quote:Original post by Anonymous Poster
I don''t consider that as a counterexample. Getting the error to reach 0 would mean an infinite loop even with infinitely accurate numbers, so the example has in fact nothing to do with float''s precicion.
Not true. Not all iterative functions are asymptotic. Some are mathematically guaranteed to reach an exact certain number after n iterations, but progressive floating point error keeps that from happening.
Don''t listen to me. I''ve had too much coffee.