What's Changed

• Fix compilation when building with msvc's new preprocessor by @pps83 in #667
• Further protect attribute to avoid warnings by @mborland in #668
• Attempt repair issue670 and add tests by @ckormanyos in #671
• Update and correct docs and badges by @ckormanyos in #672
• Update github actions by @mborland in #675
• Handle implicit fallthrough warn on no-except by @ckormanyos in #674
• Remove namespace alias from not our namespace by @mborland in #680
• Repair warning from issue 679 by @ckormanyos in #681
• A few header-only warns and warns in CI by @ckormanyos in #683
• Handle issue652 with upper-bound check by @ckormanyos in #684
• Uniform logic has or not-has GCD/LCM by @ckormanyos in #676
• Remove unnecessary moves by @pps83 in #692
• Fix warning from scoped_default_precision by @pps83 in #691
• Minor fix for mpfr error handling in log1p. by @jzmaddock in #693
• Update corrected MPIR web link by @ckormanyos in #694
• Add support for modular build structure. by @grafikrobot in #632
• Trivial doc spell correction of word fulfills by @ckormanyos in #698
• double fp backend from gsoc21 by @ckormanyos in #701
• Handle docs issue 685 by @ckormanyos in #704
• Suppress spurious array bound warning from gcc. by @jzmaddock in #702
• Enhance quality via tests and syntax by @ckormanyos in #705
• Optional double-fp finish int convert by @ckormanyos in #706
• Change CircleCI runner image to a "known good" by @jzmaddock in #707
• More cover dec_float and improve comments by @ckormanyos in #708
• Simplify and update Markdown by @ckormanyos in #709

New Contributors

• @pps83 made their first contribution in #667
• @grafikrobot made their first contribution in #632

Full Changelog: Boost_1_88_0...Boost_1_89_0

What's Changed

• Make float128 a trivially copyable type by @nmnobre in #636
• Make __float128 a floating point type. by @jzmaddock in #637
• Merge to Master for 1.87 by @mborland in #642
• Disable check for __float128 on PGI compiler by @mborland in #641
• Add bug fix for PGI compiler to 1.87 by @mborland in #643
• GHA: Use node 20 actions by @Flamefire in #645
• Try and optimize small integer powers. by @jzmaddock in #649
• Fix some gcc-13 warnings. by @jzmaddock in #650
• Suppress warnings caused by use of -Wswitch_default on GCC by @jzmaddock in #651
• fix the mpfr stack str bug of the deadloop in the while by @FAKERINHEART in #657
• Add #657 to master for 1.88 by @mborland in #659
• Suppress warnings caused by -Wimplicit-fallthrough by @koct9i in #660
• Don't use deprecated form of user-defined literal operator by @Thomas-Barbier-1A in #663
• Don't use deprecated form of user-defined literal operator around BOOST_JOIN by @Romain-Geissler-1A in #665
• Merge fix for user-defined literal operator to 1.88 by @mborland in #664

New Contributors

• @nmnobre made their first contribution in #636
• @Flamefire made their first contribution in #645
• @FAKERINHEART made their first contribution in #657
• @koct9i made their first contribution in #660
• @Thomas-Barbier-1A made their first contribution in #663

Full Changelog: Boost_1_86_0...Boost_1_88_0

What's Changed

• Type which are convertible to a number should not participate in arit… by @jzmaddock in #609
• Improve codecov reliability by @mborland in #616
• make sure eval_convert_to() do not terminate with super large number by @tchaikov in #618
• Remove overloads no longer required by Math. by @jzmaddock in #623
• Correct divide-by-zero in cpp_int modulus operations. by @jzmaddock in #625
• Ensure checked integers are in a sane state after a throw. by @jzmaddock in #627
• Update docs by @ckormanyos in #629
• Improve quadmath detection by @mborland in #633
• Bug fix for 1.86 by @mborland in #634

New Contributors

• @tchaikov made their first contribution in #618

Full Changelog: Boost_1_85_0...Boost_1_86_0

Standalone release of Boost.Multiprecision which can be used on it's own without the rest of Boost, and/or in conjunction with the Boost.Math standalone release.

Changes from 1.84:

• Fix deprecation warnings from C++23 PR#575
• Fix cpp_bin_float for small bit counts PR#577
• Fix -Wpessimizing-move PR#580
• Add covecov runs
• Reduce truncation in cpp_dec_float multiplication PR#585
• Fix complex_adaptor to match precision requirements PR#597
• Fix for breaking changes in tommath PR#603
• Fix for types that can be converted to number participting in arithmetic operator overloads #608

Can be used independently, or in conjunction with the Boost.Math standalone package.

• Change boost::array to std::array in the tests, fixes 567.
• Correct cpp_int::eval_convert_to to be noexcept, fixes 553.
• Make sure that changing the global precision, also changes the current thread precision.
• Fix rational_adaptor construction bug, fixes 540.

This is the standalone package of Boost.Multiprecision from Boost 1.83. Can be used independently, or in conjunction with the Boost.Math standalone package.

This is the standalone package of Boost.Multiprecision from Boost 1.82. Can be used independently, or in conjunction with the Boost.Math standalone package.

This is the standalone package of Boost.Multiprecision from Boost 1.81. Can be used independently, or in conjunction with the Boost.Math standalone package.

Standalone release of Boost.Multiprecision which can be used on it's own without the rest of Boost, and/or in conjunction with the Boost.Math standalone release.

Initial standalone release of Multiprecision.
This release can be used entirely independently from the rest of Boost, and/or in conjunction with a standalone Boost.Math release.