A weird quirk with numerical for-loops and decimal increments. What's happening?

As some of you may know, doing a for loop like this:

would print:

because of floating-point rounding errors. However, when I set the start and end conditions to 0 and 1, like this:

it correctly terminates at 1:

This only seems to be the case when the increment is 0.1 (or when it's a multiple of 0.5^n) and start/end is 0, 1, respectively. When I set it to 0.01, it terminates at 0.99. Furthermore, when I work backwards, from 1 to 0 with a decrement of -0.1, it terminates at a very small number (that's not quite 0):

It looks like this only ever terminates correctly when the start, end, and increment is a multiple of 0, 1, and 0.1, respectively:

prints:

My question is: is this a quirk with this particular numeric for-loop, or is there something else going on?