Fast forward to today. Most of my current code is written in Go, and the largest chunk of it is a family of DSP libraries under the algo-* umbrella — the direct descendant of everything the retrospective posts described, rebuilt on a modern stack.
The core
algo-dsp is the heart: production-quality DSP algorithms for Go, deliberately algorithm-centric and transport-agnostic — no UI, no audio device, no file format dependencies. Just the math, tested and benchmarked. Underneath it sits algo-fft, an FFT with hand-written assembly where it counts — twenty years after I was optimizing blend loops in Graphics32 with SSE2, I am optimizing butterflies the same way.
The satellites
Around the core, the suite splits by domain: algo-vecmath for SIMD vector math, algo-pde for spectral PDE solvers, algo-approx for fast function approximations (the spiritual heir of those ten tanh() approximations from the SynthEdit days), and algo-acoustics for room acoustics.
Instruments
The newest branch is physical modeling and synthesis: algo-piano, algo-glockenspiel and a drum machine (algo-drum, in TypeScript, because instruments want a UI). The Fibonacci synthesizer I once wrote for a university publication would fit right in here — some threads really do run for twenty years.
Everything is developed in the open, as always: github.com/CWBudde.