Version 1.4

Maximum optimization. Implemented loop unrolling and inlining so that each of the main work threads are about as efficient as I can make them.

Version 1.3: With Profile-Guided Optimization

Implemented Issue #33: Profile-Guided Optimization
Fixed an assert bug
Resolved all Code Analysis and Clang-Tidy warnings

Version 1.2: Redraw only necessary screen elements

Refactored mvcView, per Raymond Chen's advice in The Old New Thing, to trust WM_PAINT. So we refactored the entire drawing engine. Specifically, when we detect changes to the model (DTMF tones are turning on or off) we now invalidate that region of the screen. Then, we wait for WM_PAINT to run. This moved the entire screen redraw off of the Audio Capture Thread and to WndProc's message loop. This, in turn, reduced the number of DATA_DISCONTINUITY in Release mode to 1 (the initial audio packet).

Interim release for developing tagging and statistics content

1.1

Issue #18:  Add GitHub Statistics

Initial release

Here's the initial release of DTMF Decoder. There are still a few things that need doing, but it's good to get a 1.0 out there.