I hope it's clear now why you shouldn't use floats...

Yes - version numbers are actually integers separated by dots. To increment, just ++ the final integer.

Cheers,
Rob