I completely agree with this. Use integers, avoids all sorts of problems. If you want to use only ONE number, use an integer to store number of minutes or seconds, and then use math to figure out how many hours, days, years etc the integer value corresponds to.
I'm also agreeing that comparing float/double with == is bad idea, since floating point numbers are not always PRECISE. In particular 0.1 or multiple [except 0.5] are similar to 1/3 in decimal numbers - it is NEVER going to be precise in [binary] floating point formats.
--
Mats