Coding 2026-05-27
All right. I vendored some code from scipy and its type stubs, and that was enough for me to reconstruct the environment without numpy constantly breaking, so I can work on the synthesis code.
However, I ended up not really feeling like doing any more with any of this tonight.
...
I say that, and then I get my code to synthesize a scale. Frankly shocking that this code more-or-less worked on the first try, given the level of rewrites and rearchitectures it went through, before I tried it all out just now. I think next I'll try to get it to work in real time, but that really will need to wait until later.
Good night.