.Test) * An update to Boost.Fiber to fix compilation errors * An update to Boost.Wave to fix compilation errors on Mac OS with -std=c++17 The SHA256 checksums are as follows: 5d255d1ff8460d52dffcf9a9b55d908ae704027a66c590ec774096effaea26c1 boost_1_65_0_rc4.7z 911602d830ea5d16483d1e565b9859d120d03f2e5176017226ef391eb9d17fc8 boost_1_65_0_rc4.tar.bz2 369e06dc787d32b0fbaae020f33d814c9e7b77263acb859776968e623035f760 boost_1_65_0_rc4.tar.gz a9c6b8346b05ea43ac4278e80036a9e956ea99fc61f0b7f55469471af2df934c boost_1_65_0_rc4.zip As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. Thanks! -- The release managers

2 1

Where we are 1.65.0
by Marshall Clow 16 Aug '17

16 Aug '17

RC1 came out on Monday, and it had some problems. Specifically: * Some libraries didn't build with ToT clang/libc++ (Locale, Wave and Test) This is because in C++17, some deprecated library features have been removed. [bind1st, mem_fun, auto_ptr, random_shuffle, etc ] Boost.Config has support for these removals, but this was ... incomplete. [ This is on me; I did the partial job of adding these to Config's libc++ support a while back ] Normally, I would be tempted to say "Well, clang 5 is not shipping yet, and so we can put this off until 1.66.0". But both clang 5 and C++17 are imminent (they will both happen well before the 1.66 release, and so I think we should fix this now. Also, a future version of MSVC will have the same library issues. So I have committed a fix for Boost.Config that fixes the Boost.Test issues. Hartmut has fixed most of the Boost.Wave issues. Artymom has a fix for Boost.Locale. Windows/MSVC builds look good across the board. * Boost.Fiber fails to build on Mac OS with anything other than -std=c++03. There is a fix in develop for this; but there's still a bunch of tests failing. * There's a problem from the beta in Boost.Math. The regression regarding boost/math/common_factor_ct.hpp (vs > math_fwd.hpp) seems to be still there. (For details please see the > original mail and replies: > https://lists.boost.org/Archives/boost/2017/07/237452.php) > What I presume is the fix (https://github.com/boostorg/math/pull/75) was > merged into develop about 12 hours ago, some automatic CI tests still > running. * There's a problem with ScopeExit and Visual C++ 2017 With Visual C++ 2017 and /permissive-, ScopeExit is still broken, for which > there's an open PR: > https://github.com/boostorg/scope_exit/pull/3 > Looks good otherwise on MSVC 14.11, tested with /std=cpplatest and > _HAS_AUTO_PTR_ETC=1, targeting Windows 10 and XP, 32 bit and 64 bit, built > as static libraries. >

5 5

constructing rationals from mpfr_floats [multiprecision]
by Dani Brake 15 Aug '17

15 Aug '17

Hi, 1) I am getting signaling due to division by 0. The following lines are faulty: mpfr_float p("1.1"); mpq_rational q(p); signal: integer divide by zero; whereas this succeeds: mpq_rational q(1.1); any one seen this before? i'm on compiler Apple LLVM version 8.0.0 (clang-800.0.38) ,with homebrew-provided boost 1.64, mpfr 3.1.5, gmp 6.1.2. -- 2) I also expected to be able to construct rationals as mpq_rational q("1.1") but it fails to parse using the built-in constructor. i mean, i understand why (1.1 isn't a rational), but given that i can construct a rational from a double, i figured that i'd get something reasonable from "1.1", namely 11/10. -- thanks! dani brake university of wisconsin eau claire ps i am not disabling mixed double-multiprecision arithmetic in this scenario, in case you have been reading my other questions this last week

1 0

Re: [Boost-users] deleting operators to forbid mixed arithmetic? [multiprecision]
by Dani Brake 15 Aug '17

15 Aug '17

Paul and John, Putting a macro bracket around this might make it generally useful, perhaps > something like: > > #ifdef BOOST_MULTIPRECISION_NO_DOUBLE_OPS (BUILTIN_OPS would include > float and long double?) > > … > > template <class Backend> > struct is_compatible_arithmetic_type<double, number<Backend> > : public > mpl::false_ {}; > > say > #endif > > If you find this useful, you could create a GIT Pull Request, or just > email the file changed. You won’t be the only one J. Defining the macro > BOOST_MULTIPRECISION_NO_DOUBLE_OPS could be a way to detect dodgy > conversions that need to be changed to keep full precision. I implemented this in my codebase, and had few ill effects. Some unexpected conversions, but I did find this disabling very useful. Wrapping it in a macro bracket and making it officially supported would be fabulous. I think it should go in number_base.hpp at line 110 or so. If we wanted to make it robust and thorough, I think we need to consider turning off implicit constructors from disabled types, too. I believe that what we have so far only disables arithmetic. Dani

1 0

Re: [Boost-users] Problem building Boost 1.64.0 with Oracle Developer Studio 12.6
by Leo Carreon 15 Aug '17

15 Aug '17

It’s now nearly a week since I sent this email and I have not received a comment from anyone from Boost. Does this mean we have no experience building on Developer Studio 12.6 or Boost doesn’t care whether it works on Solaris or not. I suspect Problem #2 may be a Solaris issue but I’m not sure about Problem #1 because even the guys from Oracle have no comment about it. I would appreciate a comment from the maintainer of Boost.Optional. From: Leo Carreon Sent: Wednesday, August 9, 2017 10:52 AM To: Boost Users Mailing List Subject: Problem building Boost 1.64.0 with Oracle Developer Studio 12.6 Hi Boosters, I have been trying to build Boost 1.64.0 with Oracle Developer Studio 12.6 on Oracle Solaris 11.3 x86_64. There are two errors encountered during the build. They are: 1. Problem with Boost.Optional with this error messages: "libs/log/src/named_scope.cpp", line 118: Error: Use of overloaded binary operator '=' is ambiguous. "./boost/optional/optional.hpp", line 900: Note: Viable candidate 'boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>& boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>::operator=<boost::log::v2_mt_posix::attributes::named_scope_list&>(boost::log::v2_mt_posix::attributes::named_scope_list&)'. "./boost/optional/optional.hpp", line 962: Note: Viable candidate 'boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>& boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>::operator=<boost::log::v2_mt_posix::attributes::named_scope_list&>(boost::log::v2_mt_posix::attributes::named_scope_list&)'. 2. Problem with Boost.Python with this error message: "libs/python/src/numpy/dtype.cpp", line 61: Error: NPY_FLOAT16 is not defined. Note that these two errors are the cause of 9 Boost shared libraries to not be generated. I would appreciate it very much if someone could give me some pointers on how to resolve the above issues. Kind regards, Leo

3 2

[Log][Optional][Python] Problem building Boost 1.64.0 with Oracle Developer Studio 12.6
by Leo Carreon 15 Aug '17

15 Aug '17

Hi, This is a resend with the suggested change on the subject line. I have been trying to build Boost 1.64.0 with Oracle Developer Studio 12.6 on Oracle Solaris 11.3 x86_64 in C++14 mode. There are three errors encountered during the build. They are: 1. Problem with Boost.Log with this error message: >> Assertion: (../lnk/initializer.cc, line 1865) while processing ./boost/log/expressions/message.hpp at line 98. My comment: Acknowledged by Oracle as a compiler issue. 2. Problem with Boost.Optional with this error messages: "libs/log/src/named_scope.cpp", line 118: Error: Use of overloaded binary operator '=' is ambiguous. "./boost/optional/optional.hpp", line 900: Note: Viable candidate 'boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>& boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>::operator=<boost::log::v2_mt_posix::attributes::named_scope_list&>(boost::log::v2_mt_posix::attributes::named_scope_list&)'. "./boost/optional/optional.hpp", line 962: Note: Viable candidate 'boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>& boost::optional<boost::log::v2_mt_posix::attributes::named_scope_list>::operator=<boost::log::v2_mt_posix::attributes::named_scope_list&>(boost::log::v2_mt_posix::attributes::named_scope_list&)'. My comment: Ambiguous boost::optional<>::operator=(), possibly a Boost configuration issue or compiler SFINAE issue. I’m not that fluent with SFINAE to be able to figure this out. 3. Problem with Boost.Python with this error message: "libs/python/src/numpy/dtype.cpp", line 61: Error: NPY_FLOAT16 is not defined. My comment: Possibly a Solaris package installation issue due to no support contract. (numpy package fails to install) I would appreciate it very much if someone could give me some pointers on how to resolve the above issues. Kind regards, Leo

1 0

Boost 1.65.0 Release Candidate 3
by Marshall Clow 14 Aug '17

14 Aug '17

The third (and hopefully final) release candidates for the 1.65.0 release are now available at: https://dl.bintray.com/boostorg/release/1.65.0/source/ [ Note: The RC1s and RC2s are in that directory, too - make sure you get the right files! ] Changes since RC2: * Fix a crashing bug in Boost.ASIO * Fix a compilation error when including both <boost/math_fwd.hpp> <boost/math/common_factor_ct.hpp> * An update to Boost.Spirit to fix compilation errors on Mac OS with -std=c++17 Changes since RC1: * An update to Boost.Config so that the flags BOOST_NO_CXX98_RANDOM_SHUFFLE and BOOST_NO_CXX98_BINDERS are now set appropriately for libc++. (This fixes the compilation errors for Boost.Test) * An update to Boost.Fiber to fix compilation errors * An update to Boost.Wave to fix compilation errors on Mac OS with -std=c++17 The SHA256 checksums are as follows: 1d844a958656e76b140292ec1cbfe770d9916166c16e13b59d3dac1df3b7ef40 boost_1_65_0_rc3.7z 3fc2877cf5da4bbb0c91eab058d2e6764155e705e9c3e2dce2a0248936079675 boost_1_65_0_rc3.tar.bz2 e065b53d7f5326cc20d246b518facedb4407391aaf9fd4a1c90001e0c85accab boost_1_65_0_rc3.tar.gz 96b2ba69c6045060d1dfe012218457870200c2c22af60eb855bab5ffb75c13a2 boost_1_65_0_rc3.zip As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. Thanks! -- The release managers

1 1

Re: [Boost-users] deleting operators to forbid mixed arithmetic? [multiprecision]
by Dani Brake 14 Aug '17

14 Aug '17

> > So in principle you just define a full or partial specialization, for > example: > template <class Backend> > struct is_compatible_arithmetic_type<double, number<Backend> > : public > mpl::false_ {}; > and then cross your fingers and hope for the best ;) Thanks for the suggestion, i'll give it a try. \e crosses fingers ~dani

2 1

Re: [Boost-users] deleting operators to forbid mixed arithmetic? [multiprecision]
by Dani Brake 13 Aug '17

13 Aug '17

> > Been there, done that – fallen into the big pit awaiting the unwary mixing > double and multiprecision – again and again… L > I am being bit by this one hard. Converted a bunch of multipliers and tolerances from mpfr_floats to doubles in an effort to simplify some code, and now mixed precision is introducing error. > I think that you don’t really want to delete a family of operator*s – what > you DO want to get a warning when there is loss of significance, similar to > the warning when you fit a double into a float. > > It’s OK to multiply (or other ops) with a double as long as this doesn’t > lose any bits or digits of precision. > > So multiplying by 0.5 is OK, but not by multiplying 0.1 (the > multiprecision type needs FP_type(1)/10 to represent 0.1 as accurately as > possible for the FP type). > Multiplying by 0.1 is a common desire in my household (adaptive precision solving of polynomials). If only the compiler provided a trait bool is_exactly_representable(double) > so that a warning could be issued… > This is a bit tricky, because any double is by definition exactly representable by a double, right?. You would want is_exactly_representable to get the value before being transformed into a double? Another issue I foresee having to deal with is runtime parsing. Forbid the runtime user from feeding doubles problematic numbers. Thanks for the help, Dani

1 0