Why can a floating point number be used as an array index?  Anybody
know of a good use case for this?

irb(main):020:0> [1, 2, 3][0.3]
1
irb(main):021:0> [1, 2, 3][0.9]
1

Just curious,
Jeff