That's getting back to my point though - the integral is over measurable sets of reals, and we are approximating that with finite-difference floats. This is a c

That's getting back to my point though - the integral is over measurable sets of reals, and we are approximating that with finite-difference floats. This is a complex problem with many guises.