Readme update.

Readme update.

Readme update.

Readme update.

Changelog update.

Readme update.

Readme update.

Readme update.

Changelog update.

Version 2.11.3.

Readme update.

Log message arguments preparation refactored; tuple and pair elements are copied by value (recursively); character const arrays are considered to be string literals and passed as raw pointers, all the rest null-terminated strings are passed by copy (as it was in previous versions); added direct support for string views (passed by copy); added raw pointer value output as a hex address; version 0.99.30.

Readme update; note on MinGW-w64 GCC 16.1.0 compatibility added.

Readme update; removed misleading warning about functional broker task groups management for a rare specific case which is not relevant in case of a normal use of the task groups mechanism.

Version 2.11.3.

Readme update.

Log message arguments preparation refactored; tuple and pair elements are copied by value (recursively); character const arrays are considered to be string literals and passed as raw pointers, all the rest null-terminated strings are passed by copy (as it was in previous versions); added direct support for string views (passed by copy); added raw pointer value output as a hex address; version 0.99.30.

Readme update; note on MinGW-w64 GCC 16.1.0 compatibility added.

Readme update; removed misleading warning about functional broker task groups management for a rare specific case which is not relevant in case of a normal use of the task groups mechanism.

Processing thread affinity setup moved into thread's entry point prologue handler; fixed thread affinity setup when building in VS 2019, MinGW environments.

FFPP::Base::FunctionalQueueThread::GetNativeHandle() method refactored, fixed MinGW compatibility issue.

Target platform checks refactored (ffpp;:GetCurrentNativeThread, ffpp::SetThreadHardwareAffinity).

ffpp::ThreadHandleT type alias definition refactored, fixed MinGW compatibility issue.

Readme update.

External libraries fetch refactored wit the use of a plain `git clone` instead of a FetchContent (required by dlatency, actoqueues benchmarks and chains, requests demos).

Refactored task group finalization; fixed a leak when using the same task id for tasks with different context types, including scenarios involving ffpp::FunctionalBroker::Accept(); ffpp::FunctionalBroker::AllocateShared static method added.

External libraries fetch refactored (required by dlatency, actoqueues benchmarks and chains, requests demos).

External libraries fetch refactored (required by dlatency, actoqueues benchmarks and chains, requests demos).

Included external `utfcpp` library used in `chains` and `requests` demos for wstring to utf8 conversions (inside the Asynchrology logger).

Version 2.11.2 (2026-07-25).

Readme update.

Implemented optional support for encoding conversions with the use of utfcpp library (must be included above asynchrology.hpp, enabled by ASYNCHROLOGY_ENABLE_UTFCPP macro definition); ASYNCHROLOGY_ATTR_INLINE, ASYNCHROLOGY_ATTR_HOT_PATH macros refactored, checking for GCC and Clang first; version 0.99.29.

Included external `utfcpp` library used in `chains` and `requests` demos for wstring to utf8 conversions (inside the Asynchrology logger).

Version 2.11.2 (2026-07-25).

Readme update.

Implemented optional support for encoding conversions with the use of utfcpp library (must be included above asynchrology.hpp, enabled by ASYNCHROLOGY_ENABLE_UTFCPP macro definition); ASYNCHROLOGY_ATTR_INLINE, ASYNCHROLOGY_ATTR_HOT_PATH macros refactored, checking for GCC and Clang first; version 0.99.29.

Tuple concatenation refactored.

FFPP_ATTR_INLINE, FFPP_ATTR_HOT_PATH macros refactored, checking for GCC and Clang first.

Version 2.11.2.

Readme update.

FastMod class refactoring, hinting conditional move with the use of ternary operator; additional debug checks.

Added FFPP_DEFAULT_INTERFERENCE_SIZE macro for default resource policy and PRNG implementations.

Version 2.11.2.

Readme update.

Using FunctionalQueue::Size() method for thread pool's processing thread queue depth estimation when checking task offload conditions.

Optimized FunctionalLoop instance processing (functional broker tasks, functional chains) with the use of a fast modulo calculation; fixed possible race on FunctionalLoop instance finalization.

Enabled copy and move construction/assignment for FastMod class.

Version 2.11.1 (2026-07-11).

Readme update.

Fixed logger channels manual flush and completion; message level run-time check moved closer to the point of emission; using raw string data write in console output channel; using message representation as shared pointer to const string; version 0.99.28.

Fixed state initialization order of the ffpp::FunctionalQueueThread class (specialization with timer) and possible race during class instance construction, including the thread pool in dynamic mode.

Fixed functional broker's task group id existence check.

Version 2.11.1 (2026-07-11).

Readme update.

Fixed logger channels manual flush and completion; message level run-time check moved closer to the point of emission; using raw string data write in console output channel; using message representation as shared pointer to const string; version 0.99.28.

Fixed state initialization order of the ffpp::FunctionalQueueThread class (specialization with timer) and possible race during class instance construction, including the thread pool in dynamic mode.

Fixed functional broker's task group id existence check.

VS 2019 compatibility fix.

Thread lightweight id refactored with the use of constinit integral thread-local value; removed dtl::ThreadTopologyId.

Private thread local state of processing thread refactored with the use of a common struct (cache-aligned, trivial, constinit).

Removed c_bTopology_ constexpr flag double check.