I still want to port WLED patterns, and I also want to update existing patterns now that we have so many tools/tricks to do things “better”
An example is KITT, the 1D array-using time-less method works and it’s a great tutorial, but having a mapped arrayless time-using pattern means that it works in many more cases/situations.
Now that the coordinate API is available, rewriting code to use it rather than doing x=x-.5 is a no brainer, but we also have scale and rotate…
I was just reviewing code for the “radial sound spectrometer” and realized that using the rotation API probably solved the problem I had with rotating only clockwise (because counter clockwise required more code to handle, so I skipped it in the name of posting something working at least one way)… Now rotation is dead simpler. Less code. Easier to understand.
Making patterns use time() means they can sync, and scale up. So it’s not just a matter of doing it different, it’s doing it “better”
Any other “done better” bits people can think of?
Any still outstanding that feel critical?