Version 8.4.0 Merge pull request #440 from fastfloat/indeterminate-bigint CI: run the Alpine workflow on ubuntu-24.04 Avoid zero-filling bigint buffers in the slow path (C++26 [[indeterminate]]) Merge pull request #433 from fastfloat/benchmark-rounds-to-nearest Benchmark fast_float with and without FASTFLOAT_ROUNDS_TO_NEAREST Merge pull request #432 from fastfloat/rounds-to-nearest-macro Add FASTFLOAT_ROUNDS_TO_NEAREST to skip the rounding-mode check Version 8.3.3 Merge pull request #430 from CODER-7777/fix-427-unlikely-cxx20-ext-warning Merge pull request #431 from aikhanjum/codex/fastfloat-correctness Reject integer_times_pow10 arguments wider than 64 bits Silence -Wc++20-extensions at the remaining [[unlikely]] sites (fixes #427) Run clang-format Merge pull request #426 from sahvx655-wq/power-negative-shift Avoid right-shifting a negative value in detail::power() Merge pull request #425 from fastfloat/fix-424-fixed-width-invalid Leave std::float32_t/std::float64_t unmodified on invalid input (fixes #424) Merge pull request #423 from sahvx655-wq/fuzz-coverage extend fuzzer to wide char types, integer parsing and all formats