Infinite cycle in case of small percent and negative aug, e.g.
p0=1000, percent=1.0, aug=-100

Nice kata, but there are 2 problems with it:

1. population may become 0 because aug may be negative, but kata doesn't have tests about that special case
2. need to guess about rounding strategy when calculating the yearly increase in percent:
   Let assume you receive 50.5 after calculation - its 51? or 50?

Hello! Your proposal looks ok. However, I saw more tests in original solution.
What about adding some random tests to your translation?

One problem with this solution - appending to new_s is not perfectly optimal, because strings in Python are immutable and every time you need to reallocate memory when adding next char or ' ' to new_s.